The Ashgate Companion to the History of Textile Workers, 1650–2000 9780754664284, 9781315612683

Table of contents :
Cover
Contents
List of Figures
List of Maps
List of Tables
Notes on Contributors
1 Textile workers around the world,1650–2000: Introduction to a collective work project
PART I NATIONAL HISTORIES OF TEXTILES WORKERS
2 Textile production in Argentina, 1650–2000
3 Austria and Czechoslovakia: the Habsburg Monarchy and its successor states
4 Brazil: the origin of the textile industry
5 China
6 Denmark: the textile industry and the formation of modern industrial relations
7 Egyptian textile workers: from craft artisans facing European competition to proletarians contending with the state
8 The German wool and cotton industry from the sixteenth to the twentieth century
9 Great Britain: textile workers in the Lancashire cotton and Yorkshire wool industries
10 The long globalization and textile producers in India
11 The Italian textile industry, 1600–2000: labour, sectors and products
12 Japan
13 Mexican textile workers: from conquest to globalization
14 The Netherlands
15 Poland
17 Spain
18 The Ottoman Empire, 1650–1922
19 Turkey, 1922–2003
20 The evolution of the Uruguayan textile industry
21 USA: shifting landscapes of class, culture, gender, race and protest in the American Northeast and South
PART II INTERNATIONAL COMPARISONS
22 Global trade and textile workers
23 Proto-industrialization and industrialization and ‘modernity’ in a global perspective
24 The textile firm and the management of labour
25 Spatial division of labour, global interrelations and imbalances in regional development
26 How will we get our workers? Ethnicity and migration of global textile workers
27 Work floors under tension: working conditions and international competition in textiles
28 Gender and the global textile industry
29 Investigating identities within the global textile workforce
30 Institutions in textile production: guilds and trade unions
31 Covering the world: some conclusions to the project
Index

Citation preview

THE ASHGatE COMPaNION tO tHE HIStORY OF TEXtILE WORKERS, 1650–2000

This page has been left blank intentionally

The Ashgate Companion to the History of Textile Workers, 1650–2000

Edited by LEX HEERMa VaN VOSS ELS HIEMStRa-KUPERUS ELISE VaN NEDERVEEN MEERKERK International Institute of Social History, Amsterdam, The Netherlands

First published 2010 by Ashgate Publishing Published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business Copyright © 2010 Lex Heerma van Voss, Els Hiemstra-Kuperus, Elise van Nederveen Meerkerk and the contributors All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Lex Heerma van Voss, Els Hiemstra-Kuperus and Elise van Nederveen Meerkerk have asserted their right under the Copyright, Designs and Patents Act, 1988, to be identified as the editors of this work. Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data The Ashgate companion to the history of textile workers, 1650–2000. 1. Textile workers – History – 17th century. 2. Textile workers – History – 18th century. 3. Textile workers – History – 19th century. 4. Textile industry – History – 17th century. 5. Textile industry – History – 18th century. 6. Textile industry – History – 19th century. I. History of textile workers, 1650–2000 II. Heerma van Voss, Lex. III. Hiemstra-Kuperus, Els. IV. Nederveen Meerkerk, Elise van. 338.4’7677’009–dc22 Library of Congress Cataloging-in-Publication Data The Ashgate companion to the history of textile workers, 1650–2000 / edited by Lex Heerma van Voss, Els Hiemstra-Kuperus, and Elise van Nederveen Meerkerk. v. cm. Includes bibliographical references and index. ISBN 978-0-7546-6428-4 (hardcover : alk. paper) 1. Clothing workers—History. 2. Textile workers—History. I. Heerma van Voss, Lex. II. Hiemstra-Kuperus, Els. III. Nederveen Meerkerk, Elise van. HD8039.C6A74 2009 331.7’61677009—dc22 ISBN 9780754664284 (hbk) ISBN 9781315612683 (ebk)

2009007558

Contents List of figures   List of maps   List of tables   Notes on contributors   1 ��������������������������������������������������������������� Textile workers around the world, 1650–2000: introduction to a collective work project   Els Hiemstra-Kuperus, Lex Heerma van Voss and Elise van Nederveen Meerkerk PART I

ix xi xiii xvii 1

NATIONAL HISTORIES OF TEXTILE WORKERS

2 ��������������������������������������������� Textile production in Argentina, 1650–2000   Mirta Zaida Lobato 3 �������������z�������������������������������������������� Austria and Czechoslovakia: the Habsburg Monarchy and its successor states   Andrea Komlosy Brazil: the origin of the textile industry  4 ���z���������������������������������������� Roberta Marx Delson

17

43 75

5 �������� China  103 Robert Cliver 6 ������������������������������ Denmark: the textile industry and the formation of modern industrial relations  Lars K. Christensen

141

7 �������������������������� Egyptian textile workers: from craft artisans facing European competition to proletarians contending with the state   Joel Beinin

171

The Ashgate Companion to the History of Textile Workers 8

The German ���� ������� wool ��������������������������������������������������� and cotton industry from the sixteenth to the twentieth century  Dietrich Ebeling, Marcel Boldorf, Stefan Gorißen, Michael Mende, Anke Sczesny and Michaela Schmölz-Häberlein

9 Great ������ Britain: ��������� textile �������� workers ��������������� in the Lancashire ���������������������� cotton and Yorkshire wool industries  Alan Fowler 10

231



253

11 ������������������������������������������������������������� The Italian textile industry, 1600–2000: labour, sectors and products  Giovanni Luigi Fontana, Walter Panciera and Giorgio Riello

275

12

The long globalization and textile producers in India Tirthankar Roy

199

Japan  J������� 305 Janet Hunter and Helen Macnaughtan

13 ���������������������������������q���������������z�������� Mexican textile workers: from conquest to globalization  333 Jeffrey Bortz 14 ������������������ The Netherlands  363 Elise van Nederveen Meerkerk, Lex Heerma van Voss and Els Hiemstra-Kuperus 15 ��������� Poland  397 Piotr Franaszek 16

������������������������������� The cotton textile industry in Russia ��������������� and the Soviet ������� Union 421 ������� � Dave Pretty

17 �������Spain  Angel Smith, Carles Enrech, Carme Molinero and Pere Ysàs 18

449

���� Ottoman The ��������������������������� Empire, 1650–1922 477 � Donald Quataert

19 �������������������� Turkey, 1922–2003  497 Lisa A. Seidman 20 �������������������������������������������������� The evolution of the Uruguayan textile industry  511 María Magdalena Camou and Silvana Maubrigades

vi

Contents 21

USA: shifting ����� ���������������������������������������������������������������� landscapes of class, culture, gender, race and protest in the American Northeast and South  Mary H. Blewett

531

PART II INTERNATIONAL COMPARISONS 22 ������������������������������������ Global trade and textile workers  561 Prasannan Parthasarathi 23 �����������������z���������������������z���������� Proto-industrialization and industrialization and ‘modernity’ ‘����������� in a global perspective  Donald Quataert 24

577

The ����������������������������������������������� textile firm and the management of labour 597 �� Arthur McIvor

25 �������������������������������������������������������������������� Spatial division of labour, global interrelations and imbalances in regional development  Andrea Komlosy

621

26 ���������������������������? How will we get our workers? Ethnicity ���������������������������������� and migration of global textile workers  Roberta Marx Delson

647

27

Work ����� floors ���������������������� under tension: ���������������������������������������� working conditions and international competition in textiles  679 Peter Scholliers

28 ������������������������������������������ Gender and the global textile industry  703 Janet Hunter and Helen Macnaughtan Investigating identities within the global textile workforce  725 29 ���������������������������������������������������������������� Mary H. Blewett 30 ��������������������������������������������������������������� Institutions in textile production: guilds and trade unions  749 Lars K. Christensen 31 ������������������������������������������������������� Covering the world: some conclusions to the project  773 Elise van Nederveen Meerkerk, Lex Heerma van Voss and Els Hiemstra-Kuperus Index

793

vii

This page has been left blank intentionally

List of figures 4.1 Spinning wheel for cotton, 1784 85 4.2 Engine for ginning cotton, 1784 86 4.3 Loom used by Indian women in the village of Monte Alegre, 1785 86 4.4 Cambeba Indian male with his armament 91 4.5 Clothed female Indians of an unknown tribe, c.1787 93 14.1 Production of pieces of woollen cloth in Leiden, 1650–1810 367 14.2 Spinners and weavers in Tilburg 1810, according to age and sex 377 14.3 Strikes and lockouts in the Dutch textile industry, 1800–2000 386 15.1 Employment in the Polish textile industry, 1870–1913 404 15.2 Women and men employed in the Łódź textile industry, 1921–1936409 15.3 Production of yarn in Poland, 1946–1989 413 15.4 Production of fabrics in Poland, 1946–1989 414 15.5 Employment in the Polish textile industry, 1946–1989 415 15.6 Production of yarn in Poland, 1989–2000  419 15.7 Production of fabrics in Poland, 1989–1999 419 15.8 Employment in the Polish textile industry, 1989–2000 420 20.1 Textiles as percentage of total value added in manufacturing, 1900–2000 513 20.2 Labour force structure in La Industrial, 1918–1951 520 25.1 The synchronicity of the non-synchronous 624

This page has been left blank intentionally

List of maps 2.1 Argentina 3.1 The Habsburg Monarchy 1815–1918 4.1 Brazil 5.1 China 6.1 Denmark 7.1 Egypt 8.1 German textile areas 9.1 Textile districts of Northern England 10.1 India 1939, including locations cited in the text 11.1 Italy 11.2 Northern Italy 12.1 Map of Japan, showing major areas of textile production 13.1 Mexico 14.1 Netherlands 14.2 Woollen weavers’ guilds in 1650 (left) and 1780 (right) 15.1 Poland 16.1 Russia 17.1 Regions of Spain 17.2 Catalonian textile regions 18.1 Textile centres in the Ottoman Empire, c.1650–1914 19.1 Turkey 20.1 Textile regions in Uruguay 21.1 Textile regions of the USA

18 44 76 104 142 172 200 232 254 276 276 306 334 364 373 398 422 450 450 478 498 512 532

This page has been left blank intentionally

List of tables 1.1 Cotton and wool production figures for countries with more than a 1.5 per cent share in world production in the mid-1930s, and other countries involved in the project 12 2.1 Value added of textiles and other products, Argentina 1900–1950 26 2.2 The cotton spinning industry in Argentina, 1920–192927 2.3 Employment in the Buenos Aires textile industry, 1929–1945 30 2.4 Strikes and union meetings in the Buenos Aires textile industry, 1935–1945 32 2.5 Textile industry: workers and production, Argentina, 1939–1963 33 2.6 Evolution of the textile industry in Argentina, 1914–1939 34 2.7 Physical volume of Argentinean textile production, 1971–1981 38 2.8 Indicators of activity in the textile industry, 1974–1989 39 3.1 Industrial workers in Lower Austria, 1762–1811 54 6.1 Cotton weavers operating one or several looms, 1906, 1913 and 1937 148 6.2 Industrial textile workers, 1872–1993 149 6.3 Women as percentage of total workforce in different textile branches and jobs 165 8.1 Regional division of spindles in main German cotton-producing states 203 8.2 Shares of German states in cotton weaving and spinning, 1846/1861203 8.3 Looms weaving wool in the German tariff union and Hanover205 8.4 Percentage distribution of fine spindles in wool spinning206 8.5 Spindles, production, workers and added value in cotton spinning209 8.6 Looms, production, workers and added value in cotton weaving210 8.7 Output volume of the fine drapery industry in Monschau, 1775–1794212 8.8 Export shares of single sectors in the textile industry 214 8.9 Percentages of self-actor spindles, and water and ring spindles 217 8.10 Wages in cotton spinning, 1815–1913226 10.1 Scale of the Indian textile industry, 1920 and 1940 266 312 12.1 Output of machine reeled silk, 1895–1940 12.2 Output of integrated cotton mills (spinning and weaving), 1895–1935 312 12.3 Share of textiles in total Japanese exports (by value) 313 12.4 Textile workers by size of factory/workshop, 1920–1940 315 12.5 Employment in private textile factories, 1896–1920 318

The Ashgate Companion to the History of Textile Workers 12.6 Textile workers by branch, 1920–1940 318 12.7 Textile mills and workers employed, 1948–1984 (excluding apparel) 323 12.8 Value of textile exports, 1945–1980 323 12.9 Production and trade in textiles, 1980–2000 325 13.1 Textile workshops and looms in New Spain, 1759 336 13.2 Official estimate of textile industry in New Spain, 1801 337 13.3 Growth of the Mexican cotton textile industry, 1843–1927 339 13.4 Regional percentage share of national production of cotton textiles, 1880–1910 341 13.5 Gender composition of cotton textile industry labour force, 1927 349 13.6 Mexican cotton textile industry, average daily wage for men 351 13.7 Employment in the Mexican cotton textile industry, 1943–1950 357 13.8 Mexican textile production, 1985–2000 360 14.1 Wool weavers’ guilds in the Dutch Republic, 1600–1800 366 14.2 Weavers and spinners in the Dutch Republic, 1650–1810 369 14.3 Division of labour in spinning according to sex 374 14.4 Gender composition of the Dutch textile labour force, 1920 and 1930 380 14.5 Cotton spinning and weaving in the Netherlands, 1861–1910 381 14.6 Spinning and weaving in the Netherlands, 1921–1950 383 15.1 Textile industry workers in Galicia, 1890 and 1910 407 15.2 Textile industry workers in Poland and in Łódź, 1921–1937408 15.3 Workers in textile firms according to sex and age, Łódź 1922–1936410 16.1 Russian textile production, 1860–1912 430 16.2 Russian and Soviet textile production, 1913–1940 442 16.3 Soviet and Russian textile production, 1940–2003 447 17.1 Catalan textile industries in 1784 454 17.2 Spindles in the Catalan cotton textile industry, 1807–1929457 17.3 Workers in the Spanish cotton textile industry in 1867460 17.4 Distribution of the Catalan cotton textile labour force, 1850–1919463 17.5 Development of the Spanish cotton textile industry, 1942–1979473 17.6 Active workforce in the Spanish textile industry, 1958–1997 474 17.7 Conflicts in the Spanish textile industries, 1966–1976475 20.1 Percentage share of the textile industry in manufacturing, Uruguay, 1960–1982 513 20.2 Work histories of skilled workers, Campomar Company, 1949 521 20.3 Workers employed in textile industry and manufacturing, 1908–1997 524 21.1 Persons employed in cotton goods, Fall River, Lowell and Lawrence, 1875 540 21.2 Southern competition and the New England cotton industry, 1900–1905 547 21.3 Impact of rising tariff protection on US wool manufacturing, 1859–1919 548 xiv

List of tables 21.4 Growth of the Southern cotton industry, 1880–1900551 24.1 World’s leading producers of synthetic fibres, 1979 614 31.1 Comparative costs in cotton weaving, nine countries, c.1910 783 31.2 Relative high points of labour unrest in the world textile industry, 1870–1996 788

xv

This page has been left blank intentionally

Notes on contributors Joel Beinin (b.1948) studied Middle East studies and history at Princeton, Harvard and the University of Michigan. His Ph.D. was revised and published as Workers on the Nile: Nationalism, Communism, Islam, and the Egyptian Working Class, 1882–1954 (Princeton University Press, 1988; co-authored with Zachary Lockman). His most recent book is Workers and Peasants in the Modern Middle East (Cambridge University Press, 2001). He is currently Director of Middle East Studies and Professor of History at the American University in Cairo. Mary H. Blewett (b.1938) studied twentieth-century American political history and foreign policy at the University of Missouri. Her Ph.D. analysed the transition between the Roosevelt and Truman administrations. Caught up in the political crises of the 1960s and 1970s she retrained herself in the labour and social history of industrial New England. She is now retired from the University of Massachusetts, and her most recent book is Constant Turmoil: The Politics of Industrial Life in Nineteenth-Century New England (University of Massachusetts Press, 2000). She is currently working on Migration Lived and Imagined (forthcoming, University of Illinois Press). Marcel Boldorf (b.1965) studied history in Saarbrücken. His Ph.D. was about social welfare in East Germany after 1945, his Habilitationsschrift on the development of European linen regions. He has published on social policies, industrialization and migration. Jeffrey Bortz earned his Ph.D. at UCLA. He has published extensively on the history of wages in Mexico and also on textile workers during the Mexican revolution. He is professor of history at Appalachian State University. María Magdalena Camou (b.1953) studied history in Berlin. Her Masters Degree in Economic History was a case study of a textile factory in Uruguay. She is a researcher at the Economic History Department of the Universidad de la República of Uruguay and holds a chair in State and Economic Development. Nowadays her research topics are the labour market and the international comparison of the evolution of wages and gender. In these fields she has published many books and articles.

The Ashgate Companion to the History of Textile Workers Lars K. Christensen (b.1962) is curator and senior researcher at the Danish National Museum, department of Modern Danish History. He has an MA in history and philosophy. The topic of his Ph.D. was cultural and institutional changes in the Danish textile industry, 1895–1940. His publications include a book and several articles on industrial labour history, as well as on museology. He is a former chairman of the Danish Society for Research in Labour History. Robert Cliver (b.1969) received his Ph.D. in history from Harvard University. He is a specialist in the history of labour and industry in twentieth-century China, focusing on the 1950s. His dissertation research concerns Yangzi Delta silk workers and their employers from the end of the Second World War through the first decade of the People’s Republic of China. He currently teaches East Asian history at Humboldt State University in California. Roberta Marx Delson studied international relations, anthropology and history at Syracuse and Columbia universities. Her Ph.D. dissertation was on urban planning in colonial Brazil. She is a Research Associate in Anthropology at the American Museum of Natural History in New York City and an adjunct Associate Professor of History at Drew University (Madison, NJ). She has published on Caribbean and Brazilian economic history and Brazilian urbanism. Dietrich Ebeling (b.1950) studied history and sociology in Bielefeld. His Ph.D. was devoted to the crisis of the imperial cities in the eighteenth century, his Habilitationsschrift was on the trading system in western Germany between the sixteenth and nineteenth centuries. He has published on different aspects of the economic and social history of the early modern period. Carles Enrech (b.1966) studied history in Barcelona. His Ph.D. was devoted to the Catalan textile industry and labour in mills from 1881 to 1923, and it received the Jaume Vicens Vives award in 2003. Nowadays he is researching textile labour, wages and gender at Treball, Institucions i gènere (TIG) at Barcelona University. He has published three books and several papers, mostly about the textile industry and labour. Giovanni Luigi Fontana is Professor of Economic History at the University of Padua. He is the author of several books and articles on industrial, entrepreneurial and institutional economic history. He recently edited with Gérard Gayot a volume entitled Wool: Products and Markets (13th–20th Century) (Padua, 2004). He is President of the Italian Association for the Conservation of Industrial Heritage and Director of the MA Programme in Preservation, Management and Improvement of Industrial Heritage at the University of Padua.

xviii

Notes on contributors Alan Fowler (b.1944), having studied at the London School of Economic, taught Economic and Social History at Manchester Polytechnic/Manchester Metropolitan University from 1968 onwards. His undergraduate interest in labour history and subsequent teaching appointment coincided with the final decline of the Lancashire cotton industry. He was fortunate enough to meet and talk to a number of trade union officials and activists whose involvement in cotton went back to the 1920s and 1930s. He has written extensively on the Lancashire cotton operatives and their trade unions, his latest work being Lancashire Cotton Operatives and Work 1900–1950 (Ashgate, 2003). Now retired, he continues with his interest in researching the history of the Lancashire cotton industry. Piotr Franaszek (b.1955) studied history at the Institute of History of the Jagiellonian University in Krakow. His Ph.D. was devoted to the history of the oil industry at the turn of the nineteenth century. In 2003 he received a professorship and the Chair of Economic and Social History at the Jagiellonian University. His research has concentrated on the history of industry and agriculture in the nineteenth and twentieth centuries. Stefan Gorißen (b.1960) studied history in Bonn and Bielefeld. His Ph.D. was devoted to proto-industrialization and the history of trade in eighteenth-century southern Westphalia. He is research fellow at the University of Bielefeld and has published on the regional history of Rhineland and Westphalia, mainly in the eighteenth century. Lex Heerma van Voss (b.1955) studied history in Utrecht and Paris. His Ph.D. was devoted to the introduction of the eight-hour working day in the Netherlands. He is a research fellow at the International Institute of Social History in Amsterdam and holds a chair in the history of labour and labour relations at Utrecht University. He has published on the comparative history of dockworkers and on the history of the North Sea. Els Hiemstra-Kuperus (b.1966) studied history at the University of Groningen. At the International Institute of Social History in Amsterdam she is responsible for the organization of the bi-annual European Social Science History Conference (ESSHC). She has also contributed to a publication on the image collection of the IISH. Janet Hunter is Saji Professor of Economic History at the London School of Economics. She obtained her Ph.D. from Oxford University, and then taught at the University of Sheffield before moving to the LSE. She has published widely on the economic and social history of modern Japan, and particularly on the development of the female labour market and the history of Anglo-Japanese economic relations. Her book on textile workers, Women and the Labour Market in Japan’s Industrialising Economy: The Textile Industry before the Pacific War, was published by RoutledgeCurzon in 2003, and a Japanese version will be published by Yuhikaku. xix

The Ashgate Companion to the History of Textile Workers Andrea Komlosy (b.1957) studied history and political science in Vienna. Her dissertation (1984) was devoted to the economic and social history of a Lower Austria Region, the Waldviertel, analysing the reason for its peripheralization. Komlosy’s Habilitationsschrift (2001) dealt with migration and single market formation in the Habsburg Monarchy, looking at internal borders, which during the eighteenth and nineteenth centuries were abolished for the transport of goods, but established to control the mobility of people. Komlosy is professor at the Institute for Economic and Social History, University of Vienna. She has published on problems of uneven development on a regional, a European and a global scale. Mirta Zaida Lobato (b.1948) received her Ph.D. in History from the University of Buenos Aires. She is a Full Professor and researcher in the Facultad de Filosofía y Letras of the University of Buenos Aires. Her research work discusses both the formation of working-class culture in Argentina in the twentieth century and gender relations in the working world. She is the author of Un jirón del hogar abandonado. Trabajo femenino en Argentina, 1869–1960 (Edhasa, 2007), La vida en las fábricas. Trabajo, protesta y conflicto en una comunidad obrera. Berisso, 1904–1970 (Prometeo libros, 2001 and 2004) and co-author of La protesta social en Argentina (FCE, 2003) and Atlas Histórico de la Argentina (Sudamericana, 2000). Arthur McIvor (b.1956) studied history in Nottingham and Manchester. His Ph.D. examined employers’ organizations and industrial relations in the cotton textile region of north-west England. He is a Reader in History at the University of Strathclyde, Glasgow. He has published on the history of employers and management in textiles and on the history of work, occupational health and industrial relations in the UK in the nineteenth and twentieth centuries. Helen Macnaughtan is Lecturer in International Business and Management for Japan at the School of Oriental and African Studies, London. She received her Ph.D. in 2001 from the London School of Economics, and has been teaching at SOAS since 2002. She is the author of Women, Work and the Japanese Economic Miracle: The Case of the Cotton Textile Industry, 1945–1975 published by RoutledgeCurzon in 2005. Her areas of research interest include employment, gender and economic development in Japan. Silvana Maubrigades (b.1973) studied sociology in Montevideo. Her MA in Economic History was about female participation in the industrial sector in Uruguay during the twentieth century. She is a researcher in the Economic History Department of the Universidad de la República of Uruguay. She has published numerous articles in different publications about the labour market, gender and inequality.

xx

Notes on contributors Michael Mende (b.1945) studied art history and politics in Berlin. His Ph.D. (Technical University) was devoted to the education and vocational training of both technicians and skilled workers in mechanical engineering. At the Braunschweig University of Arts he holds a chair in the history of architecture and design. He has published on the history of technology and industrial architecture. Carme Molinero is a lecturer in modern history at the Universitat Autònoma de Barcelona. Together with Pere Ysàs she has contributed to over sixty works and they have jointly published, amongst others, the following books: ‘Patria, Justicia y Pan’. Nivell de vida i condicions de treball a Catalunya, 1939–1951 (1985), Els industrials catalans durant el franquisme (1991) and Productores disciplinados y minorías subversivas. Clase Obrera y conflictividad laboral en la España franquista (1998). Carme Molinero has also written La captación de las masas. Política social y propaganda en el régimen franquista (2005). Elise van Nederveen Meerkerk (b.1975) studied economic and social history at Utrecht University. Her Ph.D. thesis was on female textile workers in the seventeenth- and eighteenth-century Dutch Republic and was awarded at the International Institute of Social History in Amsterdam. After a period as a postdoctorate researcher at Leiden University working on civil services and citizenship, she has now returned to the International Institute of Social History to work on the history of charity in the early modern Netherlands. Walter Panciera is Associate Professor of History at the University of Padua. He has written several books and articles on the economic and social history of the Veneto region in the early modern period and on woollen manufacturing in the seventeenth and eighteenth centuries. Among his recent publications are ‘L’Arte Matrice’. I Lanifici della Repubblica di Venezia nei Secoli XVII e XVIII (Treviso, 1996) and Fiducia a Affari nella Società Veneziana del Settecento (Padua, 2000). Prasannan Parthasarathi (b.1961) studied economics and economic history at Williams College and Harvard University. His Ph.D. dissertation was a study of relations between weavers and merchants in the late pre-colonial and early colonial period in South India. He teaches history at Boston College (USA). He has also published on comparative labour history in the eighteenth century and on global economic history. Dave Pretty (b.1958) has degrees in history from the University of California at Santa Cruz, Columbia University and Brown University. His dissertation on the textile workers of the Ivanovo-Voznesensk region between 1885 and 1905 focused on modes of worker activism. His current research interests remain with lower class political culture, including the political uses of popular violence. He is assistant professor of history at Winthrop University in Rock Hill, South Carolina.

xxi

The Ashgate Companion to the History of Textile Workers Donald Quataert (Harvard MA 1968, UCLA Ph.D. 1973) is Professor of History at Binghamton University and has been teaching at Binghamton University since 1987. He is the author of six monographs and the editor or co-editor of another 10 books dealing with various aspects of the history of the Ottoman Empire. His most recent publications are History of the Ottoman Empire, 1700–1922 (Cambridge, 2000 and 2005) and Coal Miners and the State in the Ottoman Empire: The Zonguldak Coalfield, 1822–1920 (New York, 2006). He is the holder of grants from the Social Science Research Council, the National Endowment for the Humanities and the John Simon Guggenheim Memorial Foundation. Giorgio Riello was Research Officer in Global History at the London School of Economics and is now Lecturer in History at the University of Exeter. He has published on fashion, material culture and product innovation in the eighteenth and nineteenth centuries. He is the author A Foot in the Past: Consumers, Producers and Footwear in the Long Eighteenth Century (Oxford, 2006) and the co-editor of Shoes: A History from Sandals to Sneakers (Oxford and New York, 2006). He is currently writing a book entitled A Global History of Cotton Textiles, 1200–1800. Tirthankar Roy (b.1960) studied economics in India, and has done a Ph.D. on the history of the hand-loom weaving industry in India. He has published on the textile and labour history of South Asia, and teaches economic history at the London School of Economics and Political Science. Michaela Schmölz-Häberlein (b.1964) studied history at the universities of Augsburg, Guatemala-Ciudad and Freiburg. Her Ph.D. dissertation discusses the continuities and changes of living and working conditions of Kekchi and Pokomchi Indians in Guatemala during the Great Depression. She is a historian living in Bamberg and has taught history classes at Freiburg, Karlsruhe, Mannheim and Pennsylvania State universities. She has published on the history of Latin America, the Caribbean and on the social history of early modern Germany. Peter Scholliers (b.1953) studied history at the Vrije Universiteit Brussel, where he obtained his Ph.D. in 1984 (Purchasing power and index-linked wages in interwar Belgium). He is professor of contemporary history in Brussels, teaching and studying the history of labour, industrial relations, living standards and, particularly, food. Recent publications include Experiencing Wages: Social and Cultural Aspects of Wage Forms in Europe since 1500 (ed. with L. Schwarz) (Oxford, 2003). Anke Sczesny (b.1963) studied history in Augsburg and took her doctor’s degree there. She has published on the rural textile industry in Ostschwaben in the seventeenth and eighteenth centuries and on family and television. She is currently working as academic coordinator in the Institut für Europäische Kulturgeschichte at the University of Augsburg.

xxii

Notes on contributors Lisa A. Seidman (b.1962) studied economics and agricultural economics at the University of Illinois and Michigan State, and labour economics at Cornell University. Her Ph.D. is a study of female labour force participation in Denizli, Turkey. She designed and managed the first household level panel data survey conducted in Turkey (the Denizli Panel Household Survey of 1,200 households) in collaboration with the Turkish State Institute of Statistics. Prior to attending Cornell, she worked extensively in East Africa as an agricultural economist. Her Master’s thesis was on agricultural extension institutions in East Africa and she has published various articles and working papers on this topic. Angel Smith (b.1958) studied history in Manchester and at Queen Mary College, London. His Ph.D. focused on labour and anarchism in early twentieth-century Catalonia. He is senior lecturer in the Department of Spanish, Portuguese and Latin American Studies at the University of Leeds. His most recent work is Anarchism, Revolution and Reaction: Catalan Labour and the Crisis of the Spanish State (Oxford and New York, 2007). Pere Ysàs is a lecturer in modern history at the Universitat Autònoma de Barcelona. Together with Carme Molinero he has contributed to over sixty works and they have jointly published, amongst others, the following books ‘Patria, Justicia y Pan’. Nivell de vida i condicions de treball a Catalunya, 1939–1951 (1985), Els industrials catalans durant el franquisme (1991) and Productores disciplinados y minorías subversivas. Clase Obrera y conflictividad laboral en la España franquista (1998). Pere Ysàs is also the author of Disidencia y subversión. La lucha del régimen franquista por su supervivencia, 1960–1975 (2004).

xxiii

This page has been left blank intentionally

1 Textile workers around the world, 1650–2000: introduction to a collective work project Els Hiemstra-Kuperus, Lex Heerma van Voss and Elise van Nederveen Meerkerk

The relevance of a global history of textile production over a long period of time is clear: textile products cater for a basic human need, they are among the most important goods fabricated and traded by mankind and have thus played a central role in human activities throughout history. It is therefore no wonder that historians have paid so much attention to the basic processes in producing textiles: spinning and weaving. Numerous regional and national studies on developments in the production of and trade in textiles have been published. Moreover textiles have also been at the centre of several crucial historical debates. Theories on protoindustrialization, the Industrial Revolution, technological and business history, the history of taste and fashion and the gendered division of labour often take the textile industry as a point of reference.  On each of these issues there is a voluminous debate, to which justice cannot be done in a simple footnote, so we confine ourselves to pointing to F.F. Mendels, ‘Protoindustrialization: The First Phase of the Industrialization Process’, Journal of Economic History 32 (1972): 241–261; P. Kriedte, H. Medick and J. Schlumbohm, Industrialization before Industrialization: Rural Industries in the Genesis of Capitalism (Cambridge, 1981); Sheilagh Ogilvie and Markus Cernan (eds), European Proto-Industrialization (Cambridge, 1996); Hans Medick, Weben und Überleben in Laichingen, 1650–1900: Lokalgeschichte als allgemeine Geschichte (Göttingen, 1996); Peter Kriedte, Hans Medick and Jürgen Schlumbohm, ‘Eine Forschungslandschaft in Bewegung. Die Proto-Industrialisierung am Ende des 20. Jahrhunderts’, Jahrbuch für Wirtschaftsgeschicht, 2 (1998): 9–20; D.T. Jenkins (ed.), The Textile Industries (The Industrial Revolutions, Vol. 8) (Oxford and Cambridge, Mass., 1994); Mary B. Rose, Firms, Networks and Business Values: The British and American Cotton Industries since 1750 (Cambridge, 2000); Chandra Mukerji, From Graven Images: Patterns of Modern Materialism (New York, 1983); Daniel Miller

The Ashgate Companion to the History of Textile Workers We have also long known that textile labour history was potentially very international. There are several ways to explore this international characteristic. One can follow the commodity chain from cotton or wool as a core raw material to yarn to cloth to clothing and other finished textile products. This means that a change in the labour conditions at any point in the chain can have ramifications along it, or in other parts of the world, as a link in the chain is replaced, as became clear during the worldwide cotton famine induced by the American Civil War. One can also focus on the competition between producers, which had acquired a global scale by the turn of the eighteenth century, as mechanized production in Britain competed with highly productive hand-weaving in India. If the aim of the research is to cover a long period of time and at least a considerable part of the spatial diversity, nowadays textile history usually becomes teamwork. Such a joint project was undertaken by the International Institute of Social History (IISH, Amsterdam, The Netherlands) with the aim of taking stock of the work done on textile workers and comparing their history internationally over a period of 350 years. To realize this broad comparison, the project follows the approach of an earlier comparative project on the history of dock workers. In the first stage a group of specialists was asked to write national histories on textile work in different countries covering the period 1650–2000. This method has an advantage over having one historian or a small team doing comparative work on a number of countries in that it can build on existing expertise. However, if one asks experts from different national historiographies to write on the area of their expertise, all these contributions tend to have their own questions, logic and – therefore – their own answers. To reach a certain degree of comparability, these (ed.), Acknowledging Consumption: A Review of New Studies (London and New York, 1995); Peter N. Stearns, Consumerism in World History: The Global Transformation of Desire (London and New York, 2001); Gertjan de Groot and Marlou Schrover (eds), Women Workers and Technological Change in Europe in the Nineteenth and Twentieth Centuries (London and Bristol, 1995).  Sven Beckert, ‘Emancipation and Empire: Reconstructing the Worldwide Web of Cotton Production in the Age of the American Civil War’, American Historical Review 109 (2004): 1405–1438.  Louise A. Tilly, ‘Connections’, American Historical Review 99 (1994): 1–20; Prasannan Parthasarathi, ‘Rethinking Wages and Competitiveness in the Eighteenth Century: Britain and South India’, Past and Present 158 (1998): 79–109.  An impressive example in textile history, not specifically textile labour history, is David Jenkins (ed.), The Cambridge History of Western Textiles (2 vols, Cambridge, 2003).  Sam Davies et al. (eds), Dock Workers: International Explorations in Comparative Labour History (2 vols, Aldershot, 2000).  See Marcel van der Linden, ‘Doing Comparative Labour History: Some Preliminaries’, in Marcel van der Linden, Transnational Labour History: Explorations (Aldershot, 2003) pp. 173–196, for this form of organization of comparative research, and some alternatives.  An example of this is Douglas A. Farnie and David J. Jeremy (eds), The Fibre that Changed the World: The Cotton Industry in International Perspective, 1600–1990s (Oxford, 2004). 

Textile workers around the world, 1650–2000 national histories are based on a questionnaire, which in the project was dubbed the framework document. This listed a number of topics and questions to be studied in each overview. It is reproduced here as the appendix to this introduction. The national textile workers’ histories make up the first part of this book. The second part consists of comparative studies that have been written on the basis of these national histories. These comparative studies were vividly discussed at a productive conference in Amsterdam in November 2004 attended by all the authors involved in the project. This created the opportunity to suggest improvements to the national overviews and correct factual errors in the interpretation of national overviews by the authors of comparative papers. But most importantly, it brought together experts and information, which stimulated all those present to think about largescale developments. It made us acutely aware of the huge amount of variety in time and space, but it also enabled us to grasp at least some trends evident in this variety.

Choices and boundaries 1: fibres and work processes Organizing the project meant that we had to make a number of choices. The first one concerned the fibres to be taken into account. We have focused on the production of cotton and wool. Many natural and artificial fibres have been spun into yarn and woven into cloth over the past centuries and for those national overviews where other fibres are crucial, they have been included, but wool and cotton have been the most important fibres if we take global production over the entire period from 1650 until recent times into consideration. There was also a related practical reason. The prominence of cotton and wool in textile production guaranteed there would be enough expertise to write national overviews, but we foresaw that our authors would not be specialists on the whole of the 1650–2000 period. They would therefore have to base their research for the other periods on the existing literature. This was another reason to focus on the most prominent fibres, and also on the simple production processes, as these are better represented in the existing literature. In some cases the production of other fabrics, especially silk, was so important that we agreed with the specialist that it was best to include it in the national overview.

Choices and boundaries 2: period and waves of globalization As we have already mentioned, we aim to cover the period 1650–2000. This long period has been chosen to include the processes of industrialization and de-industrialization in all parts of the world. In the pre-industrial phase, market production of textiles was largely undertaken by part of the rural population on farms or by artisans in urban workshops. In some regions urban-based weavers 

The Ashgate Companion to the History of Textile Workers and entrepreneurs sought cheaper means of production, by employing waged weavers and spinners in the countryside to produce for them. The specific nature of this ‘proto-industrialization’ has been debated, especially in relation to textile production. Don Quataert, in his comparative paper (Chapter 23), highlights that one of the conclusions of this project should be that ‘proto-industrial’ is too teleological an adjective, as it suggests a phase of industrialization will necessarily follow and in practice this was often not the case. In the search for cheaper cloth production, technical improvements were made to enlarge spinning and weaving production from the 1760s onwards. Cotton and wool were among the first raw materials to be processed industrially after the introduction of new machinery. The ‘Industrial Revolution’ created immensely wealthy textile employers and condemned the actual producers to work on ever more numerous and larger machines in ever more efficient factories. The living conditions of textile workers and their efforts to organize themselves in order to improve their situation have had an immense impact on the debates in social and economic history about the consequences of industrialization. Specific stages of textile production, whether in domestic, artisan or factory industry, were almost invariably divided between men and women. In some cases spinning was performed exclusively by women, in other cases by men. Certain kinds of products were only made by one sex; others were exclusively made by the other sex. Sometimes specific machines were only operated by one sex. Usually female workers (and children of both sexes) earned far less than men, even when performing exactly the same tasks. These and other aspects of the gendered nature of textile production seem to be persistent, and therefore can probably only be explained by a long-term historical analysis that compares these mechanisms in different parts of the world. Textiles became a global product early on. Wool and cotton were cultivated in one place, transported to centres where yarn and cloth were produced, and these intermediate products were then moved to places all over the world, where they were sold for further refinement or usage. Both within and between countries, different technologies, including un-mechanized and mechanized production, often coexisted for remarkably long periods. Trade relations and political hegemony greatly influenced decisions about what processes could be performed most profitably in which place. Metropolitan authorities often prevented their colonies from protecting their local textile industries. With the process of de-colonization, however, the lower wage rates in the ex-colonies served to compete with the former colonial powers. Consequently, textile industries in the economic core regions of the world dramatically declined after 1950. Even if the direction of the flows remained in Western hands, this meant at least a partial reversal of roles. Nonetheless, long commodity chains in textile goods, linking all parts of the world, remained typical.

 See the literature mentioned in note 1.  See Chapter 22, ‘Global trade and textile workers’ by Prasannan Parthasarathi. 

Textile workers around the world, 1650–2000 To cover the whole experience of globalization as far as possible was another reason to opt for a long time frame. There is agreement that at the turn of the twenty-first century, the world experienced a period in which globalization, which had been occurring for some time, dramatically increased in pace and impact.10 Given this rapid increase, contemporary observers of globalization all too hastily jump to the conclusion that it is both a recent and a one-way process. Among historians, there is also general agreement that this phase was preceded by a phase of de-globalization lasting from 1914 to 1945, which included the war economy of the First World War, slow post-war reconstruction and war repayments, the protectionist measures induced by the crisis of the 1930s and finally another war economy. The period 1914–1945 was therefore one of closing borders, growing trade barriers and increasing limitations on international migration, as the national overviews show. Prior to that, the period 1820–1913 saw an earlier globalization trend that was in many ways quite as fundamental as that of the second half of the twentieth century. A transport revolution drove down transport costs from the 1820s onwards: railways, steamships, better roads and more canals made it possible to send bulk goods like raw cotton or coal over far greater distances and to places which had not been economically accessible before. Tariff barriers came down. Colonial powers forced open Third World territories and compelled their colonial populations to produce for world markets and consume the goods produced in the metropolis. Workers migrated in enormous numbers and capital was invested worldwide at a level not exceeded by the end of the twentieth century.11 Economic historians, who measure globalization by price convergence, argue that globalization only started in the 1820s and that the economic consequences only began to be felt in the Third World some 50 years later.12 But in other ways earlier waves of globalization are clearly visible.13 There is a particularly good 10 Pietra Rivoli, The Travels of a T-Shirt in the Global Economy: An Economist Examines the Markets, Power, and Politics of World Trade (Hoboken, NJ, 2005); Fran Abrams and James Astill, ‘Story of the Blues’, The Guardian (European edition), 29 May 2001. We are indebted to Marcel van der Linden for this reference. 11 Jeffrey G. Williamson, ‘Globalization, Convergence, and History’, Journal of Economic History 56 (2) (1996): 277–306; Kevin H. O’Rourke and Jeffrey G. Williamson, Globalization and History: The Evolution of a Nineteenth-Century Atlantic Economy (Cambridge, Mass. and London, 1999). 12 Kevin H. O’Rourke and Jeffrey G. Williamson, ‘When did Globalisation Begin?’, European Review of Economic History 6 (2002): 23–50; Jeffrey G. Williamson, ‘Land, Labor, and Globalization in the Third World, 1870–1940’, Journal of Economic History 62 (2002): 25–54. 13 Göran Therborn, ‘Globalizations: Dimensions, Historical Waves, Regional Effects, Normative Governance’, International Sociology 15 (2000): 151–179; Ronnie Robertson, The Three Waves of Globalisation: A History of a Developing Global Consciousness (London, New York and Nova Scotia, 2003); Isabelle Lescent-Giles, ‘Globalisation as a Long-term Process: A Case Study of the British Experience since 1700’, Entreprises et Histoire 32 (2003): 15–31; Dennis O. Flynn and Arturo Giráldez, ‘Path Dependence, Time Lags and 

The Ashgate Companion to the History of Textile Workers reason for historians of textile labour to start their research well before the 1820s; long distance trade in woollen and cotton fabrics pre-dated this period by many centuries. However, textile labour also became truly global much earlier. When the Dutch and English East Asian Companies reached the Indian Ocean in the early 1600s, they entered a pre-existing long-distance trading system. They soon discovered that they needed Indian cotton fabrics to purchase spices and pepper in the Indonesian archipelago.14 By the 1680s Indian cotton had also become fashionable in the Netherlands and Britain, and by the turn of the eighteenth century the British had developed a re-export trade in Indian cotton and imported large quantities of white cloth to be printed and finished in England.15 This led to the demand for cotton goods in Europe, which fuelled the Industrial Revolution and global competition in the textile markets for cotton goods. Global developments thus provide good arguments for starting a global history of textile labour in the seventeenth century, but there is more. Many of the developments that we associate with globalization, like long-distance trade in raw materials, yarn or cloth, long-distance influences in textile fashions and the displacement of production to low-wage areas are visible in the early modern period,16 and some of them long before this. It would be short-sighted to keep them out of a consideration of global developments in textile labour.

Choices and boundaries 3: spatial units The choice of spatial unit has two aspects: the type of unit and the units selected. For the type of unit, a good case could have been made for the specific textile regions, for the ‘Lancashires’ of the world. These were usually smaller than a nation-state, but some of them crossed national borders. In such cases some aspects would have been influenced by national differences (typically those in relation to national legislation and national economic policies), and others would not have

the Birth of Globalisation: A Critique of O’Rourke and Williamson’, European Review of Economic History 8 (2004): 81–108; Ulrich Pfister, ‘Die Entstehung der europäischen Weltwirtschaft (ca. 1450–1850): ein endogenes Modell’, Jahrbuch für Wirtschaftsgeschichte 2 (2003): 57–81. 14 K.N. Chaudhuri, Trade and Civilisation in the Indian Ocean: An Economic History from the Rise of Islam to 1750 (Cambridge, 1985), p. 83; Anthony Reid, Southeast Asia in the Age of Commerce 1450–1680, vol. I: The Lands below the Winds (New Haven and London, 1988), pp. 90–91. 15 John E. Wills Jr., ‘European Consumption and Asian Production in the Seventeenth and Eighteenth Centuries’, in John Brewer and Roy Porter (eds), Consumption and the World of Goods (London and New York, 1993), pp. 133–147, at pp. 136–137. 16 We are aware that this is a period based on Western history. For a – limited – global alternative, see Jerry Bentley, ‘Cross-cultural Interaction and Periodization in World History’, American Historical Review 101 (1996): 749–770. 

Textile workers around the world, 1650–2000 been.17 In the end we decided against comparing textile regions for two reasons. The more substantive reason was that we were afraid that if we were to opt for the textile regions, we would miss the change from earlier, smaller and more dispersed production centres to these textile regions. The practical reason was that we expected to be able to find specialists in national textile history, but were afraid that we would be far less able to identify specialists for the diverse textile regions, especially in non-western countries. We are aware of a number of problems in choosing countries or nation-states as units of analysis. Not all nation-states existed in all periods in the same form, and borders have shifted over time. As Nancy Green has pointed out, historians have often taken for granted the nation-state as a category of analysis because comparative history is rooted in nineteenth-century historiography. This leads to the danger of considering nations as virtually unchanging entities when it comes to national comparisons.18 Then there is also the debate on the diminishing importance of nation-states in light of globalization.19 If we nevertheless opted for nation-states as our unit of description in the overviews, it was mainly for the reason mentioned earlier: because it is the most practical and in historical and historiographical practice the most common solution. It is easier to find experts, because our colleagues tend to work on a national level with nationally oriented sources. It is therefore the state of the discipline that we felt limited in our choice here. But as Nancy Green states: Comparative history offers a means of reconciling those who wish to transcend the nation-state and those who insist upon its importance. Several points are crucial. First nation-states are very much a part of the nineteenth- and twentieth-century past … However, comparing historical nation states can help relativize their rigidities and aid us in understanding their constructed nature.20 To complicate matters of comparison, we ended up, not only with nation-states, but also with empires (the Habsburg Monarchy and the Ottoman Empire) and some of their successor states as well. Moreover, given the nature of the project there was room to focus on particular regions within a nation-state and to cross national borders.

17 R. Schüren, Staat und ländliche Industrialisierung: Sozialer Wandel in zwei Dörfern einer deutsch-niederländischen Textilgewerberegion 1830–1914 (Dortmund, 1985) offers an analysis of this problem for the textile region on the Dutch–German border. 18 Nancy Green, ‘Forms of Comparisons’, in Deborah Cohen and Maura O’Connor (eds), Comparison and History, Europe in Cross-National Perspective (New York, 2004), pp. 41–56, at p. 44. 19 R.J. Barry Jones, The World Turned Upside Down? Globalization and the Future of the State (Manchester, 2000). 20 Nancy Green, ‘Forms of Comparison’, p. 46. 

The Ashgate Companion to the History of Textile Workers This brought us to the question of which countries to select for the national overviews. Peter Baldwin quite rightly warns us that ‘The very choice of the nations, or whatever the unit of analysis may be, to be compared often determines the answers drawn.’21 The selection of countries was based on the idea that the major producers should be represented. Estimates for cotton and wool production for early periods are very scarce. The best option proved to be an overview produced by the International Labour Organization, which reflected the situation in the first half of the 1930s.22 This was clearly uneven in the sense that much of the small-scale production in the Third World will have escaped the statisticians composing the figures. However, we still adopted it for want of a better alternative. Based on this overview, we concluded that it would be manageable to include all countries that were responsible for 1.5 per cent or more of the world’s production of cotton or wool fabrics in the mid-1930s. A summary of the data derived from this source is offered in Table 1.1. The source used five indicators to measure cotton production (the consumption of raw cotton, number of spindles, number of looms, production of cotton yarn and production of cotton pieces), four to measure wool production (the consumption of raw wool, number of combing machines, number of spindles and number of looms) and the number of workers in the textile industry as a whole (which might, however, include other fabrics such as silk or linen). If a country scored more than 1.5 per cent of the world total in any of these indicators, it was added to the wish list. In this way, a list was drawn up comprising Argentina, Brazil, the United States, China, India, Japan, Australia, Belgium, Czechoslovakia, France, Germany, Italy, the Netherlands, Poland, Spain, Russia and the United Kingdom. To this list we added Mexico, Indonesia, Turkey and Austria, countries which either had huge cotton or wool production that was missed by the ILO statisticians or were prominent before or after the 1930s. For these 21 countries collaborators were sought, but we failed to find one for France and Indonesia, and the contribution of those we found for Australia and Belgium in the end did not lead to a contribution in this book.23 Experts on Egypt, Uruguay and Denmark responded to our call for papers, and national overviews on these countries are included in the book, resulting in the addition of very interesting cases to the project. How well do we cover the global experience of cotton and wool workers in this way? Of course one can raise doubts about whether the diversity of Chinese, Indian, British or US textile workers’ history over 350 years can be ‘covered’ in an essay of 10,000 words. But that is a limitation innate to overview projects of this kind. For the first half of the 1930s, the period for which we have ILO data, how 21 Peter Baldwin, ‘Comparing and Generalizing’, in Cohen and O’Connor (eds), Comparison and History, pp. 1–22, at p. 12. 22 L’Industrie textile dans le monde. Problèmes économiques et sociaux (BIT: Etudes et Documents B 27) (2 vols, Geneva, 1937). 23 However, the Belgian expert participated in the project, and one of the authors of a comparative chapter is a specialist on Belgian textile history, so the Belgian case did play a role in our discussions. 

Textile workers around the world, 1650–2000 well did we do? The countries involved in the project covered 87 per cent of the unweighted average of all indicators in some way, the national overviews included in the book cover 85 per cent.24 However, we have a somewhat better coverage for cotton (89 per cent) than for wool (75 per cent). That is a figure that can still be improved upon, as – no doubt – it is possible to improve upon the quality of the overviews and comparisons presented here. But as far as we know, this is the first project that has even posed the simple question of what the global experiences of textile workers were in the long run, has considered which national histories to look into to answer that question and has then proceeded to try and do so. The question may be simple and the procedure straightforward, but, as Deborah Cohen has remarked: ‘Comparative history is a tremendously uncertain business. There are many perils that await the unsuspecting.’25 In trying to overcome these perils, we have been fortunate to have the support of a great team of specialists. Apart from the authors of this book, they have included Danielle van den Heuvel, Marcel van der Linden, Jelle van Lottum, Jan Lucassen and Bart de Wilde, who contributed to the quality of the deliberations at the conference. The conference was made possible by the financial support of the Research Network ‘Labour 1500–2000’ of the Fund for Scientific Research Flanders (FWO), the Foundation Unger-Van Brerofund, the Royal Netherlands Academy of Arts and Sciences (KNAW) and the International Institute of Social History (IISG/IISH), to whom we are very grateful. Discussants at the Working-Class History seminar at Pittsburgh University (30 September 2004), the First European Congress of World and Global History (Leipzig University, September 2005), the 30th Social Science History Association Conference (Portland, November 2005) and the Sixth European Social Science History Conference (Amsterdam, March 2006) helped us sharpen our analysis. We also would like to thank Peter Blewett and Anne Lee for assisting with language correcting, Annelieke Vries-Baaijens for drawing the maps, Lindsey Brake for copy editing, Sally Parker for compiling the index and Sarah Charters and her colleagues at Ashgate for producing the book.

24 If all of the countries on our wish list could have been included, this figure would have risen to 94 per cent. 25 Deborah Cohen, ‘Comparative History, Buyer Beware’, in Cohen and O’Connor (eds), Comparison and History, pp. 57–70, at p. 60. Cohen discussed comparisons made by one author, not by a team, but a number of her observations apply nevertheless. 

The Ashgate Companion to the History of Textile Workers

APPENDIX 1 A global history of textile production, 1650–2000 Framework document Production

What was the size of production and how many people were involved in production? What changes occurred in the size of production and workforce? How can they be explained? What markets were exploited, and what role did regional and interregional competition play? How would you assess the impact of demand (consumption, trade cycle) on the production of textiles? What technological developments took place in textile production (spinning and weaving)? How did this influence production and labour relations?

Organization of production

How was textile production organized? What was the position of the masters/ management on the one hand and workers on the other? What changes occurred in the organization of production, and how can they be explained? What role did guilds, trade unions, employers’ organizations and other forms of labour organization play?

The workers

How were the textile workers recruited? What was their social and ethnic background? What changes took place and how can they be explained? What was the gender composition of the workforce? What changes took place and how can they be explained? Was there a specific age composition of the workforce? How is this (absence of) age differentiation to be explained? What were the labour conditions of the workers like (hours of work, payment, secondary labour conditions and so on)? What were the living circumstances of the workers? Did a specific work culture or occupational identity develop? What changes occurred in occupational identity and how are they to be explained? Were there forms of labour protest in this period? If so, how were they organized and which workers took part?

10

Textile workers around the world, 1650–2000 Environment

How important was the textile industry in the context of the national (or regional) economy? Did this change and if so, how and why? What policy and regulations did (local) governments impose on the textile industry? What was the relationship between urban and rural areas and how did it affect textile production?

11

Table 1.1

Cotton and wool production figures for countries with more than a 1.5 per cent share in world production in the mid-1930s, and other countries involved in the project Cotton Consumption of raw cotton 1933/34

Number of spindles 31.1.1936

Number of looms 31.12.33

Wool Production of cotton yarn 1935

Production of cotton pieces 1935

Consump -tion of raw wool 1934

Africa

Egypt

0.1

0.1

0.1

0.1

America

Argentine

0.1

0.1

0.1

0.1

1.8

Combing machines 1933–35

Persons

Number Number Workers of of looms in textile spindles 1933–35 industry 1933–35 worldwide(a) 1920–1933 (%) 0.5

0.5

Brazil

1.8

1.8

2.6

1.8

2.5

0.9

0.6

Mexico

0.7

0.6

1.1

0.6

1.1

0.1

0.6

0.0

Uruguay

25.4

23.3

China

10.5

3.2

1.4

8.0

2.5

1.2

India

8.8

6.3

6.1

9.5

9.4

1.5

Japan

12.8

6.9

8.9

13.1

13.3

5.5

Turkey

0.1

0.0

0.1

0.0

Australia

Australia

0.0

0.0

0.0

0.0

2.0

Europe

Austria

0.5

0.5

0.4

0.7

0.4

0.4

Belgium Czechoslovakia

1.1

1.3

1.8

1.3

1.4

2.0

4.0

3.1

1.6

1.8

1.3

2.4

3.3

2.4

2.4

1.0

1.2

5.0

4.9

2.6

0.1

0.1

0.1

0.1

0.2

Denmark

19.6

21.8

US Asia

19.0

0.0 16.9

16.3

15.3

12.2

8.7

0.3

0.4

27.5

7.0

3.1

5.9

10.6

1.7

1.2

0.8

4.3

0.6

France

4.9

6.5

6.3

4.4

3.9

11.8

17.2

11.4

11.0

6.6

Germany

6.6

6.6

7.1

6.4

6.3

10.6

21.2

14.7

19.2

8.0

Italy Netherlands

3.7

3.6

4.7

3.4

3.1

5.5

5.2

4.6

4.6

5.2

0.7

0.8

1.8

0.9

1.4

0.3

Poland

1.3

1.1

1.2

1.1

1.2

1.3

2.8

Spain Russia/ USSR

1.8

1.4

2.1

1.4

2.1

2.3

2.2

5.7

6.4

8.0

6.4

8.3

5.9

1.2

1.7

2.6

6.4

10.7

27.6

18.8

11.3

9.7

18.3

17.2

26.1

22.2

10.3

In table, of data available

96.1

96.3

95.5

94.9

94.5

89.5

97.2

90.7

89.4

97.0

95.3

Included in project

91.2

89.7

89.1

90.5

90.6

75.8

78.3

78.1

77.6

90.4

86.8

Included in book

90.1

88.4

87.4

89.2

89.2

73.8

74.3

75.0

76.0

88.6

84.9

UK

0.6 3.6

3.5

1.1 1.5

average

Total

Notes a

Large differences in figures given and years (for instance: China only in factories, India all).

When figures for the date or year mentioned are not available, figures for the nearest available date in the source have been used. Countries included in the project have country names in bold. Countries included in the project with a less than 1.5 per cent share have country names in italics. Figures in italics in country rows are not given in the source and have been estimated from the figures in other columns (yarn on the basis of spindles etc.). Data on silk, rayon, linen and jute are not included in production figures, but workers in these types of textiles are included in the figures on workers. Source: L’Industrie textile dans le monde. Problèmes économiques et sociaux (BIT: Etudes et Documents B 27) (2 vols, Geneva, 1937).

This page has been left blank intentionally

PaRt I

NatIONaL HIStORIES OF tEXtILE wORKERS

This page has been left blank intentionally

2 Textile production in Argentina, 1650–2000 Mirta Zaida Lobato

In Argentina a textile industry only developed in the modern sense in the twentieth century, mainly in the interwar period (1914–1939), when an important expansion in the sector took place. In contrast to other Latin American countries like Mexico and Brazil, the textile production from Rio de la Plata had not achieved consistent growth in the previous centuries, although different fabrics were produced in the colonial period. The main reason for this late start was the peculiar integration of Argentina in the international market as primary product supplier. In the period between 1880 and 1930 most of the manufacturing industry in Argentina was related to the growth of exports (meat-packing depots, flour mills, quebracho (tannin and hard wood) processing plants) and the construction and maintenance of railways, housing and public services. A national textile industry did not develop, due to the fact that most of the demand for textiles was met by imports. Since the First World War, however, the textile industry grew. This was a consequence of the natural protectionism created by the difficulties suffered by international trade. Besides, in Argentina, there was a growth in the supply of raw materials like cotton, companies producing synthetic fibres were created and the internal market grew. The expansion showed itself in an increase in the number of companies, workers and production. Changing economic conditions caused a decline in the textile industry from the 1970s until it almost disappeared. In 2001 an economic, political and social crisis led to a resurgence in textile manufacture. The starting year of the period covered in this book is based on developments in the European industry. Textile production in Argentina can be divided into three clearly defined periods: indigenous textile production, the colonial period and the modern textile industry. This last phase can again be divided into three stages. The first one took place from 1860 to 1914 and was characterized by a weak and slow expansion. The second stage started during the First World War and lasted until approximately 1950. It was a phase of fast expansion with important developments in the manufacturing sector. The third stage began in the 1950s when growth

Map 2.1

Argentina

Textile production in Argentina, 1650–2000 decreased. From the 1970s the industry shrank and by the end of the millennium it had lost most of its former importance. In this chapter, the periods mentioned above will be analysed, taking into account production characteristics and the organization of production, markets and the workforce. The analysis cannot be equally balanced for each period as there is not enough relevant historical literature on the earlier stages.

The indigenous textile industry Indigenous culture is of importance both to the material culture of Argentineans and to the national history of Argentina. The indigenous population used the fibres and colouring agents naturally available in the region. The sedentary population was located in some areas of the present territory, particularly in the north-west, while hunters and gatherers were scattered over the rest of the territory. Different communities lived in Tucumán, Catamarca, Salta, Jujuy and part of Santiago del Estero. Among their other skills, there were accomplished metallurgists, weavers and potters. The goods produced not only had economic and technological functions in daily life, but were also used for ceremonial purposes. The archaeological remains and references by conquerors and settlers inform us about the development of textile production over several thousands of years. Shirts or tunics, ponchos (capes), blankets, caps and corsets were woven on looms, mainly using the warp technique, from spun camel wool. In order to achieve attractive designs interwoven warps in red or blue were used. From sixteenthcentury evidence, we know that clothing was similar among the indigenous Diaguitas, Humahuaca and Chichas, but the styles were different. In the coastal areas, in the hunting, gatherer and fishing communities and among the MatacoMataguayos groups, they wove a fibre from caraguatá plants. This still exits in the well known ‘yicas’, a type of handbag. In the Cuyo area, a region dominated by the Huarpes aborigines, the pieces of clothing found suggest that part of this group had a developed textile technology that included wool and probably vegetable fibres, loom-weaving and leather and fur work. The arrival of the Spaniards in the area in the middle of the sixteenth century and the conquest and colonization of the territory influenced the local cultures deeply. This influence also changed the technology used in textiles. European ways of producing textiles prevailed and indigenous uses and techniques almost disappeared. The area of Araucania (present day Patagonia and Southern Chile)  Miriam Tarrago (ed.), Los pueblos originarios y la conquista, Nueva Historia Argentina vol. 1 (Buenos Aires, 2000).  Catalina Teresa Michieli, Aportes para la caracterización y la historia de la textilería en Cuyo (San Juan, Argentina, Universidad Nacional de San Juan, Facultad de Humanidades y Artes, Instituto de Investigaciones Antropológicas y Museo, 1984), pp. 18–19 and 24–25. 19

The Ashgate Companion to the History of Textile Workers was less affected by these changes. The textiles produced were always utilitarian: for clothing, protection or transport. Ponchos, shawls and belts were made to wear, blankets for bed coverings and saddlebags or travelling bags for riding outfits. Women did all the weaving. With the arrival of the Spaniards in what is now Argentina, the indigenous society was torn apart, although methods and styles of native production were not completely eliminated, but have been maintained as residual elements of a culture that still survives. The traditional communities and people of mixed racial background still use indigenous manufacturing techniques and visual patterns in a now capitalist society. Keeping up craft production (textiles, pottery and so on) combats rural unemployment and reconciles the backwardness and poverty with the natural beauty of the country in the development of tourism as a transnational industry. So textiles are still made domestically, and the goods produced are consumed as well as exchanged through an extended trading network that joins the local economy with the national and international capitalist system.

The colonial textile industry In the period of the sixteenth and the seventeenth centuries, the Spaniards conquered present-day Argentina. The Spanish reorganized the spatial divisions of the country and the political, economic and cultural relations. New urban centres were founded and smaller trading circuits joined together. In the northern region (Cuyo) and the central area (Córdoba) a colonial industry sprang up to satisfy the needs of the local market but little is known about its labour relations or organization of production. We have more information regarding trading circuits and thus it is possible to point out that trade, including textiles, linked cities that were located far from one another. From Buenos Aires to Alto Perú (present-day Bolivia), through the current provinces of Salta, Santiago del Estero and Córdoba, these trading circuits developed in the last third of the sixteenth century. In the seventeenth century greater specialization of production was introduced to satisfy the demand of the market in the Alto-Peruvian mining areas. The weaving and spinning took place in the area of Córdoba, where a brief expansion phase occurred between the end of the sixteenth century and the beginning of the following century. In the mid-1580s textile manufacturing started to develop with the production of weaving in the countryside (where native labour was available), the expansion of flocks of sheep and the increase in demand from the  Néstor García Canclini, Las culturas populares en el capitalismo (México, 1982), pp. 133–162.  Pedro Santos Martínez, Las industrias durante el virreinato (1776–1810) (Buenos Aires, 1969), pp. 38–50 and Vilma Miletich ‘El Río de la Plata en la economía colonial’, in Enrique Tandeter (ed.), La sociedad colonial, Nueva Historia Argentina, vol. II (Buenos Aires, 2000), pp. 145–188. 20

Textile production in Argentina, 1650–2000 Alto-Peruvian mining market. Textiles from Córdoba were traded in Potosí and other cities like Asunción and Santa Fe where they were exchanged for wine and sugar. Textile manufacturing was led by Europeans, but the actual work was performed by the indigenous inhabitants of Argentina under the encomienda system. This institution, that became of crucial importance for the development of the colonial economy, was imposed by the colonizers. Under it, the indigenous people were forced to work in personal service to the Spaniards. The Spanish encomendero did not own the indigenous encomiendas, but had a ‘concession’, generally for life, although in time the system transformed itself into a family concession that went on through several generations. The encomendero was supposed to protect the natives and convert them to Christianity, but generally these duties were not carried out. In textile production, the encomenderos directed the indigenous encomiendas in making coarse woollen clothes (sayales), linen, stockings, blankets and hats. The raw material, rough wool, came from the big flocks of sheep in the region, and cotton was obtained from the nearby regions of Santiago del Estero and Catamarca. During the last decade of the sixteenth century some Europeans established the first colonial textile obrajes (workshops). The obrajes were production units that combined the different phases of textile production (carding, spinning, weaving and dyeing) in one place. In these obrajes there was cloth, linen, hat and blanket production, mainly for consumers with limited purchasing power. Some of the encomenderos incorporated their obrajes into companies in which they cooperated with Spanish craftsmen. The former brought capital and native labour into the company while the latter provided their professional skills. The peak in obrajes production was reached in 1613, after which date the demand for Córdoba fabrics in the Alto-Peruvian market decreased due to competition from other regions such as Asunción (present-day Paraguay), Catamarca, La Rioja and Alto Perú itself. In Europe, by the mid-seventeenth century Spain and Portugal were in decline and the economic centre had moved to the north, to Holland, England and France. These changes also had an effect on the American territories, particularly in the form of an increasing demand for raw materials. Commerce moved to the Atlantic and the Spanish Crown decided to reorganize the territory and change the trading system. In 1776, the Crown created the Río de la Plata Viceroyalty, which extended from Alto Perú to Tierra del Fuego, with Buenos Aires as its capital city. New high courts (Audiencias) were established in Buenos Aires, Cuzco and Caracas. Between 1776 and approximately 1810, when the colonial order started to break down, textiles, together with mining and stock breeding, continued to be the most important industries in three areas: Alto Perú, Tucumán and Santiago del Estero. The textiles produced in Alto Perú, in Moxos, Chiquitos and Cochabamba provinces were in general rough and rudimentary, although the textiles from Moxos were considered to be of superior quality. In Cochabamba the production of textiles grew considerably, to the detriment of the industry in Moxos. The magnitude of this increase was reflected in the number of people involved in the production of cloth in Alto Perú (which had its centre in Cochabamba): the number grew from 2,000 in 1778 to 80,000 in 1799. In this province the main product was ‘tocuyo’ cloth, a combination of cotton and linen. Cochabamba cloth could be obtained in the 21

The Ashgate Companion to the History of Textile Workers whole of the viceroyalty, in Salta, Córdoba and Buenos Aires. The greatest problem for the industry was the supply of raw materials, particularly the cotton that was produced in Perú, Tucumán and other countries. The difficulties in obtaining fibres boosted cotton growing in Cochabamba itself. Although in Córdoba del Tucumán the development of the textile industry did not reach the same level as in Alto Perú, ponchos, blankets, bags, sack cloths and coarse woollen cloths were produced. In Salta del Tucumán the production was more important and the industrial centre was San Miguel de Tucumán where cotton or vicuña wool textiles were made on domestic looms. But the most important production area was Santiago del Estero, where ponchos and carpets were manufactured, mainly by women. The colonial textile industry faced some difficulties due to the importation of European goods at low prices, particularly from Britain. Between 1825 and 1850 British cotton textile prices dropped sharply as a result of the Industrial Revolution and this made the importation of cotton and woollen goods from Britain, and woollen and silk fabrics from France easier. On the other hand, the civil war that affected Argentina for more than half a century made the circulation of products difficult and increased prices. Internal and external conflicts had arisen, weakening Spanish colonial power. As a result important economic changes took place, but these did not affect textiles. In Argentina the textile industry did not have the dynamic role it had in other (particularly European) countries. The primary sector, based on cattle and the export of leather and dried beef, became stronger. Some attempts were made to set up factories for the production of cloth and paper. However the first attempts to set up textile factories were thwarted by the lack of infrastructure (the development of the railway network only took place in the second half of the nineteenth century) and the limited amounts of capital available, both within the region and from Europe. European capital was mainly oriented towards government loans, and banks charged exorbitant interest rates.

The modern textile industry In the second half of the nineteenth century, the transformation to capitalism in Argentina accelerated and the Argentinean economy became better integrated into the international market for primary goods. Wool exports to the Belgian and French markets increased sharply. However, increased wool production did not stimulate the development of a national textile industry or any other related activities, and although there was farmland suitable for cotton growing, its production only started to increase during the First World War. Wool was the main textile export until the 1880s, although the so-called wool ‘boom’ was affected by an international decrease in the demand for wool in the  Hilda Sabato, Capitalismo y ganadería en Buenos Aires: la fiebre del lanar 1850–1890 (Buenos Aires, 1989). 22

Textile production in Argentina, 1650–2000 1860s. The drop in prices, monetary problems and the economic crises of 1866 and 1873 created a suitable climate for the discussion of new economic policies. In fact, the first attempt to set up a wool textile industry followed the crisis of 1873. The necessity of protecting the industry and the diversification of production were briefly discussed. A company was even set up to found a factory, but due to a number of factors these efforts failed. International wool prices fell to less than during the 1866 crisis. When the economy had recovered, interest in stimulating a local wool industry had disappeared. The lack of customs protection and the fact that capital was more profitably invested in the primary sector did not help either. The crisis of 1890 did not have a stimulating effect on the sector either, as imports were not substituted by local products. This contrasted with what happened in other countries, like Brazil, where the crisis of 1890 favoured the national textile industry. The cotton industry was also hampered by the lack of cotton. Yields were low at the end of the nineteenth century, and the competition from British goods was very strong. For a long time, the main local textile companies survived only because of demand from the state. However, some experiments in cotton growing were performed, aimed at obtaining the necessary raw material for cotton mills. Between 1862 and 1864, during the cotton famine created by the American Civil War, some cotton plantations were set up by the Cotton Supply Association of Manchester to demonstrate the possibility of growing this crop. However, when US cotton production resumed its normal level the plantations were abandoned. At the beginning of the twentieth century cotton growing spread to the colonies of Chaco and Formosa with good results, but lack of credit and the distance between the production centre and manufacturers made the raw material expensive and the expected development of a textile industry did not occur. The first cotton mill was established in 1906, with a capital of one million pesos invested in machines and with a production capacity of 500,000 kilograms per year. However, the company could not compete with the price of imported goods. Only a 5 per cent import tax was charged, and this was not enough to make the protection effective. In 1910 there was only one cotton mill, which produced about 650 tonnes annually. The first wool-knitting factory was organized in Buenos Aires in 1873, but it was not successful. The factory had wool washing, carding, spinning and weaving sections and the plant had a steam machine that kept 19 looms going. The firm employed 60 workers, mainly women and children. Businessmen pressed   

José Panettieri, Síntesis histórica del desarrollo industrial argentino (Buenos Aires, 1969), pp. 20–28; José Carlos Chiaramonte, Nacionalismo y liberalismo económicos en la Argentina (Buenos Aires, 1971). Fernando Rocchi, ‘Consumir es un placer. La industria y la expansión de la demanda en Buenos Aires a la vuelta del siglo pasado’, Desarrollo Económico 148, vol. 37 (1998): 533–558. Alberto Cassagne Serres, La política comercial argentina relacionada con las industrias nacionales (Tesis presentada a la Facultad de Ciencias Económicas para optar al grado de doctor, Universidad de Buenos Aires, Facultad de Ciencias Económicas, 1916). 23

The Ashgate Companion to the History of Textile Workers Parliament to pass protectionist laws or to support the industry. In 1869 they introduced a proposal to subsidize wool washing and weaving, and in 1877 the Board of Directors of the cloth factory asked for a law that would prescribe that army uniforms had to be made with cloth produced by this factory. In 1878 they asked Parliament to remove the duty on the import of machines, and in 1886 they requested privileges to make wool and vegetal silk. Later on, in 1888 and 1892 they asked for further privileges to establish a weaving mill. In 1901, 1902, 1903 and 1922 different companies asked to be exempted from the customs duties on imported machines, and in 1914 they presented a project to encourage cotton growing and mills. All these initiatives show that businessmen envisaged an important role for the state in making the exploits of their companies viable. The national state promoted private economic activities on several occasions, but it was not enough to make the textile industry take off. In spite of all this, there were several factories in Buenos Aires at the beginning of the twentieth century, such as the Societé per l’esportazione e per l’industria Italo Americana that had been founded by the Italian businessman Dell’Acqua, the French Prat’s, Barolo and Company factories and Remigio Montero’s. As is clear from the names of the companies, but also from the publicity and the goods sold, these firms were involved not only in production, but also in commerce. This corroborates what we know about Argentinean businessmen from existing literature.10 These importers of industrial products started to substitute what they bought from abroad in the 1930s and 1940s, becoming ‘pure industrialists’. However, it is clear that in many cases they continued to import goods when it was convenient for them.11 Many businessmen, as well as workers, were immigrants, mainly from Italy and Spain, and had come to the country attracted by the opportunities of a promising economy. It is important to remember that European immigrants were an important component of population growth in Argentina during the last quarter of the nineteenth century. Immigrants from Italy were always the largest group, followed by the Spanish, French, Russians and Syrian-Lebanese. Workers demanded improvements in working conditions, particularly in the working day and salaries. They also organized unions in some factories in 1901 and 1905, and newspapers record partial conflicts, although these issues have not been investigated in depth yet. In the period 1890–1910, there are only a few strikes organized by textile workers, but we find striking stevedores, bakers, printers, railway drivers, and sailors. During the first decade of the twentieth century, the textile industry developed slowly if we compare it with other sectors of industry. There were 2,200 companies with a capital of only 121 million pesos, compared with, for example, 800 million  F.J. Scardin, La Argentina y el trabajo (Buenos Aires, 1906) pp. 520–528; Adolfo Dorfman, Historia de la industria argentina (Buenos Aires, 1970), pp. 79–81, 135–141, 373–380. 10 Fernando Rocchi, ‘En busca del empresariado perdido: los industriales argentinos y las tesis de Jorge Federico Sábato’, Entrepasados 10 (1996): 67–106 and Jorge Schvarzer, Empresarios del pasado. La Unión Industrial Argentina (Buenos Aires, 1991). 11 Rocchi, ‘En busca del empresariado perdido’. 24

Textile production in Argentina, 1650–2000 pesos invested in the food industry. The number of workers was smaller than in other industries and workers’ organization was still limited.

The expansion of the textile industry during the interwar period The First World War had a positive impact on the growth of some industrial activities. The war caused restrictions on the importation of yarn and woven fabrics so other sectors that were part of the textile industry started to grow and change. From that point on, the activity in the textile sector became more complex. In different sectors (cotton, silk and wool) specialized firms appeared in areas such as spinning, silk and knitted fabrics, ribbon and elastics. The sector focused on wool production, but in the 1920s, changes in government policy favoured the expansion of cotton. By that time, a pest had affected US production. This led to a decrease in the international supply of cotton, prices rose and this encouraged the expansion of cotton growing in the north-eastern part of the country. The Ministry of Agriculture fostered the expansion of crops through studies, experimental plants and distribution of free seeds among small producers, and by stimulating immigration. The production of cotton fibres was oriented towards exports and gave rise to only a limited process of industrialization. The development of the textile industry accelerated due to the economic crisis after 1929, when the barriers applied to international commerce led to a process of import substitution. According to Alberto Petrecolla, in the 1930s the Argentinean textile industry industrialized swiftly. The depression did not affect investments in industry, and the number of spindles grew, both through a growing number of companies and because of the increase in average company size.12 The first textile mills used natural fibres and had four kinds of specialization: 1) spinning, weaving and different articles made in wool, cotton and other fibres, including silk mixtures; 2) woven and knitted fabrics, cotton or mixtures; 3) textiles and silk articles; and 4) stockings. From the 1930s, artificial fibres, cellulose or synthetic were introduced. The coexistence of capital-intensive firms and family companies, small workshops and self-employed workers (cuentapropismo) was typical in the textile industry. The value of the textile industry doubled between the beginning of the twentieth century and the 1915–1919 period, and a new leap was made in 1935–1939 (Table 2.1). According to a report of the Chamber of Commerce and Industry, in 1923 there were 162 textile mills, with a total invested capital of US  $131 million and 16,600 workers, but national production only covered 26.5 per cent of domestic consumption. By 1932 the number of textile mills had risen to 289, with US  $191 million in capital and 35,850 workers.13 In 1939, the National Labour 12 Alberto C. Petrecolla, Prices, Import Substitution and Investment in the Argentine Textile Industry (1920–1939) (Buenos Aires, Instituto Torcuato Di Tella, November 1968), pp. 57 and 60. 13 Claudio F. Bellini, Industria y Política. Estados, empresarios e industrialización en los años de Perón, 1943–1955 Capítulo: La industria textil (Tesis doctoral, Facultad de Filosofía y 25

The Ashgate Companion to the History of Textile Workers Department (DNT, Departamento Nacional del Trabajo) reported that there had been 157 textile companies with 9,260 workers in 1914, and by 1935 the number had risen to 436 with 36,815 workers.14 Table 2.1 Value added of textiles and other products, Argentina 1900–1950 (%) Textiles

1900–1904 1905–1909 1910–1914 1915–1919 1920–1924 1925–1929 1930–1934 1935–1939 1940–1944 1945–1949

4.1 3.3 3.6 6.4 5.5 4.5 7.1 10.8 13.2 14.6

Food, beverages, tobacco 34.3 27.8 28.8 35.0 31.2 27.7 25.6 26.5 25.6 21.7

Clothing Wood Chemicals/ Leather Metal footwear products petroleum/ products products coal 11.4 14.6 11.3 6.4 8.7 7.7 6.3 4.7 5.2 5.2

14.9 12.8 10.9 13.5 11.1 10.2 10.3 7.6 6.6 7.1

6.7 7.2 7.5 9.6 8.2 9.4 11.5 10.0 11.8 10.6

5.1 4.9 4.8 6.9 5.7 5.2 5.2 3.2 4.3 3.8

0.7 0.6 1.1 1.4 3.0 4.8 5.0 10.8 10.3 13.4

Source: Alberto C. Petrecolla, Prices, Import, Substitution and Investment in the Argentine Textile Industry (1929–1939) (Buenos Aires, Instituto Torcuato Di Tella, November 1968), pp. 22–23.

Among textile companies, the change was particularly important in cotton spinning, as Argentinean products replaced imported goods. Originally, the production of yarn was small compared to woven products, but during the 1930s the spinning industry expanded. The expansion in cotton spinning took place through the foundation of new plants, the enlargement of existing ones and the introduction of new technology that increased productivity.15 In 1920 there were only three cotton spinning factories, by 1931 the number had doubled, and it reached 22 by 1939 (Table 2.2). In 1930 the number of spindles was 50,000, in 1933 it was 100,000 and in 1935, it had increased to 215,000. If we take into account the number of spindles, the Argentinean industry ranked 24th in the world in 1935. In wool spinning, on the other hand, the investment of capital remained low, possibly due to the industry not fulfilling its existing capacity. Both the wool and cotton

Letras, Universidad de Buenos Aires, 2003). 14 Industria Textil. Capacidad normal de trabajo de los obreros de la industria textil, especialmente mujeres y menores (Departamento Nacional de Trabajo, Buenos Aires, 12 de junio de 1939), pp. 5–6. Alejandro E. Bunge, Una nueva argentina (Buenos Aires, 1984), pp. 193–218. 15 Industria Textil. Capacidad normal de trabajo, pp. 5–6. 26

Textile production in Argentina, 1650–2000 sectors introduced newer technologies, combining more up-to-date machinery with an intensification of the work pace. Table 2.2 The cotton spinning industry in Argentina, 1920–1929 Mills Number of Cotton consumption Yarn produced spindlesa (tonnes) (tonnes) 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

3 3 4 4 5 5 5 5 5 5 5 6 6 7 10 18 19 22 23 22

14,100 14,100 18,300 26,500 34,600 40,600 43,000 43,000 43,000 50,000 52,000 60,000 80,000 100,000 140,000 215,000 262,000 309,034 328,906 331,950

645 938 1,291 1,292 2,156 2,624 3,282 3,488 4,111 4,052 4,068 5,023 6,959 8,177 11,542 15,897 20,792 15,713 24,372 29,016

802 1,232 1,779 1,755 2,524 3,078 3,774 4,069 4,766 4,751 4,818 5,944 8,152 9,647 13,444 18,945 24,397 30,589 28,853 34,568

Spindles per mill 4,700 4,700 4,575 6,625 6,920 8,120 8,600 8,600 8,600 10,000 10,400 10,000 13,333 14,286 14,000 13,440 13,817 14,000 14,300 15,088

Notes a

Spindles in the industry on 31 December of each year.

Sources: Junta nacional de Algodón, La industrialización de la fibra de algodón en la República Argentina, Annual Reports, mimeos. Alberto C. Petrecolla, Prices, Import, Substitution and Investment in the Argentine Textile Industry (1929–1939) (Buenos Aires, Instituto Torcuato Di Tella, November 1968), p. 59.

The knitting industry also expanded in the 1920s, possibly due to the customs reforms in 1923. When the international crisis broke out, the conditions in which the industry developed changed. The provisional government of Uriburu (1930–1932) introduced currency controls and raised customs rates by 10 per cent. The peso suffered a devaluation of about 40 per cent between June 1931 and the beginning of 1932. Although there was a reduction in the income generated by exports, national consumption was enough to encourage an increase of capital investment and textile production. The knitting sector grew because of the installation of small plants, 27

The Ashgate Companion to the History of Textile Workers which were less expensive and used simpler technology. In 1933, the knitting industry had 270 factories and 16,000 workers, and almost fully met the demand from the domestic market. Government customs policy did not try to support the development of the sector, although it took some ambiguous measures regarding silk textiles, an activity that went back to the 1920s. Uriburu’s government reduced the customs duties on silk imports to 50 per cent in 1930. The next year there was a 10 per cent rise in duties to increase tax income. In 1933 the Roca–Runciman Agreement between Britain and Argentina split the consignment between natural silk textiles and those mixed with artificial silk, giving discounts to the latter. In addition to this, there was an internal tax that returned the level of duties to that of 1930. A 1937 law cancelled the extra 10 per cent introduced in 1931 for silk textiles.16 During this period new foreign companies were established to produce locally the goods that had previously been imported. Between 1931 and 1943, about seven large textile factories were installed, including Ducilo S.A., which opened its plant in 1937. This firm belonged to Industrias Químicas Argentina Duperial (Duperial Chemical Industry), which had been founded by the American Company Du Pont and the British Imperial Chemical Industries in 1934, a fact that emphasizes the process of import substitution in the silk industry.17 Other American companies founded cotton spinning and weaving plants, like Sudamtex S.A. (1934) and Anderson Clayton (1935), which also undertook cotton cleaning and oil production. The expansion lasted for the whole of the 1930s, and included not only cloth production, but also cotton, wool and rayon spinning. It was only partially interrupted by imports from Japan and Italy, which generated a crisis in 1937 and 1938. In the post-war period, it was necessary to re-equip the textile industry and make decisions on economic policy regarding the textile market. Originally, during the First World War a textile machinery industry had developed to satisfy local demand. In 1938, Sedalana S.A. created an auxiliary section to repair machines. In August 1942 Talleres Coghlan opened with 600 workers and 40 technicians, and apart from repairing textile machines, it produced new machines for carding cotton, weaving, dyeing and cutting. The products of this company competed with European manufacturers who were trying to sell their obsolete machines to Argentina, a policy Talleres Coghlan rejected. By the end of 1945, the Economy and Industrial Policies Office of the Secretary of Industry favoured the reintroduction of permits to acquire machinery in order to avoid the import of obsolete capital goods, a measure supported by the Confederación Argentina de Industrias Textiles (CAIT, Argentinean Textile Industry Confederation).18 16 Bellini, Industria y Política. 17 Oscar Colman, ‘La industria textil y la reconversión extensiva del sector industrial argentino’, Ciclos 2 (1992): 149–150; Eduardo Jorge, Industria y concentración económica. Desde principios de siglo hasta el peronismo (Buenos Aires, 1973); and Ducilo, 1937–1967. Crónica de una industria para industrias (Buenos Aires, 1967). 18 Claudio F. Bellini, Las políticas industriales del peronismo, Capítulo1, La industria textil 28

Textile production in Argentina, 1650–2000 As for state supervision of the domestic textile market, in 1939 Congress passed Law 12591 in order to limit price increases. The state imposed restrictions on textile exports as the government considered that internal demand was not being suitably met. In April 1943, the export of silk and rayon goods was prohibited, a year later permits were introduced for the export of cotton goods, and in June 1944 the export of cotton, wool and blended woven goods was banned. At the beginning of the following year, the exportation of weaving machinery was prohibited, and the requirement of permits for the exportation of fibre, cotton waste and woven goods was extended. All these measures aimed at maintaining the internal price level through the limitation of the newly developed export. The development of the industry was unequal. There were companies with large, modern plants, which introduced machinery and could adapt to changes more quickly, and others, which were smaller with higher expenses and depended on market protection. In the 1930s, the biggest factories were Alpargatas, Campomar, Ducilo, Grafa and Sudamtex, with between 2,000 and 7,000 workers each. For instance, Campomar had 2,500 workers, Grafa had about 3,000 workers and 200 white-collar employees and Alpargata employed about 7,000 people. There were also other smaller factories, for example the Patent Knitting Company, which had about 500 workers. This company was part of an association of smaller and medium-sized factories located in areas of Buenos Aires City and in its suburbs, such as San Martín.19 Many of these small and medium-sized companies grew under the government’s protection and a gradual increase in consumption, mainly in popular sectors that widened the opportunities for commerce and production. A recent research report mentions a rise of 79.7 per cent in textile production between 1939 and 1963 and this figure was higher than that for manufacturing in general (51.4 per cent).20 As for the work process, little is known. The basic process of the transformation of textile fibres (cotton) was relatively simple. First, the fibres went through a mechanical process of cleaning and then they were spun into continuous and regular threads of pre-determined diameters, which underwent stretching and twisting processes. Once the thread was obtained, it was woven and finally high-finish characteristics were given to the fabric, taking care of its appearance and texture.21 This process was performed in basic spaces, where machines and workers were distributed in such a way that all tasks were carried out in a continuous process. (Manuscrito, 2003). 19 Mirta Zaida Lobato, La vida en las fábricas. Trabajo protesta y política en una comunidad obrera (Tesis de doctorado, Universidad de Buenos Aires, 1999), p. 11 and Torcuato Di Tella, ‘La Unión Obrera Textil, 1930–1945’, in Torcuato S. Di Tella (ed.) Sindicatos como los de antes… (Buenos Aires, 1993), pp. 169–214. 20 Liliana Acero in collaboration with Claudia Minoliti, Alejandra Robania and Irma Nora Perez Vichech, Textile Workers in Brazil and Argentina: A Study of the Interrelationships between Work and Households (Hong Kong, 1991), pp. 165–177. 21 Enrique Pedraza, La industria textil argentina (Tesis presentada a la Facultad de Ciencias Económicas para optar al grado de doctor, Universidad de Buenos Aires, Facultad de Ciencias Económicas, 1946), pp. 510–520. 29

The Ashgate Companion to the History of Textile Workers As in any industrial activity, jobs were divided, but the conceptual system and a scientific work organization were not incorporated as much as in the meat-packing industry for instance. This does not mean that there was a lack of technicians or modern machinery, or that the workers´ tasks were not measured or supervised.22 The transformation of the textile industry was accompanied by an increase in the number of employees. Between 1930 and 1934 workers in the silk-weaving factories increased from 2,300 to 8,990, in wool-weaving and spinning mills from 7,150 to 11,000, in cotton-weaving mills from 5,850 to 9,770 and in the knitting industry from 14,000 to 17,000. Between 1929 and 1945 employment in the textile industry in Buenos Aires almost tripled (Table 2.3). Most of the workers were young women (the percentages range between 63 and 75 per cent) and even children, although their employment cannot be compared with the child labour in the earlier period of industrialization in other countries. The textile industry workers were not an elite. Most of them were unskilled workers and could easily be replaced by others, although they needed more training than other unqualified workers such as those in the meat-packing industry.23 Weaving was the most skilled sector. In 1936, 60 per cent of the workers (men and women) were Argentinean, about 11 per cent were Spanish and Italian, and between 2 and 4 per cent were Polish, German or Yugoslavian. But with the decrease in immigration, the workforce was nationalized and foreign workers vanished from the scene. Table 2.3 Employment in the Buenos Aires textile industry, 1929–1945 (1929 = 100) Employment 1929 1930 1931 1932 1933 1934 1935 1936 1937

Employment 1938 1939 1940 1941 1942 1943 1944 1945

100.0 100.8 105.2 117.5 132.7 151.6 181.2 209.7 224.5

211.5 218.9 241.0 254.1 276.3 298.2 289.2 290.4

Sources: DNT, Investigaciones sociales, 1940, p. 36; Dirección de Estadística Social, Investigaciones sociales, 1943–1945, p. 62.

22 Mirta Zaida Lobato, El ‘taylorismo’ en la gran industria exportadora argentina, 1907–45 (Buenos Aires, 1988). 23 Mirta Zaida Lobato, La vida en las fábricas. Trabajo, protesta y política en una comunidad obrera (Buenos Aires, 2001). 30

Textile production in Argentina, 1650–2000 Working conditions depended on the size of factories and their location. In the smaller ones the hygiene and security conditions were worse than in the big factories. In addition, the ones located in the provinces had a greater chance of escaping the control of weak state institutions like the DNT, which was in charge of inspecting labour conditions. Heat, humidity and cotton particles floating in the air were health hazards. As for trade unions, some had been organized in some factories in the previous period, but in 1921 the FOIT (Federación Obrera de la Industria Textil/Workers’ Federation of the Textile Industry) was formed. In contrast to the previous period, there are records of more labour conflicts, mainly in plants located in the city of Buenos Aires. A new feature in the 1930s with regard to workers’ organization was the existence of more than one trade union and the arguments between unionists, Socialists and Communists. In the 1920s, the UOT (Unión Obrera Textil/Textile Workers’ Union) was led by Syndicalists, and for a short period by Communists who lost that position in 1928. After the economic crisis and the military coup that took place in 1930 the FOIT almost disappeared and although the UOT was still functioning, it could not stop the dismissals or the salary reductions that followed the military coup. Qualified weavers led the defence of working conditions and salaries. The Socialists also led the UOT, which engaged in some conflicts, but as they were not as popular they did not attract the government’s attention in the same way as strikes by other groups of workers. Besides, the government was not as hostile towards Socialists as it was towards Communists. Communism was the ‘red threat’ for the conservative governments of the period, because, among other reasons, the Communists had already organized several industrial unions for construction and food. The textile workers organized in a union that was against the Socialists, but in 1936 they amalgamated and two years later the Communists dominated the UOT. In 1936 there were two trade unions in the textile sector with a total of 5,550 members and by 1941 this figure had risen to 12,504.24 In that same year, the UOT, led by Jorge Michellón, had about 10,000 members.25 The presence of the Communists in the UOT combined with the general growth in membership and the trade unions led strikes in the textile sector in the late 1930s. The relationship between the Socialists, Communists and Syndicalists was visibly difficult not only in the textile sector but also within the CGT (Confederación General del Trabajo/General Labour Confederation) that had been formed in September 1930. The demands were similar in the different companies: wage increases and the reinstatement of dismissed workers and/or union leaders. The most important change was the pressure from trade unions to sign a collective agreement in the cotton, wool and silk sectors. In April 1936, the UOT pressed for the signing of a collective contract, especially in the 24 Boletín informativo del Departamento Nacional del Trabajo, Primer Censo de Asociaciones Profesionales Obreras (Septiembre octubre de 1936), p. 47 and DES investigaciones Sociales (1943–1945), p. 29. 25 Hugo del Campo, Sindicalismo y peronismo. Los comienzos de un vínculo perdurable (Buenos Aires, 1983), pp. 94–95. 31

The Ashgate Companion to the History of Textile Workers wool sector. The clauses were related to salaries, freedom of association, abolition of favouritism, the rotation of workers instead of dismissals, the reinstatement of workers dismissed because of union activism and the creation of a joint committee between workers and employers to discuss problems. In addition, it required the supervision of the DNT and the Provincial Labour Department in the province of Buenos Aires.26 The textile workers’ movement was very active between 1935 and 1937 (Table 2.4) and its collective bargaining power in the wool sector helped the cotton, silk and stocking sector negotiations. But this success was not permanent and the political and economic conflicts produced by the Second World War affected the trade unions and deepened the gap between Communists, Socialists and Syndicalists. Table 2.4 Strikes and union meetings in the Buenos Aires textile industry, 1935–1945 Strikes 1935 1936 1937 1938 1939 1940

13 19 12 7 4 10

Union meetings

Strikes 1941 1942 1943 1944 1945

460 506 574 539 737 682

3 10 12 3 7

Union meetings 448 431 155 119 315

Sources: DNT, División de Estadística, Estadística de las huelgas (Buenos Aires 1940), p. 49; Estadística de las huelgas 1940 (Buenos Aires, 1941), p. 9; Investigaciones sociales 1940, p. 45; Investigaciones sociales 1941 (Buenos Aires, 1942), p. 88; Investigaciones sociales 1942 (Buenos Aires, 1943), pp. 88, 95; Dirección de Estadística Social, Investigaciones sociales, 1943–1945, pp. 17, 56–59.

In the 1930s, new Chambers of Commerce were created to represent the different sectors into which the industry was divided. In 1932 the Confederación Argentina de la Industria Textil (Argentinean Confederation of the Textile Industry) and the Cámara Industrial de la Seda (Industrial Chamber of Silk) were set up, and they joined the UIA (Unión Industrial Argentina/Argentinean Industrial Union). In 1934 the Asociación Textil Argentina (Argentinean Textile Association) was created, in 1936 the Asociación Fabricantes de Punto y Anexos (Association of Knitting Manufacturers) and in 1944 the Asociación Fabricantes de Tejidos de Rayón (Association of Rayon-Knitting Manufacturers) were founded.

26 Joel Horowitz, Argentine Unions, the State & the Rise of Perón, 1930–1945 (Institute of International Studies, University of California, Berkeley, 1990). 32

Textile production in Argentina, 1650–2000 By 1945 the textile industry had completed the development that had started as a result of the international crisis of 1929. The change in relative prices, the reduction in the Argentinean capacity to import and the introduction of currency controls had given the incipient textile industry, concentrated mainly in the manufacture of wool, a chance to begin an expansion process that continued through the war years. The expansion not only included the substitution of weaving imports but also wool, cotton and rayon-weaving. This expansion took place without modern machinery and through the engagement of a cheap workforce, which had increased in size as a result of the crisis. War made textile imports impossible and intensified the import substitution process. As conditions on the international markets changed, Argentina managed to export to South American and South African markets. Table 2.5

Textile industry: workers and production, Argentina, 1939–1963 Workers

Growth (%)

Production

Growth (%)

Manufactures 1939

619,721

56,970,000

1946

938,387

51.4

111,714,280

96.1

1950

923,824

–1.6

129,931,000

16.3

1963

908,753

−1.6

197,881,772

52.3

Textiles 1939

96,732

2,664,577

1946

117,110

21.1

13,133,923

392.9

1950

153,876

31.4

19,297,725

46.9

1963

106,029

−31.1

20,973,306

8.7

Source: Liliana Acero in collaboration with Claudia Minoliti, Alejandra Robania and Irma Nora Perez Vichech, Textile Workers in Brazil and Argentina: A Study of the Interrelationships between Work and Households (United Nations University Press, Hong Kong, 1991), p. 177.

33

The Ashgate Companion to the History of Textile Workers Table 2.6 Evolution of the textile industry in Argentina, 1914–1939 Mills

1914 1923 1932 1935 1937 1938 1939a a

2,317 250 380 672 908 921 753

Capital (US  $)

Value of production (US  $)

Raw materials used (US  $)

Personnel

Machine capacity (hp)

27,200,000 160,000,000 205,000,000 211,200,000 – – –

29,200,000 120,000,000 210,000,000 261,200,000 348,000,000 324,000,000 285,900,000

16,400,000 – – 181,500,000 233,000,000 213,000,000 163,100,000

12,500 19,000 37,000 49,500 61,000 54,000 63,000

8,600 – – 58,000 82,000 – –

Excluding cotton cleaning and wool washing.

Source: Adolfo Dorfman, Cincuenta años d eindustrialización en la Argentina. Desarrollo y Perspectivas (Buenos Aires, 1983), p. 376.

The textile industry under Perón’s government The textile industry continued to expand during the government of Juan Domingo Perón (1946–1955) but at a slower pace. Unlike other sectors, no specific policy was devoted to textiles during this period, although the Ley de Fomento Industrial (Stimulation of Industry Bill) supported wool and cotton production.27 The first five-year economic plan stated production objectives that the sector should reach by 1951 and in 1952, the second five-year plan placed the industry in last place, together with the food industry, in the priorities of industrial policy. On the other hand, the price control policy that was intended to set limits on the increase in the cost of living generated problems within the sector. As the government was worried about rising prices the cotton textiles industry had to adapt to the maximum prices policy. The government ordered the industry to set aside 10 per cent of its production to ensure the supply of the knitting industry. In addition, to avoid speculative activities in the trade chain, business deals between wholesalers and retailers were banned. Maximum prices for textile articles for domestic use and women’s clothing were set and later men’s and children’s clothes were included. The price control system involved manufacturers, dressmakers, wholesalers and retailers. The reaction of businessmen and traders was to sit on their textiles, causing shortages. All these measures increased the shortage of some textile goods such as cotton products. The policy of maintaining prices failed and in 1947 the government changed course. In June 1947 a new Bill confiscated stocks and stopped production, revoked the 27 Bellini, Industria y Política, pp. 18–60. 34

Textile production in Argentina, 1650–2000 agreement between manufacturers and wholesalers and froze prices at the August 1946 level. A mixed Committee for the Repression of Speculation was created. The government’s decisions caused anxiety among the factory owners who expressed their disagreement and started to take measures against the workers in the sector. For instance, some expansion plans were cancelled at important companies such as La Fábrica Argentina de Alpargatas, which stopped the construction of a fine canvas factory, blaming rising costs and conflicts with the unions. The government implemented policies in order to oppose speculation and those policies included measures against business interests. On 19 September 1947 an official report revealed that Sudamtex and Sedalana S.A. had made huge profits, by increasing prices artificially for retailers through business deals with big firms connected with them. Difficulties persisted during Peron’s two presidencies and the government had to change the maximum price system for another one that transferred the increases of production costs to prices. Trade was seriously affected by government policies as business deals between factories and wholesalers were practically banned and they had to work with their habitual customers. Between October 1947 and February 1949 the textile manufacturers were freed from the imposed prices because of the constant increase in production costs due to the application of the Régimen de Utilidades Máximas. But in February 1949 the economic policy changed when Alfredo Gómez Morales took over the Economic Ministry and pressures on the industry started to increase again. Bill 4995 reduced the percentages of maximum profits established and the transfer of salary rises to price increases was banned in order to stop inflation. Maximum prices and the maintenance of wage levels reduced profits at a time when sales started to stagnate as a consequence of inflation. In addition, businesses were damaged due to a lack of raw materials, parts and machinery because imports were slack due to a lack of purchasing power in the international market, caused by the drop of foreign currency reserves. Between 1946 and 1951 the textile industry had to deal with a policy of price control because its production influenced the cost of living. In the first stage, between June 1946 and June 1947 the bills that set maximum prices stayed in place, particularly for spinning, and at the same time, agreements with businessmen were made to set up production lines subject to maximum prices, and the policy of price control was emphasized. A month later, maximum prices were reset, and cost increases were taken into account. This system lasted until September 1949 when maximum prices were reintroduced. As happened during the first stage, this new freeze reduced company profits at a point when demand for textiles was low. Almost a year later, in August 1950, higher costs were again taken into account, eliminating the pressure on profits. In March 1952 maximum prices were reintroduced but by this time the industry was facing recession because of the decrease in domestic demand. Due to the slow recovery that followed, the failure of official policy was clearly seen in 1954 when the renovation of collective bargaining made the state accept price rises caused by new wage agreements. Price control is not the only matter to be considered in an analysis of the industry under Perón’s government. We also have to consider the customs policy. When the 35

The Ashgate Companion to the History of Textile Workers Second World War came to an end, businessmen were worried about international commerce and the British recovering business in the local markets. The official customs policy was based on the previous permit system and import quotas, and not on the change of customs rates as manufacturers suggested. In fact, there was no elaborate policy of import substitution in textiles, because the government gave preference to the supply of the domestic market. The importation of textiles included aspects that were not directly connected with the needs of the industry, for example bilateral trading treaties. From 1946 to 1948 treaties with Great Britain, Brazil, Spain and Switzerland were signed, which had an impact on the imports of spun yarns and textiles. Textile exports were hurt by the 1943 export legislation, when exports that had previously been permitted were banned until the end of the decade, with the objective to supply the domestic market with domestic products. In addition, industrial competition was also affected by the exchange revaluation that was steady between 1945 and 1955. Although from 1950 the government created exchange facilities to encourage exports, they did not use these because the facilities did not counteract money revaluation. At the same time, it was impossible to import modern machinery, and obsolete technology was a factor that made local textile competition worse, especially in the wool industry. As for labour policy during Peronism, it was aimed at giving social benefits to all workers, controlling labour conflicts and supporting unions in all industrial sectors.28 With regard to trade unions, 1945 was a key year for Perón’s relationship with the workers. The basis of his policy was to create unions parallel to those controlled by the Socialists and Communists, as happened with the UOT and in industries such as food, clothing, footwear and wood. The new trade unions grew very quickly, and as a result, there was a close relationship with Perón.29 However, during the years of Peronism there were conflicts and many strikes, despite the national authorities’ wish to maintain social peace. The reasons for these conflicts were numerous. In 1946 a salary rise was requested that resulted in a wool strike that included 30,000 workers, lasted three weeks and eventually ended favourably for the workers. Workers could also go on strike due to something as small as the dismissal of two AOT delegates, as happened in the Villa Flandria textile company, a strike that ended up including all of the workers in the province of Buenos Aires because of the intransigent stance of the company management. Once international trade had returned strikes were also feared in some sectors, such as stocking manufacture. Some strikes took place even after collective agreements had been signed, requesting a minimum working week of 36 hours. The ATA (Asociación Textil Argentina/Argentinean Textile Association) and the Asociación Fabricantes de Medias de la Confederación Argentina e Industrias Textiles (Association of Stocking Manufacturers of the Argentinean Confederation of the Textile Industry) asked Labour Secretary José María Freyre for state 28 Juan Carlos Torre et al. (eds), La formación del sindicalismo peronista (Buenos Aires, 1988) and Los años peronistas, Nueva Historia Argentina, vol. VIII (Buenos Aires, 2002). 29 Del Campo, Sindicalismo y peronismo, pp. 186–187. 36

Textile production in Argentina, 1650–2000 supervision, as the chance that manufacturers could not carry out the agreements of the Industry Office was high. Moreover, the president of the Mixed Committee for the Repression of Speculation was warned that the AOT wanted to extend the movement to all unions in the country. The firms requested that unemployment insurance would be paid for by the state, and they considered the working hours guarantee as in the exclusive sphere of state intervention. They also claimed that the strikes were illegal as they amounted to extortion of the state and industry owners. They held their positions firmly as they expected that the Labour Secretary would not be willing to negotiate. In addition to this, businessmen refused the union request of including management, security and administration personnel in the agreement, and changes to the night-work laws and the seniority scale. After 46 days of strikes the 36 hours’ guarantee was granted to the union and it was rejected by the ATA. In October 1947 another strike took place. This lasted eight days and included the cotton industry. Organized by the AOT, it involved more than 70,000 workers. The companies dismissed union delegates in over 250 factories and the government negotiator favoured businesses. In 1948, the CGT supervised the trade unions due to conflicts in the industry workers’ strikes. After the strikes in the cotton sector in October 1947 there were no conflicts in the industry. In 1950 the ATA and the Federación Argentina de la Industria Textil (Argentinean Federation of the Textile Industry), formerly the CAIT, claimed the authorities’ support against the requests for salary rises that were considered inadequate at a time when there was a decrease in sales and maximum prices were being established. For businessmen the problem was in the workers’ lack of discipline and they reported absences and a reluctance to work.

The decay of the textile industry The period 1930 to 1970 was the most favourable for the development of the textile industry. The sector grew rapidly, serious conflicts with the Peronist government notwithstanding. Between 1925 and 1950 the average growth rate was 10 per cent per year helped by the internal market, and, during the Peronist period, by the income distribution policy. Although the price control policy led to many problems, between 1950 and 1970 the annual average growth rate was 2.5 per cent per year. This slow growth can be seen in the decrease in the number of workers employed. In 1960 there were 26,530 people employed in cotton mills, in 1970 this figure had fallen to 16,487.30 Eighty-five per cent of production and employment came from companies with more than 10,000 spindles. Although in decay, the industry

30 Adolfo Canitrot, Julio Fidel, Milton Juillerat and Jorge Lucángeli, ‘El empleo en la industria textil argentina. Análisis de comportamiento y elección tecnológica’, Desarrollo Económico 63, vol. 16 (octubre-diciembre 1976): 349–364 and Acero et al., Textile Workers in Brazil and Argentina, pp. 170–177. 37

The Ashgate Companion to the History of Textile Workers was modern and mature compared to international competitors, as small family factories were not important. Overall, textile industry activity fell in the 1970s, with the exception of the period 1973–1975 when there was some stimulation as a result of the law of Industrial Encouragement (Tables 2.7 and 2.8). In the 1970s the companies producing cotton and wool textile fibres had to face a lot of problems due to fluctuations in demand and from 1983 the conflicts got worse.31 In 1980 there were 58 cotton mills with over 1,100,000 spindles.32 Most of the factories were located in the capital city and in the province of Buenos Aires. The industry comprised 45 cotton mills, 200 textile mills and 5,900 looms. The difficulties in the sector were linked to financial problems that made employers close some factories, as was the case for Linotex and Nihon Reori. Dismissals and the suspension of staff became widespread. Table 2.7 Physical volume of Argentinean textile production, 1971–1981 (1970 = 100) Production 1970 1971 1972 1973 1974 1975

Production 1976 1977 1978 1979 1980 1981

100.0 104.5 104.6 111.6 119.8 114.2

97.1 118.2 102.0 114.2 94.7 77.8

Source: Evolución de la industria manufacturera 1970–1981 (República Argentina, Ministerio de Economía, Instituto Nacional de Estadísticas y Censos), pp. 15–20.

31 Anuario 90, Tendencias económicas y financieras. La economía argentina. Consejo Técnico de Inversiones (Buenos Aires, 1991), pp. 318–319. 32 Jorge Katz and Bernardo Kosacoff, El proceso de industrialización en Argentina: Evolución, retroceso y prospectiva (Buenos Aires, 1989); Adolfo Dorfman, Cincuenta años de industrialización en la Argentina, 1930–1980. Desarrollo y perspectivas (Buenos Aires, 1983); and Anuario 80, Tendencias económicas y financieras. La economía argentina. Consejo Técnico de Inversiones (Buenos Aires, 1981), p. 332. 38

Textile production in Argentina, 1650–2000 Table 2.8 Indicators of activity in the textile industry, 1974–1989 (1970 = 100) Production 1974 1975 1976 1977 1978 1979 1980 1981

119.8 114.2 100.5 118.2 102.0 114.2 94.7 77.8

Workers

Production 1982 1983 1984 1985 1986 1987 1988 1989

105.6 109.2 105.3 96.9 84.7 75.8 60.3 45.6

79.2 92.7 98.0 75.2 93.1 87.7 83.6 84.3

Workers 48.4 52.7 58.6 52.1 50.3 49.4 44.1 47.7

Source: Anuario 90, Tendencias económicas, La economía Argentina, Instituto Nacional de Estadísiticas y Censos (INDEC), p. 320.

As for textiles made from cellulose and synthetic fibres, their manufacture grew at the expense of the rest of the textile industry. From 1980 there was an important decline in manufacturing and trading, with the industry operating at an average of 30–35 per cent less than full capacity.33 The causes of this were the increase in production costs, the duties that prevented the setting of competitive prices and the low purchasing power of the popular sector due to a fall in working hours, the cancellation of shifts and extra hours, suspensions and personnel dismissals. As a consequence of all these factors, several factories such as Sudamtex, AMAT, Alpesa and La Bernalesa were in crisis and others, for example Hirlon, Prenyl, Beccar, Reysol and Poliamidas Argentinas, closed. The textile factories were seriously damaged due to imported textiles and clothing and several factories ceased working. The decay of the industry was faster during the last military dictatorship (1976– 1983) and the economic policies taken by Minister José Alfredo Martínez de Hoz. The new military government intended to modify the dynamics and productive conditions produced by substitutive industrialization due to its supposed inefficiency on the international market, the discriminations produced in relation to agricultural production and in particular due to social conditions and the kind of political alliances created, as they were the basis for successive popular projects. The economic re-arrangement was based on a strict control of the state system and the discipline of workers. The new economic policy eliminated the redistribution of income towards popular sectors, dramatically reduced education and health expenses, and increased the military and security budget. Productive companies that belonged to the state were privatized and the public companies linked closely with oligopolistic firms. The economic programme of military management led to de-industrialization.34 The opening of the economy and the importation of 33 Anuario 80, Tendencias económicas y financieras. La economía argentina. Consejo Técnico de Inversiones (Buenos Aires, 1981), p. 335. 34 D. Azpiazu, M. Khavisse and E.M. Basualdo, El nuevo poder económico (Buenos Aires, 39

The Ashgate Companion to the History of Textile Workers clothes and textiles from eastern Asia, especially from China, affected the industry considerably. In the 1990s, neo-liberal policies opened the economy even wider and caused the industry to practically disappear. Factories and workshops were closed, producing massive de-industrialization that also affected other sectors of industry. Unemployment grew, seriously restricting the population’s consumption capacity. There was a general fall in the demand for industrial textiles products, which was accentuated by the illegal importation of articles of clothing that, according to textile businessmen, produced a US  $700 million tax evasion. There was also a considerable volume of smuggled goods sold in local markets as a result of illegal imports from countries with low production costs, mainly in Asia.35 Since the serious crisis that broke out in December 2001 a steady recovery of the sector has started. The devaluation of the peso caused a fall in imports and this allowed the industry to regain part of the domestic market. In addition, imports from Brazil and Asia stopped. During 2002, the import substitution of textiles reached 50 per cent and 70 per cent in garments. The activity in the sector increased fourfold between the last three months of 2001 and the same period in 2002. In 2002 and early 2003, according to press information and the Chambers of Commerce, around 100 small and medium-sized companies (with between 4 and 10 workers each) reopened their doors and many more began investing and opening new shops.36 As for labour, the working experience can be divided into two periods. From the military coup of 1955 to the one in 1976, workers achieved better salaries and got involved in the political conflicts of the different governments. At the beginning of this period there was a strong feeling of ‘revenge’ in all companies, which produced resistance in the factories. To face ‘de-Peronization’ policies, workers chose Peronist delegates in their unions. For instance, at Alpargatas, the most important textile company in the country, over 12,000 workers voted for ‘Peronist’ union lists. The confirmation of ‘Peronist’ power among factory workers was obvious whenever a salary rise was requested and when looking at labour agreements. This power produced debates about the clauses related to rationalization and work incentives and the power to be invested in the internal committees. Employers only accepted the proposal of a discussion about the renewal of collective bargaining when some trade unions started to become weak. The textile industry led the employers’ intransigence. The FITA (Federación de la Industria Textil Argentina/Federation of Argentinean Textile Industry) agreed to wage rises on the condition that unions accepted clauses on the rationalization of production. Textile workers went on strike often, sometimes with demoralizing results, as in 1959 and 1960, when the AOT demanded that employees work to rule 1988), pp. 151–191. 35 Anuario 2002, Tendencias económicas y financieras. La economía argentina. Consejo Técnico de Inversiones (Buenos Aires, 2003), p. 181. 36 Clarín, Suplemento de Economía y negocios, Buenos Aires, 11 de mayo de 2003, ‘La recuperación de la industria del tejido’. 40

Textile production in Argentina, 1650–2000 in the main plants of the country. The employers’ reply was immediate. Alpargatas and Sudamtex, companies that led the sector, dismissed a mass of employees including whole factory committees. Workers took control of the plants and the police cleared them out of the factories immediately. As the internal organization became weaker, companies imposed their rules as they wanted. Rationalization meant, at least in big companies, the realization of time-movement studies. Finally, in 1961, the national union agreed to sign an agreement declaring the applicability of the productivity plans.37 Regarding the organization of trade unionism, a centralized model was being consolidated. The leadership was then able to control any section that, according to the authorities, committed any undisciplined or irregular act. Moreover, the Professional Association Law guaranteed the negotiation rights of the union leaders regardless of the existence of rival trade unions. All unions, not only those in textiles, controlled important financial resources from the workers’ compulsory contributions and from assistance fees that were used for different services. Although it was a common feature of trade union organizations to be in charge of large sums of money, this made Peronist unions corrupt. An important but poor union like the AOT received millions of pesos under the heading of assistance fees and the employers’ contributions to social services. The AOT gave medical care and tourist facilities to its members, as it owned hotels and camping sites in Córdoba and Mar del Plata. However, the power of textile union leaders weakened when problems in the industry increased. Following the military coup of 1976 the union faced large difficulties due to two converging reasons. On the one hand, the repression of the military dictatorship extended to all sectors of work. They wanted to destroy the type of union organization established during the second half of the twentieth century, and to some extent, the trade union organizations had to resist the measures imposed by the military regime. The military government closed labour organizations and supervised the CGT and the most important unions like the metallurgic union workers (UOM), the mechanics union (SMATA), the telephone workers (FOETRA), the textile workers association (AOT) and the union of state oil workers (SUPE). The armed forces were implacable in chasing union leaders, including those who in the past had had fluent communication with military governments. Their policy was to achieve the subordination of the unions or make them illegal with the aid of state bodies like the Ministry of Work.38 On the other hand, the industry’s decay, factory shutdowns and workers’ dismissals during the 1980s and 1990s made trade unions weak and almost led to their disappearance.

37 Daniel James, Resistencia e integración. El peronismo y la clase trabajadora, 1946.1976 (Buenos Aires, 1990) and ‘Racionalización y respuesta de la clase obrera: contexto y limitaciones de la actividad gremial en la Argentina’, Desarrollo Económico 83, vol. 21 (1981). 38 Mirta Zaida Lobato and Juan Suriano, La protesta social en Argentina (Buenos Aires, 2003). 41

The Ashgate Companion to the History of Textile Workers

Conclusions Regarding historiography, we can conclude that the discipline of economic and social history in Argentina has not paid appropriate attention to the study of different industrial sectors, including the textile industry. The period chosen for this project (1650–2000) corresponds to the development of textiles in Europe. In the Argentinean case, the indigenous and colonial production had to be examined, but there are no studies about the organization of production. Observations on these subjects came from the literature on trade. The biggest development in the textile industry took place between the First and Second World Wars, due to the difficulties in importing textile goods. In spite of the importance of wool production and cotton growing, this did not lead to industrialization, but in many cases the exportation of raw materials was maintained. The development of synthetic polyamides, polyester and polypropylene fibres was the result of the provision of foreign capital. The growth of the sector did not come from a transformation of the colonial industry as happened in other Latin American countries. By 1945 the industry had completed its development cycle without importing modern machinery. The key to expansion was the incorporation of more workers. The most favourable period for the development of textile production was between 1930 and 1970 due to internal market control. From 1976 problems in the sector became more serious because of imports as well as smuggling. From the point of view of companies, the sector was heterogeneous, as the big companies that controlled the market coexisted with a larger number of mediumsized and small businesses. Companies did not integrate vertically as occurred in the meat-packing industry and only in some periods did they sign commercial agreements. As for the organization of work and control techniques, some aspects of Scientific Management were introduced from the 1930s. Companies were located mainly in the city and province of Buenos Aires, and, later on, with the industrial location plans of the 1970s, some were opened in other provinces. Industry was mainly urban and factories with small company towns in which workers were housed, which were typical of rural locations, were not common. Workers were often foreign at the start of the industry, but due to the decrease in immigration, by the middle of the twentieth century, they were mostly native Argentineans. The textile industry employed women. There was a clear difference between qualified and unskilled workers. Foremen and chiefs of sectors were the product of long careers. Working conditions depended on the size of the production unit but with the extension of social benefits that went with the consolidation of trade unions, improvements were clear. It is more difficult to give an accurate view of living conditions, as there are no available reports about real salaries and purchasing power. Trade unions were weak at the start of the industry, grew stronger in the 1930s and stronger still during Peronism. With the decay of the industry they again lost their importance.

42

3 Austria and Czechoslovakia: the Habsburg Monarchy and its successor states Andrea Komlosy

National histories of textiles tend to neglect the regional character of industrial production, which is always located in specific towns, industrial districts and regions. Studies of industrial relations with domestic as well as international locations incorporate these districts into a broader network. The size and character of these networks are not static but respond to economic cycles and social and political change. Another factor influencing the industrial landscape(s) is the extension of state territory and its borders. In the time under consideration, the Habsburg Monarchy underwent significant territorial changes. Basically it consisted of the constitutive Austrian territories, the Archduchies of Lower and Upper Austria (in the following referred to as ‘Austrian Lands’), Inner Austria in the south (Styria, Carinthia, Carniola, Istria), Tyrol and the scattered territories of the Vorlande (Further Austria) in the west, and Bohemia, Moravia and Silesia (in the following referred to as ‘Bohemian Lands’) in the north. Except for a narrow region in western and northern Hungary, the majority of the Hungarian Kingdom was under Ottoman rule from 1526 until 1699. Repelling the Ottoman army, the imperial troops expanded the Habsburg possessions towards the European south-east, conquering Hungary and Transylvania in 1699 as well as Banat and western Serbia in 1718. At the same time, Habsburg Austria made territorial acquisitions in western Europe in the wake of the War of Spanish Succession. While the Bourbons succeeded in conquering the Spanish Crown, the Habsburgs received the Spanish possessions in Europe outside the Iberian Peninsula, that is Naples and Sicily (until 1734), Lombardy and the southern Netherlands. Austrian ambitions to establish East Asian trade relations from Ostende and eventually build a small colonial empire overseas were thwarted by the Western Sea Powers’ veto. Moreover, England and France made their acceptance of a female successor to the Habsburg throne conditional

Map 3.1

The Habsburg Empire 1815–1918

Austria and Czechoslovakia on Austria’s renunciation of any trade activity from the Atlantic coast. Thus Maria Theresia (1740–1780) and Joseph II (1780–1790) could not make full profit from the southern Netherlands nor were they successful in integrating Lombardy – both regions were incorporated into Napoleonic France in 1797. Being prevented from expanding westwards or to the Indian Ocean, the Habsburg rulers started to concentrate on territorial acquisitions in east and southeast Europe. After having lost Silesia to Prussia in 1742, they made territorial gains through the partition of Poland (Galicia), the conquest of Moldavia (Bukovina) and the Venetian heritage in Dalmatia, which they accepted in exchange for the southern Netherlands and the Vorlande. After 1815 the Habsburg borders saw only minor changes with the loss of Lombardy (1859) and Venetia (1866) and the occupation of Bosnia and Herzegovina (1878). However, irrespective of the exact state borders at a given time, the Monarchy always encompassed a broad variety of regions differing in landscape, agricultural and civic constitution, nationality, religion, as well as economic assets and specialization. A first dividing line demarcated the provinces with a moderate feudal regime (Rentengrundherrschaft), which permitted peasants’ private agriculture, municipal autonomy and private industrial initiative, from those where the landed aristocracy exercised stronger power over its subjects (Gutsherrschaft). This line divided the Austrian Archduchies and the Alpine provinces from the eastern part of the empire with the Bohemian Lands representing a zone of transition. However, after the defeat of the Protestant provincial nobility in 1621 (Battle of the White Mountain), which set out to oppose re-Catholization and absolutistic advances, Bohemian manorial estates came into the hands of nobles who were loyal to the Emperor. Their properties were bigger, the amount of land cultivated by the feudal lord was higher and they developed a special type of manorial industrial enterprise, based on coerced labour (Wirtschaftsherrschaft). A second line separated the industrial provinces, which were situated in Lower and Upper Austria, Styria and the Bohemian Lands, from the agrarian provinces in the Alps, in Hungary and in the Carpathian and southern lands. Finally, a third line divided the territorial heartlands, characterized by a longer lasting affiliation to the Habsburg Empire, from marginal territories with only transitory status, such as the southern Netherlands, Lombardy and Venetia, and those old provinces which were lost to rival Empires, such as the Vorlande and Silesia. Textile regions were primarily situated within the moderate feudal zones and the Bohemian transition zone and less so in the Gutswirtschaft-zone. Textiles were of major importance in the industrial parts of the country, especially in Lower and Upper Austria, in Bohemia, Moravia and Silesia, in Vorarlberg and the Vorlande. However, two of these more advanced textile centres, Silesia and the Vorlande, were lost in the eighteenth century. The most advanced textile regions were located in the transitory margins of the Monarchy, that is in the southern Netherlands and Lombardy-Venetia. While they belonged to Habsburg Austria, these regions played a crucial role in the transfer of know-how and skilled labour. Due to their innovative character, these lost and transitory provinces will be included in this

45

The Ashgate Companion to the History of Textile Workers survey, while vast regions that were irrelevant for the production of textiles are excluded. The textile regions in the Habsburg Empire were included in various networks of differing scale. Within and between neighbouring textile regions, we observe a significant specialization, which encouraged trade and enabled merchants and producers to exploit the different skills, qualities and costs. Furthermore, many merchants and putting-out agents (Verleger) came from outside these regions, incorporating them into networks of trade that facilitated a transregional division of labour. Until the Thirty Years’ War, Austrian and Bohemian textile regions had strong ties with the textile centres of western and southern Europe. However, their position was subordinate to Venetian, Dutch and German merchants who controlled export trade. At the same time as the heart of Europe moved from the centre of the continent to the north-western Atlantic coast, east central Europe was de-linked. As a consequence, regional markets and networks became more important than transregional ones, and trade relations oriented towards east and south-east Europe. However, it was the internal market, unified by the abolition of customs and duties during the eighteenth century, which absorbed the major share of the growing market production. The variety of regional specialization in textiles affected a broad range of techniques and raw materials. Techniques included spinning and weaving, knitting, manufacturing of hosiery, lace-making, embroidery and so on. The main textile fibres were linen and wool from domestic production. Cotton only gained in significance when the Habsburg Monarchy reached a trade agreement with the Ottoman Empire following the Peace Treaty of Požarevac in 1718. Although the use of cotton rose quickly, it did not surpass linen as the basic textile fibre until the end of the Monarchy. Silk-weaving was an important industrial trade in Vienna, where the court was located. Propelled by the incorporation of Lombardy, a district of silk production, this branch came to dominate urban proto-industry in eighteenth- and early nineteenth-century Vienna. While the manufacture system was flourishing in the Austrian and Bohemian Lands and – supported by Swiss putting-out capitalists – to a lesser extent in Vorarlberg, textile activities in the rest of Habsburg Austria are not worth mentioning. These vast territories had neither relevant central manufactories nor any significant proto-industrial districts. The Kingdom of Hungary was almost devoid of textile market production. Some of the non-textile regions were purely agricultural, eventually supplying the industrialized districts with food; others were sites of different branches, such as mining, iron or glass industries. Both served as an important market for the textiles produced in the Bohemian and Austrian Lands. After the collapse of the Monarchy, the existing networks and the established division of labour were severely disturbed by the new national borders and the attempts of the successor states to sever the ties with Vienna, the former political 

Miroslav Hroch and Josef Petráň, Das 17. Jahrhundert – Krise der Feudalgesellschaft? (Hamburg, 1981), pp. 167–170. 46

Austria and Czechoslovakia core. The surviving regional patterns of textile production were integrated into the economic landscape of the new nation-states and their international alliances. However, the textile geography developing in the interwar period hardly had any time to settle. From 1938 to 1945 it was again being re-shaped in line with the political interests and the military-industrial priorities of the Third Reich, which annexed Austria and subjected Bohemia and Moravia. After the Second World War Austria and Czechoslovakia joined two opposing projects of European economic integration, which for a long period disconnected the most developed and interlinked areas of the Habsburg Monarchy until they were to meet again as competitors on the world market in 1990. The following chronological overview is guided by a periodization that combines territorial modifications, legal conditions and economic cycles with changes in technology, raw materials and the organization of production. One particular focus is on the specialization of and the division of labour between the different regions involved in the process. As will become apparent, each period was characterized by a specific spatial arrangement of interregional cooperation. Another focus is on the different types of labour and employment. Conceiving historical change in terms of inequality, unevenness and the contemporaneous overlapping of methods usually attributed to different eras, my explorations will reveal the simultaneity of different types of labour at a given moment rather than a chronological sequence of modes of production.

Rural subsistence and urban crafts (1650–1720) In the period prior to 1720, two organizational forms of textile production prevailed. Wherever textile fibres, mostly flax but also wool, were locally available, homemade textiles, chiefly for direct consumption, were one of the pillars of the rural economy. Domestic producers also provided the yarn that the textile artisans used for their more sophisticated processing of linen and woollen cloth. Located in towns, the artisans organized in guilds. A significant number still had an agricultural base – especially in small rural towns. Apart from these two dominant forms, the Viennese court supported a considerable number of silk weavers producing luxury goods. Cotton, by contrast, played only a marginal role, when it was woven with linen warp (fustian). The fustian industry had reached a first peak in late medieval times, with a centre in Lombardy and southern Germany, from where it spread into Austria. However, it was largely abandoned when the Ottoman wars impeded cotton supply and transport. Trade with the Ottoman Empire was in the hands of Greek or Serbian merchants, Vienna serving as their emporium for central Europe. Concentrated in the Austrian Archduchies and the Bohemian Lands, which joined the Habsburg Empire in 1526, rural crafts overwhelmingly produced for local and regional markets. Only very few locations, including the Mühlviertel as well as Jihlava and Liberec, specialized in export production of linen and draperies. 47

The Ashgate Companion to the History of Textile Workers There were a broad variety of producers. The majority of work was done within rural households, usually on a seasonal basis, by those members who were qualified to do the work and were not needed elsewhere; spinning was mainly, but not exclusively, a female occupation; weaving was chiefly male; children and old people fully participated in the work process. Artisan households showed a more distinct gender division of labour. While men – masters, journeymen and apprentice boys – carried out the artisan tasks, the female members of the family took care of the house and the garden and prepared the food. However, they were also responsible for the administration and marketing of the trade. As the households accommodated work – professional as well as reproductive – and life under one roof, they included not only family members but also migrant artisans and seasonal workers. Although men usually headed these units, women could also be in charge of them under certain conditions. Textile crafts were highly specialized. Depending on skills, workshop equipment and costs of material, income margins and social status differed. Besides different types of weaving, there were specific workshops for bleaching (linen), dyeing and fulling (woollen cloth). As water and water power were indispensable for their production, these workshops were usually located outside the urban centres along the rivers, often forming separate industrial districts. After the Thirty Years’ War, the nobility in the Austrian and Bohemian Lands invested in industrial sectors in order to make up for the loss of revenues and subjects. Apart from construction works, brewing and distilling, these aristocratic efforts targeted any trade that depended upon water access and water power. The resulting manorial workshops benefited greatly from the power the feudal lords exerted over their subjects – they could force them not only to work for them but also to buy their products – and therefore were a severe threat to private artisans as well as to municipal workshops. While in the Austrian Lands this type of manorial economy (Wirtschaftsherrschaft) lost significance by the middle of the eighteenth century, the nobility in the Bohemian Lands developed a more sophisticated system, which spread into glass works, mining, iron-works and textile manufacture during the eighteenth century.

Proto-industrialization (1720–1780) The period of proto-industrialization was characterized by an expanding puttingout system, organized by merchants, putting-out capitalists and manufacturers, which all relied on domestic work carried out by decentralized local producers. Their relationship with the merchant or putting-out capitalist varied: while some producers, who sold semi-finished products, were relatively independent 

Herbert Knittler, ‘Between East and West: Lower Austria’s Noble Grundherrschaft, 1550–1750’, in Charles Ingrao (ed.), State and Society in Early Modern Austria (West Lafayette, Ind., 1994), pp. 154–180; Herman Freudenberger, Lost Momentum: Austrian Economic Development 1750s–1830s (Vienna, Cologne and Weimar, 2003). 48

Austria and Czechoslovakia (Kaufsystem), others had a similar status to subcontractors being supplied with raw material and eventually also with devices (Verlagssystem). In many cases, domestic labour und putting-out agents were part of a larger unit of organization, which combined centralized activities (such as administration and marketing, preparation of raw materials and finishing, carried out in factory headquarters) with decentralized activities (such as spinning, winding and weaving, carried out in the homesteads of mostly rural producers). Centralization not only facilitated controls of labour and quality, it also served to bring together those stages of production that needed water and energy supply. Sometimes the centralized manufacture also included hand-weaving and, albeit less so, spinning. In the Austrian Lands manufactories were huge enterprises employing hundreds of workers in the central manufactories and thousands of putting-out workers. In the Bohemian Lands, however, manufactories were much smaller, usually not exceeding more than some hundred workers. Furthermore, they tended towards a more centralized structure. In Vorarlberg, the third major textile region, the putting-out system did not depend upon manufactories but on Swiss merchants who profited from the lower wages in Vorarlberg. Hoping for an increase in tax revenues, the state government promoted the establishment and growth of manufactories, which were also exempted from the guild regulations determining the number of masters, the apprenticeship and the control of quality. Conversely, guild rights were legally restricted. Facing fierce competition, they opposed the governmental privileges granted to manufactories. However, guilds were not generally hostile to new organizational patterns of work. On the contrary, they even set up their own networks of putting-out and joined the putting-out systems organized by manufactories. The first manufactories were founded in the second half of the seventeenth century (1662 Linzer Wollzeugfabrik; 1674 Manufakturhaus am Tabor) when  Markus Cerman and Sheilagh C. Ogilvie (eds), Protoindustrialisierung in Europa. Industrielle Produktion vor dem Fabrikszeitalter (Vienna, 1994), especially the contributions on Austria (Markus Cerman) and on Bohemia, Moravia and Silesia (Milan Myška); Andrea Komlosy, ‘Textiles Verlagswesen, Hausindustrie und Heimarbeit. Prototypen des informellen Sektors im 18. und 19. Jahrhundert’, in Andrea Komlosy et al. (eds), Ungeregelt und unterbezahlt. Der informelle Sektor in der Weltwirtschaft (Vienna, 1997), pp. 63–86, at p. 71.  Herbert Matis, ‘Betriebsorganisation, Arbeitsmarkt und Arbeitsverfassung’, in Herbert Matis (ed.), Von der Glückseligkeit des Staates. Staat, Wirtschaft und Gesellschaft in Österreich im Zeitalter des aufgeklärten Absolutismus (Berlin, 1981), pp. 411–479.  Arnošt Klíma, Economy, Industry and Society in Bohemia in the 17th to 19th Centuries (Prague, 1991), pp. 99–116.  Albert Tanner, Das Schiffchen fliegt, die Maschine rauscht. Weber, Sticker und Unternehmer in der Ostschweiz (Zürich, 1985), p. 23; Hubert Weitensfelder, Industrie-Provinz. Vorarlberg in der Frühindustrialisierung 1740–1870 (Frankfurt and New York, 2001), p. 19.  Leopoldine Hokr, Groß-Siegharts, Schwechat und Waidhofen/Thaya. Das Netzwerk der niederösterreichischen Baumwollindustrie (Waidhofen and Thaya, 2001), chapter 3.3; Komlosy, ‘Textiles Verlagswesen’, p. 68. 49

The Ashgate Companion to the History of Textile Workers mercantilist reforms came on the agenda. The new mode of production was given a boost in 1718. Repelling the Ottoman power, the Habsburg Monarchy conquered Hungary, Transylvania and Banat. Furthermore, the Treaty of Požarevac, which legitimized the eastward expansion, was followed by a trade agreement with the Ottoman Empire. With hopes for East Asian trade under the Habsburg’s flag having been shattered, the access to eastern and south-eastern Europe provided Austrian mercantilist ambitions with new territories. The Ottoman Empire, which suffered economic decline, not only opened new markets to Austrian industrial products but also promised cheap and regular supply of cotton. Trade was organized by the (second) Oriental Trade Company founded in 1719 – a first one, set up in 1667, stopped operating when the Ottoman troops seized Vienna in 1683. Endowed with a state privilege, the Oriental Company took over the Linz woollen cloth manufacture in 1722; it founded a cotton manufacture in Gross-Siegharts, Lower Austria, in 1720, and established the famous cotton manufacture in Schwechat near Vienna (Schwechater BaumwollManufaktur) in 1724. This enterprise was furnished with a cotton monopoly, which was only lifted in 1762 when other manufactories were granted similar privileges. In 1787 a general law codifying the foundation of factories replaced these privileges for select producers. Both the Linz cloth manufacture and the Schwechat cotton manufacture were centres of mass production. In 1790, when they reached their peak of employment, the Linz company employed 2,289 people on the site of the manufacture (4.6 per cent of the total workforce) and 47,947 putting-out workers (95 per cent).10 The Schwechat manufacture had 3,306 employees in Schwechat (10.5 per cent of the total workforce) and 28,046 putting-out workers.11 Enjoying specific imperial support because of their loyalty in the Thirty Years’ War, Bohemian nobles were most active in founding manufactories in the eighteenth century. As the Bohemian nobles wielded greater power over their subjects, who were forced to work in their manufactories, centralized workshops played a greater role in these enterprises. While these were smaller than in the Austrian Lands, they were much more numerous. Nevertheless, they also combined centralized elements with external domestic production.12 When the Habsburgs lost almost the whole of Silesia, their most developed industrial province, to Prussia in 1740, they were forced to concentrate their industrial policy on the remaining parts of the country. In 1755 Johann Ludwig  Viktor Hofmann, Die Wollzeugfabrik zu Linz a. d. Donau (Vienna, 1919); Viktor Hofmann, Die Anfänge der österreichischen Baumwollwarenindustrie in den österreichischen Alpenländern im 18. Jahrhundert (Vienna, 1926).  Gustav Otruba, Österreichische Fabriksprivilegien vom 16. bis ins 18. Jahrhundert (Vienna, Cologne and Graz, 1981), p. 30. 10 Helmut Lackner and Gerhard Stadler, Fabriken in der Stadt. Eine Industriegeschichte der Stadt Linz (Linz, 1990), pp. 91–93. 11 Andrea Komlosy, An den Rand gedrängt. Wirtschafts- und Sozialgeschichte des Oberen Waldviertels (Vienna, 1988), p. 29. 12 Jana Machačová and Jiří Matějček, ‘Entwicklung der Textilindustrie in den böhmischen Ländern in den Jahren 1781–1848’, Hospodárské dějiny, 18 (1990): 211–244. 50

Austria and Czechoslovakia von Chamaré, a Silesian aristocrat who had moved to Bohemia, founded the Potštejn linen manufacture.13 A prominent example for the wool sector is the Kladruby manufacture, founded in 1749 and transferred to Brno in 1764 (Brünner Wolltuchfabrik) because of the strong government support bestowed on the industrial development of this region.14 Cotton manufactories were concentrated in northern Bohemia, where the end of the Schwechat cotton monopoly gave rise to new establishments. The first one was the noble manufacture at the Kosmanosy manor. The biggest was founded in 1770 by a bourgeois entrepreneur, Johann Josef Leitenberger, in Verneřice/Wernstadt, occupying 100 persons in the central workshop and 2,000 dislocated spinners and weavers in their cottages.15 In the next decades Leitenberger and his sons bought several other cotton factories, establishing Leitenberger among the leading cotton producers. Textiles were by far the largest sector of industrial activity in the eighteenth century and they expanded quickly. Their expansion depended upon two crucial factors: the supply of raw materials and the availability of cheap labour. In order to increase output, the production of raw textile fibres had to be raised and imports promoted. Between 1650 and 1750 it was the flax industry that grew the most, while in the second half of the eighteenth century it was woollen cloth, essential for army supply. The manufacturing of cotton only started in 1720. After a first period of steady growth, the introduction of cotton mills in 1800 initiated the exponential increase of this branch at the expense of flax. The Viennese silk industry on the other hand profited from the incorporation of Lombardy, a silk-producing province, into the Austrian Monarchy. The manufacture of each of these fabrics was usually concentrated in a particular area and characterized by a specific arrangement of labour organization. While the weaving of linen remained a local craft, woollen cloth and cotton (with or without linen) were produced in the manufacture system. Within this system cotton and fustian manufacturing, relying on domestic labour, differed from woollen cloth manufacturing, where artisans played an important role. The Viennese silk industry, by contrast, used a continuum of labour forms, including traditional craft practices, putting-out systems and various forms of subcontracting and proletarianized wage work, which proves that a proto-industrial mix of labour was not restricted to the countryside.16 13 Klíma, Economy, pp. 105f; Bohumír Smutný, Potštejnská manufaktura na česko-kladském pomezí (Hradec Králové, 2002). 14 Herman Freudenberger and Gerhard Mensch, Von der Provinzstadt zur Industrieregion. Ein Beitrag zur Politökonomie der Sozialinnovation, dargestellt am Innovationsschub der industriellen Revolution im Raume Brünn (Göttingen, 1975), pp. 13f; Freudenberger, Lost Momentum, pp. 154–162; Jan Janák, Hospodárksy rozmach Moravy 1740–1918 (Brno, 1999), pp. 13f. 15 Johann Slokar, Geschichte der österreichischen Industrie und ihrer Förderung unter Kaiser Franz I. (Vienna, 1914), pp. 288f. 16 Helene Deutsch, Die Entwicklung der Seidenindustrie in Österreich 1660–1840 (Vienna, 1909); Josef Ehmer, Familienstruktur und Arbeitsorganisation im frühindustriellen Wien 51

The Ashgate Companion to the History of Textile Workers The recruitment of labour varied according to the type of enterprise. Artisan workshops relied on patrimony, marriage and professional careers within the guild system, each linked with specific patterns of migration. Centralized manufactories recruited not only domestic but also foreign artisans. The latter usually originated from those regions which belonged to Austria in the eighteenth century: linen and woollen cloth weavers came from the Austrian Netherlands; Emperor Franz Stephan imported Lorraine specialists for the imperial gobelin works; Svabia provided fustian weavers; and Viennese silk artisans were of Italian origin. A great number of foreign technicians, apprentices and other artisans followed their masters or entrepreneurs from the industrial provinces on the Habsburg fringes, when these were successfully established in the Austrian heartlands. Thus the Austrian Netherlands, the Vorlande and Lombardy, which did not take part in the unification of the Habsburg market, were of critical importance for the transfer of capital, technology and know-how, and gave important impetus to the industrial development of mainland Austria.17 Unskilled labour for domestic industries and manufactories was recruited in the rural neighbourhoods of the proto-industrial districts. A set of laws weakening the landlord’s grip over his subjects mobilized rural society. First and foremost serfdom, which was still in force in the Bohemian Lands, was finally abolished in 1781. Furthermore, liberalization of marriage, travelling and choice of profession severed the strong ties between subjects and the land. However, as the manufacturing system also needed domestic labour, the entrepreneurs encouraged new settlement and family foundation in proto-industrial areas. This development created a growing class of people who owned a house and eventually a small plot of land, which served to feed the family working in the textile industry.18 Because of traditional gender responsibilities in the rural economy, women were generally better prepared than men to respond to the proto-industrial mobilization of additional domestic labour. As long as spinning was carried out by hand, women dominated the textile industrial labour force.

Responses to the British challenge (1780–1873) Until 1780 the development of the textile sector was characterized by gradual improvements of techniques and organization, linked with a growing output, an (Vienna, 1980); Sylvia Lausecker, ‘Vor- und frühindustrielle Produktionsformen am Beispiel der Seiden- und Baumwollindustrie in Wien 1740–1848’ (unpublished dissertation University of Vienna, 1975). 17 Herman Freudenberger, ‘Technologie-Transfers von England nach Deutschland und insbesondere Österreich im 18. Jahrhundert’, in Ulrich Troitzsch (ed.), Technologischer Wandel im 18. Jahrhundert (Wolfenbüttel, 1981), pp. 105–124. 18 Michael Mitterauer, ‘Lebensformen und Lebensverhältnisse ländlicher Unterschichten’, in Matis (ed.), Von der Glückseligkeit, pp. 315–338, at pp. 324ff. 52

Austria and Czechoslovakia increasing number of workers and a regional expansion of textile districts. This slow progress would probably have resumed after the worldwide recession of the early 1770s, which did not spare the Habsburg Monarchy, had it not been for the mechanization of the spinning process in Great Britain. This invention had a massive impact on the worldwide production of textiles. If traditional manufactories did not want to cede the market to machine-yarn produced in and around Manchester, which continental merchants and weaving manufacturers were eager to buy, they had to react. The manufacturers in the Habsburg Monarchy came up with four strategies, which sometimes complemented and in other cases contradicted each other. Firstly, a highly protectionist duty system was introduced in 1784 prohibiting the import of nearly all finished cotton cloths and putting high taxes on cotton yarn. At the same time the government pursued the integration and liberalization of the internal market. Secondly, the proto-industrial manufactories markedly increased their production. As each additional weaver needed eight to ten spinners to supply him or her with yarn, the growing production of the manufacture system was linked with a surge in spinners, which tripled their numbers in Lower Austria between 1779 and 1790. Because of the decentralized structure of the system, the demand for labour led to a geographical expansion of proto-industrial districts, which massively increased the expenses for transport, management and control. Thirdly, the Austrians attempted to develop their own competitive spinning machines. From 1780 onwards the innovative initiatives multiplied. Many of these received government support and were granted privileges for the use of certain devices.19 However, for various reasons, the multifarious attempts did not engender a decisive breakthrough towards an effective spinning machine. British textiles were attractive and merchants as well as producers imported them in spite of the restrictions. Only the fourth response put a stop to this attraction: single producers as well as government agents travelled to Britain and tried to spy out the technologies, which were highly protected. At the same time Austrian industrialists encouraged British engineers to join their companies in order to develop spinning machines nach britischer Art (‘in the British way’). Bohemian and Lower Austrian manufacturers competed to be the first to introduce the mill system in cotton spinning with the help of British technicians. A first cotton mill was opened by the Leitenberger company, Verneřice, in 1797, another one on the Auersperg estate of Seč (Bohemia) in 1801.20 In the same year a group of prominent noble shareholders set up the 19 Leopoldine Hokr, ‘Johann Michael II. von Grosser und die Maschinenspinnerei in Niederösterreich’, Das Waldviertel 46 (1997): 30–44; Günther Luxbacher, ‘“Alles Baumwollengespinst ist nunmehr Maschinengespinst.” Die Maschinisierung der österreichischen Textilindustrie und ihre Ergebnisse im National-FabriksproduktenKabinett’, in Technisches Museum Wien (ed.), Das k.k. National-FabriksproduktenKabinett. Technik und Design des Biedermeier (Munich and New York, 1995), pp. 68–89, at p. 70. 20 Luxbacher, ‘“Alles Baumwollengespinst ist nunmehr Maschinengespinst”’, p. 71. 53

The Ashgate Companion to the History of Textile Workers k.k. priv. Garnmanufakturgesellschaft in Pottendorf (southern Lower Austria). Much bigger in size, the latter came to be known as the first cotton mill in the Habsburg Monarchy. Protected from British competition by the Continental Blockade, many mill projects were realized in the following years. The biggest companies were concentrated in southern Lower Austria; two smaller centres of mechanical cotton spinning developed in northern Bohemia and in Vorarlberg.21 The introduction of the cotton mill was restricted to single towns and villages. Moreover, it only affected spinning. Table 3.1 shows the sudden decline in the number of spinners between 1790 and 1811. The substitution of this essential process by machines heralded the end of the decentralized manufacturing system. The biggest manufactories, the Schwechater Baumwollmanufaktur and the Linzer Wollzeugfabrik, closed down in 1820 and 1850 respectively. Conversely, the power-loom took much longer to be incorporated into the factory system. Although Edmund Cartwright invented it in 1785, it was not until 1834 that the first power looms were successfully set up in an Austrian factory.22 Only after 1873 did power-loom weaving become the dominant mode of production. Table 3.1 Industrial workers in Lower Austria, 1762–1811 1762 Men and widows

1769

1775

1779

1783

4,349 8,338 14,013 19,235 17,036

1785

1788

1790

1811

25,483

35,531

35,825 43,054

Womena

15,378 27,986 22,117 41,995 77,058

Total

19,727 36,324 36,130 61,230 94,094 117,014 161,381 182,473 75,435

a

91,531 125,850 146,648 32,381

mostly spinners

Source: Komlosy, An den Rand gedrängt, p. 35.

After the end of the Napoleonic Wars, cotton manufacturing used a variety of techniques that also determined the organization of the work process. By the second half of the nineteenth century, spinning was centralized in mills, driven by water power, while weaving was still put out to handloom weavers scattered over rural textile regions. The centralized manufacture was replaced by a new type of enterprise whose headquarter workshops were usually located in a capital or a central industrial town. Their owner, called a Fabrikant, acted as a kind of manager, supplying decentralized home-weavers with machine-spun yarn. Finishing was carried out in centralized units. The most important of the finishing processes 21 Slokar, Geschichte der österreichischen Industrie, pp. 293–312. 22 Weitensfelder, Industrie-Provinz, p. 108. 54

Austria and Czechoslovakia was printing, which mechanized during the first half of the nineteenth century. Lace-making, which was equally based on cotton, was organized in a similar way, combining mechanized spinning with decentralized lace-making. Domestic lacemakers were concentrated in Vorarlberg, where the rural population worked for Swiss entrepreneurs.23 Technological innovations were first adopted in the cotton sector. They paved the way for similar experiments in the processing of flax, wool and silk, which were, however, introduced with a delay of some decades. Although these sectors reproduced the dualism between mechanized spinning and domestic weaving, mill spinning was restricted to cotton for a quarter of a century. All other materials were still spun by hand. This technological advantage accounts for the quick expansion of cotton at the expense of other textile sectors. Nevertheless, those branches remained important until the end of the nineteenth century. Technological improvements, however, were not restricted to spinning. The mechanization of block printing and the introduction of rotary printing were important innovations for the cotton sector in the first half of the nineteenth century. Another crucial invention was the Jacquard machine. First fitted to handlooms and later adapted to power-looms, this mechanical device permitted the weaving of sophisticated patterns. It was widespread in the Viennese silk industry, which specialized in complicated designs. Another mechanism, which was incorporated into ribbon looms prior to the introduction of the power-loom, was the ribbon mill. While traditional craftsmen could only ever weave one ribbon at a time, this invention facilitated the simultaneous production of several ribbons. The Austrian Monarchy granted regional privileges for the introduction of the ribbon mill, which explains the narrow regional concentration of this branch, strongly and persistently opposed by the ribbon craft. The centre of the Austrian ribbon industry was located around Waidhofen/Thaya and Gross-Siegharts in the north-western part of the Waldviertel (Lower Austria).24 Apart from enhancing selectivity and unevenness, the mechanization of textiles modified the geographic patterns, entailed new forms of labour and changed the organization of the manufacturing process. The number of textile centres in rural regions increased markedly. Eventually single factory villages or towns became so dense that contemporary observers were struck by the ‘continuous series of factories, just like in certain English regions’.25 One of the rural areas particularly affected by this transformation was southern Lower Austria, which came to be known as the Industrieviertel (industrial district) after numerous spinning mills had been set up along its rivers. Others were the Vorarlberg Rhine valley, the area surrounding Liberec in northern Bohemia and Brno in southern Moravia, which was to become the ‘Manchester of the Habsburg Monarchy’. These towns and districts attracted migrant labour from both neighbouring and more distant regions. 23 Tanner, Das Schiffchen fliegt, pp. 99 ff; Weitensfelder, Industrie-Provinz, pp. 144–152. 24 Leopoldine Hokr, ‘Bandel in Handel und Wandel’, Das Waldviertel 38 (1989): 124–134. 25 Wenzel C.W. Blumenbach, Neueste Landeskunde von Oesterreich unter der Enns (Güns, 1834/35), p. 135. 55

The Ashgate Companion to the History of Textile Workers The proto-industrial putting-out districts, which did not qualify for mill plants, suffered a rapid decline in domestic spinning, which had provided the rural population with an additional income. In some cases, the workers who were made redundant turned towards agricultural production, reducing their textile activities to wool and linen.26 Other proto-industrial districts, namely in northern Lower (Waldviertel) and Upper Austria (Mühlviertel), in rural Vorarlberg as well as in many peripheral rural textile areas in Bohemia and Moravia, experienced a boom in hand-weaving which replaced spinning. Hand-weaving from now on was generally considered a male profession in public opinion and in trade statistics. Women who concentrated on preparing and finishing tasks were no longer considered ‘workers’. According to the modern family ideology, they were transformed into helping hands of husbands and fathers, their textile activities placed by definition among housewifely duties. Rural weavers mainly were cottagers with small plots of land for self-sufficient production of food. Although self-employed, their income entirely depended upon urban textile entrepreneurs, who supplied them with orders and raw materials. As long as weaving depended on their hands, they enjoyed a certain prosperity shaping their identity as rural peasants or cottage weavers. However, when mechanization set in, they had to increase their output in order to compensate for the loss in wages beaten down by the mechanical rationalization of production.27 The industrial branches, which settled in towns with court and capital functions, were usually textile crafts. Nevertheless, these crafts also came under the pressure of technological and organizational change caused by the introduction of the factory system, which they both resisted and adapted to. Vienna was the most important centre of demand for tailor-made textiles. It housed a luxurious silk industry, which not only employed all types and combinations of craftsmen but also used centralized manufactories as well as putting-out activities to homeworkers and decentralized subcontractors. Moreover, Vienna was the home of many entrepreneurs investing in other textile sectors that relied on traditional craft workshops. Apart from carrying out minor artisan tasks, such as preparation and some finishing, the workshops were mainly responsible for the administration of the business and put out weaving into rural weavers’ households, especially in north-western Lower Austria (Waldviertel). Finally, Vienna accommodated numerous artisans specializing in specific luxury products. Giving rise to a growing craft sector, this new niche of textile production resulted in an urban variant of the putting-out system, which lasted until the end of the Monarchy.28 26 Alfred Hoffmann, ‘Die Agrarisierung der Industriebauern in Österreich’, East European Quarterly III (4) (1972): 449–468; Mitterauer, Lebensformen, p. 329. 27 Andrea Komlosy, ‘Stube und Websaal. Waldviertler Textilindustrie im Spannungsfeld zwischen Verlagswesen, Heim- und Fabriksarbeit’, in Andrea Komlosy (ed.), Spinnen – Spulen – Weben. Leben und Arbeiten im Waldviertel und anderen ländlichen Textilregionen (Krems-Horn, 1991), pp. 119–138, at p. 136. 28 Susan Zimmermann, ‘Frauenerwerbsarbeit und Haushalt im Wien der Jahrhundertwende’, Archiv 1990. Jahrbuch des Vereins für Geschichte der Arbeiterbewegung 56

Austria and Czechoslovakia While Vienna did not attract any industry that was dependent on water power, the situation was entirely different in the provincial capitals. Brno (Moravia) and Linz (Upper Austria) became centres of the woollen cloth industry, Prague a centre of cotton printing. Combining administrative and industrial functions, these capitals not only attracted migration but also employed handloom weavers in adjacent textile regions. Some towns in rural textile districts became sites of industrial production themselves, setting up a putting-out system of weaving on a local or regional scale. Textile production still used multifarious forms of labour. Earlier forms and combinations, including domestic production crafts, centralized mechanized workshops and activities put out into rural producers’ households, were supplemented by a new type engendered by the factory system: the proletarian textile worker in spinning or printing mills. At the time, industrial wage work in factories involved extreme exploitation.29 The new factory villages disregarded traditional social norms, and new conventions such as social bargaining, charity or workers’ organization had not yet developed. Mills were comparable to military barracks: they had a stranglehold over their employees from morning to evening. Workers were shut up in the factories, which they only were allowed to leave on Sundays. Providing food and lodging, spinning mills recruited children and young adults and made them work as many hours as possible until they were replaced by newcomers. Only very few industrial workers, who were skilled and exercised control over other workers, received stable and secure working conditions. Labour protest took various forms. On the one hand, skilled manufactory workers developed forms of social and cultural bonding already practised by guilds and journeymen. On the other hand, protest was expressed by means which popular movements had always used to blame famine, rising prices, layoffs, income or legal degradation by directly addressing the person or the object considered to be responsible. This protest arose spontaneously; it broke out when new machines put pressure on wages and employment. A wave of labour unrest occurred throughout the major textile regions of the Monarchy when the introduction of the powerloom and new machines for textile printing coincided with the severe economic crisis of the 1840s, which caused wage cuts and mass dismissals. Workers’ actions in Prague and northern Bohemia, as well as Vienna and southern Lower Austria, culminated in 1844, when they were inspired by the Silesian Weavers’ Uprising, and in 1848, when they were fuelled by the Bourgeois Revolution, of which they formed an integral part.30 The workers’ protest went hand in hand with bread-riots 6 (Vienna, 1990), pp. 92–122. 29 Andrea Komlosy, ‘Alles spinnt. Die frühe mechanische Baumwollspinnerei in Niederösterreich’, in Roman Sandgruber et al. (eds), Magie der Industrie. Leben und Arbeiten im Fabrikszeitalter. Katalog zur NÖ. Landesausstellung 1989 in Pottenstein an der Triesting (München, 1989), pp. 302–313, at p. 306; Weitensfelder, Industrie-Provinz, pp. 254–275. 30 Klíma, Economy, pp. 197–204; Karel Novotný, Severočestí tiskari kartounu v první polovině 19. století, 1 (Prague, 1993); Wolfgang Häusler, ‘Die Petition der Spinnereiarbeiter des 57

The Ashgate Companion to the History of Textile Workers and machine breaking, mainly carried out by weavers and printers. At the same time new forms of collective bargaining came up, not only on the level of the single enterprise, but for all workers of a specific textile branch. Mule spinners of the Lower Austrian industrial district submitted a petition to the new Parliament of 1848, asking for legal regulations and social rights for workers. With the repression of the Revolution, workers were equally silenced until 1867, when a Constitution granted the ‘right of assembly’. Two years later, the ‘freedom of association’ legalized unions, strikes and collective bargaining.31 From 1815 to 1873 the mechanization of spinning and printing gradually expanded. Furthermore, it spread from cotton to other fibres, involving more and more people and locations in the process of industrialization, which increased the number of proletarian workers. Although the mechanization of weaving started in the 1840s, until the crisis in 1873, weaving largely remained a manual activity, which entailed the simultaneous expansion of semi-proletarian hand-weaving. The improvement of transport facilities, especially the railway system which was built from 1840 onwards, allowed the creation of networks which facilitated an unequal division of labour between various locations with their different organizations of labour. Operating on both a regional and a supra-regional scale, these networks not only included the industrialized regions but also incorporated those provinces that supplied migrant labour and raw materials or served as marketing areas.

Changing regional and sectoral patterns (1873–1918) The year 1873 was a turning point for the world economy which after an expansive cycle with high rates of growth fell into a period of crisis characterized by economic decline on the one hand, and a desperate search for means to initiate renewed growth on the other. The economic crash occurred in the very year that Vienna hosted the World Exhibition. As a result of the collapse, many companies faced bankruptcy. In order to stay competitive, producers had to reduce costs. Hence they rationalized, lowered wages and looked for new geographical arrangements. Investment was most profitable in the sectors serving as new locomotives of growth, such as steel, machines, electricity, food processing and the chemical industry. Compared with the textile industry, these branches attracted higher rates of investment and increased the number of their employees in the last decades of the nineteenth century. While the highly profitable branches were concentrated in the industrial centres, less profitable sectors were often located in peripheral regions. In the textile industry, however, central as well as peripheral regions were affected, and the interregional division of labour within the textile sector underwent significant changes. Viertels unter dem Wienerwald an den österreichischen Reichstag (1848)’, Unsere Heimat 48 (Vienna, 1978): 3–21; for Vorarlberg: Weitensfelder, Industrie-Provinz, pp. 311–314. 31 Fritz Klenner, 100 Jahre Österreichische Gewerkschaftsbewegung. Entstehung und Entwicklung (Vienna, 1981), pp. 20f. 58

Austria and Czechoslovakia A major consequence of the crisis was that all those industrial producers who still operated manually introduced mechanical power. This process exercised enormous pressure on handicrafts, which were reduced in number and significance, and on handloom weavers, who often found alternative sources of income in the knitting and lace-making industry. Producing broad pieces on mechanical knitting frames in factories, this branch put out the knitting of smaller items to hand-machine workers. Furthermore, knitwear required many very labour-intensive finishing processes like seaming, linking and so on, which were usually carried out by female homeworkers.32 Similarly, lace-making machines, which demanded special skills, were most suitable for operation in domestic workshops.33 Rationalization also seized the garment industry. Handmade clothes became a luxury product, when the tailor’s business in the last third of the nineteenth century turned from a handicraft into an industrial activity. However, the ready-made production of clothes was neither mechanized nor centralized. Since the sewing machine was suitable for decentralized use and clothes-making included other labour-intensive operations, which had to be carried out manually, the branch was decentralized, putting out the single steps to individual workshops and households. Located in the centres of consumption, namely in Vienna, the clothing industry became the dominant urban textile branch of the late nineteenth century. Irrespective of the branch and material, mechanization seized the majority of the stages in the individual production chains. Factories spread all over the countryside, with single regions specializing in specific stages of the working process. Central industrial agglomerations faced a shift from textiles towards other sectors, while often maintaining those textile factories founded in earlier periods. Furthermore, new weaving or knitting factories were set up in rural textile areas that used to be putting-out regions. The silk branch, which had been concentrated in Vienna until 1873, transferred the whole production chain to the Bohemian Lands. In many cases the headquarters of the companies stayed at their (previous) urban locations in consumer centres, chiefly in Vienna. However, they were often transformed into mere managing and marketing units, while the manufacturing processes were pushed into the periphery.34 Traditional crafts fell victim to the overwhelming competition of the factory system. However, many of them were able to find specific niches in which to survive. Equally, machine-weaving did not put an end to homework. On the contrary: moving into new fields of operation, handicrafts and homework expanded along with the factory system. On the one hand homeworkers took over those steps of the production process which were not yet mechanized, such as the weaving of complicated and irregular patterns on a Jacquard loom or specific finishing and application tasks which demanded manual skills. On the other hand they carried 32 Komlosy, An den Rand gedrängt, pp. 75f. 33 Weitensfelder, Industrie-Provinz, pp. 151f; cf. Santina M. Levey, ‘Machine Made Lace: The Industrial Revolution and After’, in David Jenkins (ed.), The Cambridge History of Textiles, II (Cambridge, 2003), p. 858. 34 Komlosy, An den Rand gedrängt, pp. 78–80. 59

The Ashgate Companion to the History of Textile Workers out ordinary winding and weaving, serving as a cheap labour force, which could be taken on in times of high demand and dropped in periods of stagnation.35 These homeworkers were in direct competition with the machines, thus suffering from enormous pressure on their wages. The spread of the factory system gave rise to an organized labour movement, which started to take shape in the 1870s.36 It was characterized by heavy disputes on ideological questions as well as on the tactics of political action. The split between ‘Radicals’ and ‘Moderates’ was overcome by the foundation of a single Labour Party (Sozialdemokratische Arbeiterpartei, SDAP), which voted for a Marxist programme and at the same time for participation in the parliamentary system. Regional disparities and the striving for national liberation, however, led to a split of the SDAP along ethnic lines. The Czech Party split from the German SDAP, the same applied to the trade unions’ congress. The diversity in social and national terms was reflected by an organizational heterogeneity of the labour movement. Although textile workers were numerous among party and union members and carried out a number of important disputes and strikes, they did not belong to the labour movement elite. On the one hand, this under-representation is due to the declining dynamics of textiles as a leading industrial branch, on the other hand, it corresponds with the feminization of textile production as a consequence of mechanization and the stagnation of wages in comparison with the new leading sectors. At the same time the rural preponderance and the high percentage of craftsmen and homeworkers prevented textile labour from developing a typical proletarian character. Craftsmen and homeworkers, who often were self-employed, rejected the proletarian identity propagated by the organized labour movement. Sometimes the rural poor had no other choice than to accept their precarious working conditions, especially if their area did not offer any alternative sources of income. As official labour records complained, organization outside urban and industrial centres was very scarce. The dire living conditions of Austrian and Bohemian home-weaving families at the end of the nineteenth century were well documented in a new genre of journalism, which published heartbreaking on-the-spot accounts in the social democratic press.37 Government inquiries also dealt with the weavers’ poverty.38 Notwithstanding these circumstances, the inhabitants of many former proto-industrial regions preferred homework to industrial wage work because they cherished their economic independence and the possibility to be their own masters. Home-weaving families showed great flexibility and initiative in earning their living. They accepted seasonal wage work as well as employment as servants

35 Komlosy, ‘Stube und Websaal’, p. 124; Hugo Herz, Die Heimarbeit und der Notstand der Heimarbeiter in der mährischen Textilindustrie (Brünn, 1904). 36 Klenner, 100 Jahre Österreichische Gewerkschaftsbewegung, pp. 23, 34–39. 37 Stefan Riesenfellner, Der Sozialreporter Max Winter im alten Österreich (Vienna, 1987). 38 Handels-Ministerium (ed.), Bericht der k. k. Gewerbeinspectoren über die Heimarbeit in Österreich (3 vols, Vienna, 1900/1901). 60

Austria and Czechoslovakia and were also prepared to migrate to other places.39 Hence, resembling the old manufactory handwork putting-out system, rural weaving factories were de facto ‘putting-out factories’. The same applies to knitting and lace-making factories. They combined factory work with specialized external production, which offered alternative sources of income to former weavers.

The Czechoslovak textile industry until the collapse of the Iron Curtain (1918–1989) The collapse of the Austro-Hungarian Monarchy in 1918 broke up the established interregional division of labour and a large internal market. From one day to the next, different parts of the textile production chain were situated in separate states, and interregional became international relations. Czechoslovakia inherited the major share of Austro-Hungarian textile capacities. Furthermore, including spinning, weaving, knitting and finishing units processing cotton, wool, silk, flax hemp and synthetics, these were more diversified than in other parts of the former empire. The territory of the Czechoslovak Republic incorporated 71.7 per cent of all cotton spindles of the former Habsburg Monarchy, 89.9 per cent of cotton looms, 74.3 per cent of wool spindles, 96.9 per cent of linen spindles, 52.1 per cent of hemp spindles, 66.4 per cent of jute spindles and 89.9 per cent of silk facilities.40 Profiting from these existing structures, post-war reconstruction simply had to fill the gaps in the production chain of the new national market. The most important Czech textile regions bordered Austria in the south and Germany and Silesia, which became part of Poland after the Second World War, in the north. A large majority of the population in these regions was German-speaking. Accordingly, textile capital was mainly in the hands of German-speaking owners, many of them Jewish, except for a few branches located in eastern Bohemia, whose owners were predominantly Czech. In order to diminish the Austrian influence, the Czechoslovak government promoted the so called ‘nostrification programme’ in 1919, which aimed to replace German and Austrian owners with Czechs or at least with West Europeans. By 1924, the textile sector accounted for 33 of the 133 ‘nostrified’ companies.41 Comprising only 13 million inhabitants, the Czechoslovak market was far too small to absorb the domestic production that used to supply 52 million people. 39 Andrea Komlosy, Grenze und ungleiche regionale Entwicklung: Binnenmarkt und Migration in der Habsburgermonarchie (Vienna, 2003), p. 256. 40 Ivan Jakubec, ‘Textilindustrie in der Exportpolitik der Tschechoslowakei (1918–1938)’, unpublished paper presented at the conference Textilregionen an Böhmens und Mährens Regionen (Museum Alte Textilfabrik Weitra, 12–14 June 1992), p. 1. 41 Jakubec, ‘Textilindustrie’, p. 2; Alice Teichova, Wirtschaftsgeschichte der Tschechoslowakei 1918–1980 (Vienna, 1988), p. 61. 61

The Ashgate Companion to the History of Textile Workers Hence the future of the Czechoslovak textile industry depended on the search for new markets. These were mainly found in Austria, which lacked certain products or semi-products traditionally manufactured in Bohemia or Moravia, and in other Eastern European successor states as well as in south-east Europe, whose textile capacities were lagging behind. In 1920 Czech factories exported 10,779 tons of cotton products, 48 per cent of which went to Austria, 12 per cent to Hungary, 6 per cent to Germany, 6 per cent to Yugoslavia and 5.5 per cent to Poland. Manufactured woollen exports amounted to a total of 10,565 tons, 52 per cent of which were sold in Austria, 14 per cent in Hungary, 14 per cent in Germany, 4 per cent in Yugoslavia and 3 per cent in Poland.42 Textiles accounted for 85 per cent of total Czech imports in the Balkan states in 1920/21. In the following years Czech textiles encountered several barriers.43 When the east and south-east European states established their own textile industry, they protected their markets against Czech manufactures. While Czech textiles were not completely replaced by domestic production, the imports shifted from finished to semi-finished products, such as yarn. Czech companies reacted by founding local subsidiaries in order to circumvent these protective tariffs. Another obstacle for trade was the insufficient currency reserves for imports, which Czechoslovakia tried to counter by resorting to barter. For example, textiles were exchanged for tobacco. The Czech efforts met with similar German strategies that ousted Czechoslovakia from east and south-east European markets and eventually integrated the whole region into the ‘German Grossraum’.44 Like other consumer industries, the textile sector was severely hit by the world economic crisis of 1929. Internal and external demand collapsed. In the wake of the crisis West European states as well as Czechoslovakia introduced restrictions against agricultural imports from south-east Europe. As a consequence, these countries intensified their exchange with Germany, which accepted raw materials for German manufactured goods. Specialized in manufactured exports to southeast Europe, the Czechoslovak industry fell victim to this strategy, losing markets in east and south-east Europe. Austria also reduced its imports from Czechoslovakia. Between 1924 and 1937 Czechoslovak exports to Austria and the other successor states of the Monarchy fell from 40.3 per cent (1924) to 35.1 per cent (1929) and 29.9 per cent (1937).45 Textiles were particularly affected by this loss in market shares and, as a consequence, lost their dominant position in the Czechoslovak economy. In the 1930s many textile workers were made redundant, several companies went bankrupt, hand-weaving and homework faced a final collapse and unemployment rose. Since the unemployment rate was higher in the German-speaking border regions than the national average, German nationalist parties blamed the Czech government for neglecting the German minority. This situation fuelled the idea 42 Jakubec, ‘Textilindustrie’, p. 3. 43 Ibid. 44 Alice Teichova, Kleinstaaten im Spannungsfeld der Großmächte. Wirtschaft und Politik in Mittel- und Südosteuropa in der Zwischenkriegszeit (Vienna, 1988), pp. 188ff. 45 Teichova, Wirtschaftsgeschichte der Tschechoslowakei, p. 54. 62

Austria and Czechoslovakia of joining Germany, which, under Hitler’s government since 1933, agitated for the annexation of the Czech border regions. Following the Munich Agreement in September 1938, where Great Britain, France and Italy had approved the German annexation, the so-called Sudetengebiete were incorporated into the Third Reich. The major share of Czech textile regions were thus integrated into Germany, which supplanted Jewish with German owners. During the Second World War, when the rest of the Czech territories became a German protectorate, priority was given to heavy industry and arms production. Textile industries had to serve military demand, and factories often were used for other purposes than textiles. After the Second World War, when Czechoslovak state unity was restored, private property faced two waves of nationalization.46 In 1945 the government expropriated German citizens and so-called collaborators, which affected the majority of textile firms in the annexed border areas. When the Communists assumed power, they initiated a second wave of nationalization, which was directed against private ownership in general, and seized all of the remaining private textile companies. As the limited space of this chapter does not allow a detailed description of a planned economy, I will summarize those features that had a direct impact on the Czechoslovak textile industry.47 Firstly, in the post-war reconstruction period priority was given to heavy industries and the production of machines, while consumer industries were neglected. As a consequence the share of textile production decreased. This process effected a diversification of the economic structure, including a specialization in the production of textile machines. Secondly, as in the other sectors, the nationalization of the textile industry involved a reorganization of the branch. Thus single units were incorporated into big company structures, which gathered all of the stages of production for a particular material. This structure facilitated the introduction of the latest techniques, including automation, but it did not necessarily promote high quality. The planning mechanism did not only destroy the established geographical patterns, it lacked any continuity, as each planning period entailed a change of the industry’s organizational structure. Thirdly, location and investment followed a political logic. Previous patterns of local specializations were replaced by investments, which served regional or social purposes, such as the creation of jobs for women. The newly established state-owned branch companies sometimes comprised decentralized factories in rural areas, thus promoting industrial work all over the country. For the first time in Bohemian history, textile labour was purely proletarian. Workers enjoyed job security and a high level of social rights. They did not have a voice in shaping industrial relations, which were determined by state, party and unions; unions acted under state control. Investment priority was given to Slovakia in order to promote the industrialization of this predominantly agrarian area of the country. Slovak industrialization efforts concentrated on heavy industries and arms production whereas textiles sometimes featured in but never dominated the 46 Ibid., p. 88. 47 Ibid., pp. 112–128. 63

The Ashgate Companion to the History of Textile Workers investment. As a consequence, the textile sector remained much more important in Bohemia and Moravia, where it continued old traditions. Slovakia served as an internal market. At the beginning Czech textiles had no problem in finding new export markets. Since the level of industrial development was very low in the majority of the COMECON member states, when these joined the Soviet project, they were very interested in importing goods from Czechoslovakia until they had developed their own textile capacities. However, these COMECON obligations did not markedly change the geographical export patterns, which still very much resembled the pre-Communist period, when the markets absorbing Czech products were also chiefly located in Eastern Europe. Czech textiles were highly diversified, manufacturing all kinds of materials, including synthetic fibres, and all types of cloth. Their centres of textile research and innovation placed them among the world’s leading textile industries. However, despite this good international reputation, Czech textiles had neither a tradition nor a realistic chance to succeed in competitive Western markets. Unlike these, COMECON demand did not promote growing productivity, innovative designs and increasing quality. When other COMECON producers had established textile capacities in the 1960s and 1970s, Czech imports had to compete with import substitution, which found expression in a diminishing share of textiles in the industrial sector.48 Still, because of their variety and quality, textiles and textile products remained an important commodity, occupying 11.5 per cent of the industrial labour force in 1990.49 Only when communism collapsed in 1989 and COMECON was dissolved in 1991 did Czech textiles have to adapt to a new situation.

Austrian textile industry until the new international division of labour (1918–1980) The Republic of Austria was not the result of a striving for national independence. Founded in November 1918, it included those territories of the former Monarchy that were not claimed by other successor states. Irrespective of political boundaries, the predominantly German population of the new Austria proclaimed itself to be part of Germany (Deutschösterreich) and wanted the German-speaking provinces of Bohemia and Moravia, of Hungary and Italy to join Deutschösterreich. However, the Peace Conference in 1919 only recognized the Austrian claim for western Hungary, while Italy and Czechoslovakia received southern Tyrol and German Bohemia and Moravia respectively. The international community would not allow a union with Germany. However, Austria was not only an unwanted remnant of 48 WIIW (Vienna Institute for Comparative Economic Studies), Comecon Data 1989 (Vienna, 1989). 49 WIIW, Handbook of Statistics. Countries in Transition 2003 (Vienna, 2003). 64

Austria and Czechoslovakia the Habsburg Monarchy at the time of its foundation; it was also cut off from its former economic partners. The split up of the Monarchy and the proclamation of the successor states that precipitated the national integration of their economies broke up the existing regional division of labour. Austria was confronted with the major task of overcoming this transformation, which apart from economic also had social and political implications. Thus Vienna, a capital administering an empire of 52 million inhabitants, was left with less than 8 million people. Furthermore, the luxury sectors of the textile and garment industry, concentrated in Vienna, had lost their primary customers, the court and the nobility. The textile industry serves as a good example to illustrate the difficult Austrian situation. While the majority of its capacities were inherited by Czechoslovakia, only negligible quantities of the textile industry were located in Hungary, Poland, Romania and Yugoslavia. Southern Lower Austria had been the centre of huge spinning and printing units, while weaving factories were located in Bohemia and Moravia. While the sheer size and versatility of the Czech textile industry allowed it to counter these regional imbalances and thus to promote the transformation into a national economy, the disruption of the interregional division of labour caused the most severe regional imbalances of textile production in Austria. It was thus also for economic reasons that Austria laid claim to the Germanspeaking regions of the Bohemian borderlands. Their textile capacities were closely interwoven with the Austrian industry. Losing them, Austria had to either import or replace the lost capacities. However, in spite of its attempts to sever economic relations with Austria, Czechoslovakia remained an important trading partner, accounting for the biggest share of Austrian exports (until 1927, then second after Germany) and the second largest (after Germany) of Austrian imports.50 At the same time the mutual imposition of duties on imports granted space for domestic developments. Austrian producers established weaving capacities in traditional textile regions, namely in Lower and Upper Austria and in Vorarlberg, and they expanded the weaving branch in the Waldviertel. German-speaking Bohemian textile producers, who were afraid of losing access to the Austrian market, also invested in these regions, either by moving their production sites to Austria or by establishing local subsidiaries, usually located near the border. Vorarlberg was traditionally incorporated in a textile region centred in eastern Switzerland and south-western Germany. In the eighteenth century the small province served as a proto-industrial district for Swiss putting-out capitalists. However, at the turn of the nineteenth century Austrian protectionist policies forbade domestic workshops to carry out finishing processes for Swiss companies, which, in turn, encouraged entrepreneurs to turn Vorarlberg into an industrial centre for printing.51 When the first spinning mills were established, Vorarlberg 50 ÖSTAT (Österreichisches Statistisches Zentralamt), Beiträge zur österreichischen Statistik, 1, Der Außenhandel Österreichs in der Zeit zwischen den beiden Weltkriegen (Vienna, 1946), p. 34. 51 Andrea Komlosy and Hubert Weitensfelder, ‘Regionen vergleichen. Am Beispiel Vorarlbergs und des Oberen Waldviertels im 18. und 19. Jahrhundert’, Geschichte und 65

The Ashgate Companion to the History of Textile Workers was under Napoleonic rule, thus profiting from the Continental System and the close relationship with Switzerland. Vorarlberg was returned to Habsburg Austria in 1814 and joined the Habsburg market area in 1825/26. Therefore, Swiss capital invested in Vorarlberg in order to gain access to a market including the province of Lombardo-Venetia, which joined the Habsburg single market in the same year.52 Thus traditionally being part of the Lake Constance region, Vorarlberg was never interlinked with the textile districts in the eastern parts of Austria. After the foundation of the Republic of Austria in 1918, the provincial government, backed by a referendum, petitioned for integration into Switzerland. For various reasons its demand was rejected. However, this decision did not damage Vorarlberg textiles. On the contrary, benefiting from the loss of the Bohemian Lands, they improved their position in the Austrian internal market. At the same time, they maintained close ties with Switzerland and the west of Europe which, in combination with the more diversified structure of their trade and production, rendered their companies less vulnerable to recessions. In eastern Austria the textile industry depended upon East European markets where it now faced fierce competition from Czech companies. While the Czechs proved to be better equipped for this new situation, 31 per cent of Austrian exports in 1930 still went to Hungary, Poland, Czechoslovakia and Yugoslavia, with the textile industry producing a major share of these goods.53 Austria was also confronted with new competitors from the European south-east, where a domestic textile industry had been established. When it had finally recovered in 1924, the post-war Austrian economy only saw a short period of stability which came to an abrupt end with the economic crash in 1929, which had a catastrophic impact on profits and employment. The textile industry was swept by bankruptcies, while the companies that survived the severe blow had to reduce their labour forces significantly. The rates of unemployment surged in all textile areas. Yet, they were particularly high in those regions that were more or less dependent on one single branch. At the same time, the government cut social benefits and reduced the claiming periods. In this dire situation many families were saved by their small plots of land, which were common in many rural textile regions (for example in the Waldviertel and the Mühlviertel). Between March 1938 and April 1945 Austria was finally united with the Czech borderlands when both were annexed by Germany. German industrial planning resulted in the Austrian economy turning from eastern and south-eastern Europe towards the (German) west.54 The ties among the successor states of the Habsburg Monarchy were subordinated to German capital interests, which took over command and ownership of large parts of the Austrian industry. Equally, the production of Region 6 (1997): 234. 52 Ibid., p. 210. 53 ÖSTAT, Beiträge zur österreichischen Statistik, 1, p. 34. 54 Hans Kernbauer and Fritz Weber, ‘Österreichs Wirtschaft 1938–1945’, in Emmerich Talos, Ernst Hanisch and Wolfgang Neugebauer (eds), NS-Herrschaft in Österreich 1938– 1945 (Vienna, 1988), pp. 49–68. 66

Austria and Czechoslovakia consumer goods was replaced with basic (steel, aluminium) and military industries. The latter in particular was regularly modernized and expanded quickly. Textiles became a marginal sector that was of no interest to German capitalists. Still, in the last years of the Second World War textile factories served military purposes, especially in rural regions where the risk of bombing was lower than in towns and industrial centres. Both the Allied forces occupying Austria until 1955 and Austria’s participation in the West European recovery project, initiated by the Marshall Plan, imposed major limitations on the post-1945 reconstruction of the textile sector.55 First and foremost, the post-war recovery programme aimed at the reconstruction of the Austrian economy as it was shaped by the Nazi war machine. The investments therefore targeted basic and heavy industries at the expense of light industries. Furthermore, they aimed to link Austria with the European West, that is the countries taking part in the Marshall Plan. The Marshall Plan funds were designated for the reconstruction of those basic industry sectors that supplied OEEC markets, above all Western Germany, with essential semi-finished products. Consumer industries did not benefit from such support at all. The neglect of textiles entailed that former exchanges, which had been set up among the East European successor states of the Habsburg Monarchy, were discontinued. The European Recovery Programme thus initiated a second wave of Westernization. The reconstruction of the textile sector was further exacerbated by the fact that – with the exception of the Vorarlberg Rhine valley – practically all Austrian textile regions were situated within the Soviet occupation zone, which was excluded from Marshall Plan grants for political reasons. Moreover, the Treaty of Potsdam, signed in August 1945, determined that the Soviets were allowed to expropriate the so-called ‘German property’, by way of compensation for war destruction, while all other firms remained in private property. Textiles were rarely ‘Sovietized’. Private textile producers in eastern Austria were left to their own devices. Because of the lack of capital and credits, most enterprises were small in size, output and labour force. They used outdated equipment and often employed a number of homeworkers. Some bigger owners held back investment in order to open subsidiaries in the Western occupation zones. They thus reinforced an industrial structure characterized by small-scale enterprises and high elasticity of employment and produced mainly for local and regional markets.56 The census of 1951 shows that 17.5 per cent of the people employed in industry in Lower Austria were textile workers, another 11.5 per cent worked in the garment industry; in textile districts this percentage was as high as 50 per cent.57 After the withdrawal of the occupation forces in 1955, public support concentrated on those industries taken over from Soviet administration, mainly branches of the metal, machine, oil and chemical trades. The consumer industries only caught up in the 1960s when mass consumption rose. Starting in the urban 55 Hannes Hofbauer, Westwärts. Österreichs Wirtschaft im Wiederaufbau (Vienna, 1992). 56 Komlosy, An den Rand gedrängt, pp. 220f. 57 ÖSTAT, Volkszählung 1951. 67

The Ashgate Companion to the History of Textile Workers centres, entrepreneurs founded new factories and invested in existing textile producing facilities. This process quickly spread into rural regions, with a first wave targeting traditional textile areas such as southern Lower Austria, the Waldviertel and the Mühlviertel, and a second wave reaching more distant rural areas without any industrial traditions.58 The availability of cheap land, buildings and labour force in the latter regions attracted not only textile companies but also investors in garments and electronics. Yet, the ensuing process of industrialization differed markedly from earlier periods. Unlike the old textile enterprises, the new establishments were subsidiaries of multi-regional or multinational companies whose headquarters were located outside the region. They were bigger in size and workforce, which was often composed of low-skilled and very young, mostly female workers, who accepted low wages. While an increasing number of people started to work for these industrial branches, the small textile producers were in decline and had to reduce their workforce. Only Vorarlberg did not conform to this pattern. The process of industrialization in Vorarlberg mainly drew on local capital. Since domestic workers went into more attractive sectors in Switzerland or at home, the textile companies in Vorarlberg employed a high number of foreign workers, amounting to 30 per cent of the textile workforce in the early 1970s, and they established subsidiaries in the eastern parts of Austria themselves.59 The textile and garment sectors saw high rates of growth in the 1960s and 1970s and still employed an important share of the Austrian industrial labour force (1970: 11 per cent in the textile and 7.5 per cent in the garment industry in Lower Austria). However, they continuously lost ground to other industrial branches as well as to the service sector, which was to become the branch of the future. The world economic crisis of 1973, which marked the end of the economic cycle fuelled by post-war reconstruction, triggered the great decline of the Austrian textile industry. When growth margins decreased and competition grew, capital developed new strategies to lower costs. The entrepreneurs rationalized the work processes and moved labour-intensive branches to Third World countries. As a consequence textile and garment production decreased rapidly in all of Western Europe. This decline had a major impact on the peripheral rural regions where these industries were concentrated. While the industrial centres replaced the declining branches with new sectors of growth, the rural regions that depended upon a limited number of industrial branches faced de-industrialization.60 As early as 1974 the first textile companies moved production from the peripheral Austrian regions to the global peripheries. Both Third World and Eastern European countries, namely Hungary and Romania, began to take over the labour58 Andrea Komlosy, ‘Die niederösterreichische Wirtschaft in der Zweiten Republik’, in Michael Dippelreiter (ed.), Niederösterreich. Land im Herzen – Land an der Grenze (Vienna, 2000), pp. 296–300. 59 Stephan H. Lindner, ‘Der lange Abschied vom ‘Textilland’ Vorarlberg’, Alemannia Studens 7 (1997): 55–87, at pp. 74f. 60 Komlosy, ‘Die niederösterreichische Wirtschaft’, p. 301. 68

Austria and Czechoslovakia intensive steps of textile production for Western multinational companies. One early prominent example was the spinning and knitting company Patria-Piering. Established in the Waldviertel at the end of the nineteenth century, it founded several subsidiaries in this region in the 1960s. Yet, in 1974 a first manufacturing line for pants was transferred to Romania, dismissing 200 local workers.61 Austrian locations were not closed down immediately. For the next 15 years the Austrian government subsidized attempts at restructuring, such as mergers with other companies, and the communities offered financial incentives to save old businesses or to encourage new enterprises. In spite of these desperate efforts, most of the companies eventually left the Austrian periphery. Thus Patria-Piering shut down a knitting branch in 1982. Its spinning mill was included in the so-called Textillösung Ost, a joint regional rescue programme which aimed to coordinate the restructuring of several big textile firms in eastern Austria. When this project failed in 1978, the Patria spinning plant changed owners several times until it was bought by an Indian-based world producer of textile machines in 1984, who received high government subsidies to modernize the equipment, before closing down the site in 1987. The transfer of labour-intensive industries to cheaper locations in the south or the European East is usually traced back to the world economic crisis. However, these interpretations do not take into account that it was not until the 1980s and 1990s that entrepreneurs closed down their companies in the Austrian textile regions. Only some of them managed to survive by specializing in market niches. In 1981 textiles accounted for 5.5 per cent and garments for 6.5 per cent of industrial employment in Lower Austria.62 There are obvious parallels between the crises of 1873 and 1973. Putting an abrupt end to a cycle of expansive growth, both initiated a period of restructuring, which involved the search for strategies to lower the costs of production and led to new geographical arrangements and a new interregional division of labour. The crisis of 1873 precipitated the mechanization of the weaving process and the foundation of weaving mills in rural textile regions. The crisis of 1973 again accelerated rationalization and diminished the demand for labour. At the same time multinational companies founded new production sites in global peripheries, introducing a new division of labour on a world scale. In the textile and – even more so – in the garment industry, the global peripheries replaced the rural regions. These had served as cheap locations for production since proto-industrial times and were temporarily revived in the periods after 1873 and after the economic restructuring in the 1920s as well as in the years of economic expansion in the 1960s when big multinational companies used them for branch establishments. However, from 1980 onwards the companies started to locate their production sites in global peripheries, leaving the rural textile regions in eastern Austria without prospects. While Vorarlberg was equally affected by this change, it was more successful in

61 Komlosy, An den Rand gedrängt, p. 236. 62 ÖSTAT, Volkszählung 1981. 69

The Ashgate Companion to the History of Textile Workers finding market niches and in replacing wholesale commodities with specialized articles for export.63 While Austria was a main industrial producer of textiles between 1873 and 1973/80, labour organizations were substantially involved in shaping working conditions, first on the factory level, subsequently on the branch level and finally by Trade Union Federations which acted nationwide. Major social achievements were legally admitted in the 1880s and the 1920s. After the Second World War unions were co-opted into a so-called ‘Social Partnership’ (Sozialpartnerschaft) with the employers’ organization, exercising far-reaching legislative and executive power in controlling wages, prices and employment. Hence Austria adopted a class consensus with high social standards; it became famous for its social peace and the absence of industrial action. The corporatist management of social affairs was carried out without active workers’ participation. In spite, however, of their common interests, textile unions and the federation of industries were not able to prevent outsourcing, global relocation and the sharp decline of domestic textile production.

Central European textiles facing globalization The collapse of Communism and the capitalist transformation of Czechoslovakia opened new possibilities for East–West cooperation. Attempting to overcome their structural crisis, Western textile companies welcomed the free access to East European markets and labour forces from 1989 onwards. Czech and Slovak industry, whether state-run or privatized, was poised to work for Western companies. While Slovakia, which became an independent state in 1993, played a minor role in the textile and garment sector, Western entrepreneurs often used Czech textile and garment facilities for their production, either by subcontracting specific tasks or by taking over whole companies. Production costs in the Czech Republic were much lower than in Austria, and at the same time the workers were regarded highly for their skills and their discipline. So did the outsourcing of industrial production to the Czech Republic rescue the Czech textile and clothing sector? After 1989 Czech textile companies were confronted with the capitalist world market. When COMECON was abolished in 1991, they lost their protected markets in Eastern Europe, where they faced competition from Turkish or East Asian products, while Western markets were closed off by protectionist barriers. Many enterprises did not survive the complicated process of restructuring and privatization. However, others succeeded by serving as cheap extended workbenches for Western labels. After the fall of the Berlin Wall, many small and medium textile producers of the Waldviertel, which borders on the southern Czech Republic, seized the opportunity to move some of their production lines to a cheaper location with a view to 63 Lindner, ‘Der lange Abschied’, pp. 82f. 70

Austria and Czechoslovakia stabilizing their Austrian firms. Local producers hoped to profit from the unequal international division of labour on a regional scale, for the Czech Republic offered the advantages which until then could only be found in Hong Kong, Singapore or in free production zones in South America and in the Philippines. Furthermore, several multinational companies transferred their textile production from Austria to the Czech Republic. Thus Steilmann, a worldwide producer of ready-to-wear clothes abandoned its sites in the Waldviertel, which were only established in 1979. ‘Hongkong is on our doorstep’; with these words a manager of a multinational firm, which had just moved part of its production from the Waldviertel to the Czech Republic, expressed his feelings about the new situation in 1994.64 As there are no major multinational companies with their headquarters in Austria, the country did not play a leading role in transforming Eastern Europe into a cheap production site for the West. The majority of transactions were undertaken by companies from the European Union (EU), which signed a business agreement with a number of Central European countries in 1992. Exempting semi-finished products of Central European origin from import duties in the European Union, this deal opened the textile and garment sector of these countries to outward processing. When Austria became a member of the European Union in 1995, local companies also profited from this arrangement. While the improved access to eastern labour strengthened the market position of EU textile companies, the dislocation of production sites, whether to a close or a distant ‘Hongkong’, was a severe threat to the jobs of EU textile workers. Czech textile and clothing companies’ acquiescence in the role of dislocated production sites was a clear sign of their economic dependency. Outward processing was their only way to enter Western markets. Nevertheless, Czech textiles have steadily declined since 1989. On the one hand, this is due to the fact that Czech sites competed with other East European and Third World locations where the production costs were even lower. For many companies the Czech Republic was one of the many transitional pieds-à-terre on their continuous search for the cheapest arrangement. When living costs and wages rose, the Czech Republic started to lose its relative advantage to other countries. On the other hand, the practice of transferring production to Eastern Europe was adopted by other branches attracted by the Czech combination of lower wages and high skills, such as the automobile or the machine industry. From the 1990s onwards, leading Western producers in these sectors have invested large sums in the Czech industry and have taken over major Czech companies.65 The decline of the textile industry was mirrored in a decreasing number of workers employed in this sector. While in 1990 the textile and garment industry still accounted for 11.5 per cent of employment in manufacturing industry, this number fell to 9 per cent in 2002. Nevertheless, compared to the textile sectors in Austria, Czech textiles employ a relatively high share of workers, which proves that 64 Komlosy, ‘Die niederösterreichische Wirtschaft’, p. 337. 65 Hannes Hofbauer, Osterweiterung. Vom Drang nach Osten zur peripheren EU-Integration (Vienna, 2003), pp. 191f. 71

The Ashgate Companion to the History of Textile Workers the dislocation of labour-intensive industries had a positive effect on employment figures. It is interesting to note, however, that employment in machinery faced an even bigger decline, from 22.7 per cent to 12 per cent.66 The biggest decline took place between 1990 (22.7 per cent) and 1995 (15.6 per cent). These numbers reflect the process of de-industrialization, which immediately followed the change of system, while foreign investment, which took place in a second phase, was very selective and went hand in hand with closures and rationalization. The Austrian textile industry underwent a rapid decline in the 1990s, which caused severe problems in the rural textile regions. In 1991 the textile sector accounted for 5.8 per cent of industrial employment, the garment sector for 4.2 per cent. Of the 31,380 people employed in textiles, 41 per cent worked in Vorarlberg, 21 per cent in Lower Austria. The garment sector employed 22,790 workers, 20.5 per cent of which were located in Lower Austria, 18.5 per cent in Upper Austria and 15 per cent in Vienna.67 From 1991 to 2003, the number of jobs in these industries constantly decreased, from 31,380 to 18,900 employees in textiles (−39.8 per cent). In the clothes sector the reduction was from 22,790 to 9,600 (−57.9 per cent).68 Now Vorarlberg was also affected and many well-reputed textile firms had to close.69 As a result of this declining significance, the Textile Workers Union (Gewerkschaft Textil Bekleidung Leder) gave up as an autonomous organization in 2000 and merged with the Union of Metal Workers. There are some success stories of companies that managed to maintain their market position or to find a specialized niche. However, the era of mass production is over, leaving behind thousands of jobless textile workers. The decline of textiles coincided with a general crisis in industrial mass production, which was the foundation of post-war growth. While the regions with a more balanced structure of employment were able to counter the effects of rationalization and the new international division of labour, the industrial crisis had its most severe impact on those districts that specialized in textiles. Textile workers, who earned very low wages and had almost no qualifications and no chances for advancement, went from bad to worse when the era of textile production ended. Textile and garment production employed a high percentage of women, hence unemployment has a female face. Many female former rural textile workers have as yet been unable to find a new stable job. Tied to the houses they built in textile villages, they often depend on temporary contracts or return to hearth and home. Many of them try to make up for their income losses by growing their own food or by carrying out tele-homework, thus adopting their ancestors’ strategies to combine paid and unpaid labour, periods of domestic employment with periods of migration. Like the labour force, former textile buildings have lost their specific 66 WIIW, Handbook of Statistics 2003. 67 Österreichisches Institut für Raumplanung (ÖIR), Entwicklungsperspektiven der Textilund Bekleidungsbranche im nördlichen Niederösterreich (Vienna, 1993), pp. 7–9. 68 Industrial Association of Austrian Textil Producers (Vienna), Industrial Association of Austrian Garment Producers (Vienna), Annual Reports (Vienna, 1991–2003). 69 Lindner, ‘Der lange Abschied’, p. 85. 72

Austria and Czechoslovakia function. Ruins evoke memories of an industry that has moved to other places in the world. At the same time, textile museums and trails,70 which were set up in many textile regions, try to conserve the heritage of textiles and reflect on the social and economic transformation.

70 Andrea Komlosy, Waldviertler Textilstraße. Reisen durch Geschichte und Gegenwart einer Region (Vienna, 1994). 73

This page has been left blank intentionally

4 Brazil: the origin of the textile industry Roberta Marx Delson

This chapter considers the origin of Brazil’s important modern textile industry. Rejecting previous theories of textile manufacture, the position presented here is revisionist in nature, suggesting a direct, if uneven, line of development for the industry from the eighteenth century onwards. The 1700s witnessed metropolitan support for an expanding proto-industrialized cotton cloth industry in Brazil in conjunction with an expansion of textile production in Portugal, a significant departure from mercantilist colonial paradigms. Where cloth production had previously been limited to home or plantation settings, heavily capitalized factories now appeared, especially in the north and the far west of Brazil. These mills produced surplus cloth for trade, employing a wage labour force that was mostly indigenous. Cloth was produced in both urban and rural settings, becoming a proxy for currency; salaries were calculated in rolls of fabric, while cloth became a medium of commercial exchange in many parts of the colony. Notwithstanding eventual metropolitan directives to halt this colonial production, cotton cloth of all categories continued to be manufactured in Brazil. The nineteenth century saw the expansion of Brazilian cotton mills which prospered even in the face of competition from inexpensive British textile imports. The diverse nature of the Brazilian textile workforce, which included slave and free labourers, and male as well as female weavers, is also considered. It is widely acknowledged that ‘the textile industry has a major and longstanding role in Brazil’s industrial geography’. Estimates of Brazilian trade in clothing and textiles for the year 2000 indicate an income of US  $22 billion, representing one of the strongest growth sectors of the present economy. 

The author thanks Professor Sharon Sundue (History Department, Drew University) for her useful comparative comments on the early textile industry in the United States and Professor Eric Delson (American Museum of Natural History) for his editorial assistance.  John Dickenson, Brazil: Studies in Industrial Geography (Folkestone, 1978), p. 77.  Anne Sobotta, ‘Survey of the Brazilian Clothing and Fashion Industry’ (London, 2003)

Map 4.1

Brazil

Brazil: the origin of the textile industry But while the importance of textile manufacture in the recent past is undeniable, the origins of this industry in Brazil have largely been obscure. The very notion that colonial and early nineteenth-century Brazil had a manufacturing past contradicts long-standing historical constructions predicated on cash crop economies dominated by exported sugar, cacao, tobacco, indigo and to a lesser degree raw cotton. With specific reference to the manufacture of cloth three arguments have been advanced: the first eschews the existence of Brazilian textile manufacture, or any other manufacture for that matter, prior to 1840; the second suggests that although textiles were woven in colonial Brazil, such activity was halted with the implementation of official prohibitions in 1785; the third position reckons that textiles continued to be produced post-1785, but that cloth manufacture ultimately fell victim to competition offered by inexpensive British goods after 1830 – this view further holds that textile production was not revived until passage of protectionist legislation in the middle of the nineteenth century. Perhaps more telling, these arguments tend to characterize all early manufacture of cloth in Brazil as essentially pre-industrial, dismissing the textiles produced in the colony as artisan or ‘craft’. That categorization, however, is no more satisfactory in describing the economic past of Brazil than are those narrow historicist constructions which have overemphasized the export cash crop sector. My colleague John Dickenson and I have challenged these assumptions in a general survey of proto-industrialization in colonial Brazil. The present discussion focuses on the background of modern textile production in Brazil, and rejects all three of these models cited above for being essentially counter-intuitive. Colonial Brazil possessed many of the requisite factors for industrialized cloth manufacture: these included abundant raw material, of which as cited at http://www.brasil-consult.com/B_R.html. This amount seems high and might represent an error of an order of magnitude, but it is the only numeric estimate I could locate.  See Luiz Carlos Soares, ‘A Manufatura na Sociedade Escravista: O surto manufatureiro no Rio de Janeiro e nas suas Circunvizinhanças (1840–1870)’, in Frédéric Mauro (ed.), La Préindustrialisation du Brésil: Essais sur une Économie en Transition, 1830/50–1930/50 (Paris, 1984), pp. 13–49.  For example, Dauril Alden, Royal Government in Colonial Brazil: With Special Reference to the Administration of the Marquis of Lavradio, Viceroy, 1769–1779 (Berkeley, 1968), p. 385 and Stanley Stein, The Brazilian Cotton Manufacture: Textile Enterprise in an Underdeveloped Area (Cambridge, Mass., 1957), p. 2.  This is the position of Douglas Cole Libby, ‘Proto-industrialisation in a Slave Society: The Case of Minas Gerais’, Journal of Latin American Studies 23 (1991): 1–36.  For the clearest statement of this position, see Caio Prado, Jr., The Colonial Background of Modern Brazil (Berkeley, 1969), p. 260, who characterizes most colonial cloth production as craft.  John Dickenson and I have rejected the characterization of mining and textile production in late colonial Brazil as craft. See John Dickenson and Roberta Delson, Enterprise under Colonialism: A Study of Pioneer Industrialization in Brazil, 1700–1830 (Working Paper 12, Liverpool, 1991). 77

The Ashgate Companion to the History of Textile Workers native, long-staple cotton was only one fibre (wool and linen would be added after colonization); a vast available labour force composed largely of African slaves or coerced indigenous workers; the potential of a substantial domestic market with the sanctioned development of internal trade networks; and the encouragement, albeit sometimes wavering, of the metropolitan government in the eighteenth century. This contribution addresses what logic can merely hint at, that is the coalescence of all these factors in the eighteenth century, resulting in the proto-industrialization of cotton cloth manufacture using Brazilian raw materials, as well as local workers to supply local markets. It is the contention of this discussion that the protoindustrialization of the 1700s is the origin of the present textile industry in Brazil. Nevertheless, while this textile industry has continued to expand from that period until the present, it is also clear that its evolution was neither completely linear nor even. For example, through most of the eighteenth century textile production was characterized by the coexistence of two separate configurations, home/estate manufacture and factory production; only the latter can be considered the root of today’s modern industry. Additionally, there were decided obstacles which proto-industrialized textile manufacture had to overcome in the course of its development. Among these was competition between the mother country and her colony for textile production supremacy and shifting paradigms of mercantilism; the late eighteenth-century growth of a substantial trade in raw Brazilian cotton to England, whose mills represented the principal competition to both the Portuguese and the Brazilian textile industries; and the Álvara (Law) of 1785, alluded to above, which ostensibly banned Brazilian textile production, whether industrialized or not, except for coarse goods. For the nineteenth century there is the question of whether the Brazilian textile-producing capacity could continue to grow given the entrance of cheap British fabrics into the country. All of the above will be considered here as well as the diverse nature of the Brazilian textile workforce which encompassed not only coerced labourers but also freemen, and male textile workers as well as the traditional female spinners and weavers.

Pre-colonial period Textiles were initially produced in Brazil by indigenous tribes using widely available cotton and other fibres. Unfortunately, absent any written data relating to this production, historical interpretations have taken their cue from the salacious early accounts of Portuguese visitors to Brazil, which focused on the nakedness of the first peoples encountered. These commentaries ignored the fact that many Brazilian Indian tribes (especially in the interior) wore some kind of woven covering, ranging from tangas (small skirts) to full tunics with leggings.10 The obvious skill of  For example, the account of Pero Vaz de Caminha, cited in John Hemmings, Red Gold: The Conquest of the Brazilian Indians (Cambridge, Mass., 1978), p. 3. 10 The Omagua, known as the Camemba in Portuguese, were known to have worn tunics 78

Brazil: the origin of the textile industry autochthonous weavers was evident in their elegant redes (hammocks), baskets and ropes manufactured from vines and other fibres. Weaving techniques were, in point of fact, so important ethnographically that, even among peoples of apparently the same tribe, differences in the final twist direction of cordage served to delineate the identity of internal groups within the larger ethnic group.11

The sixteenth and seventeenth centuries Coincident with the beginning of direct colonization of Brazil (around mid-sixteenth century), Portuguese settlers began to spin and weave cloth in their homes. The output of this labour was impressively diversified even though it was confined chiefly to urban and rural domestic settings. In the region of São Paulo, for example, everything from canvas to fine fabrics and trimmings was produced.12 Most textile manufacture, however, was limited to lesser (and coarser) fabrics, from which were fashioned the simple wool and cotton garments worn by the common classes.13 Local production satisfied this market, suggesting a situation in Brazil analogous to Portugal where coarse textiles for the poor were also woven by peasants in home settings.14 We can surmise that production of these ‘subsistence textiles’ during the period 1550–1700 proved to be economically supportable, since evidence of their manufacture exists from almost every region of the vast Brazilian colony. Additionally, cloth was produced on fazendas (principally sugar plantations), where slaves served as both spinners and weavers.15 The goods which they turned out were sewn into the basic clothing worn by field hands. Estates also produced coarse sacking materials used for packaging the sugar and tobacco that was destined for the export trade. We may view these estates, therefore, as providing a type of self-contained, non-industrialized, rural workshop setting for textile production, in which the output was intended for internal consumption rather than trade, or for collateral usage as shipping materials. While house slaves dressed in a more sophisticated manner than lowly field hands they likely wore locally produced fabrics, as well. Ferreira Lima reports that fashioned from cultivated cotton. See William M. Denevan, Cultivated Landscapes of Native Amazonia and the Andes (Oxford, 2002), p. 63. 11 As cited in C. Jill Minar, ‘Motor Skills and the Learning Process: The Conservation of Cordage Final Twist Direction in Communities of Practice’, Journal of Anthropological Research 57 (2001): 381–405, at p. 391. 12 Ernesto Silva Bruno, História do Brasil: geral e regional (6 vols, São Paulo, 1966), vol. 5, p. 42. 13 Ibid., p. 46. 14 Although some fabrics such as linen and silk were produced in small factories. On this point, see Nuno Luís Madurerira (org.), História do Trabalho e das Ocupações (3 vols, Oeiras, 2001), vol. 3 (A Indústria Têxtil), p. 78. 15 H. Schlesinger, Geografia industrial do Brasil (São Paulo, 1966), p. 178. 79

The Ashgate Companion to the History of Textile Workers clothing sewn from local materials was considerably cheaper than that fashioned from imported cloth.16 Indeed, the more refined (and imported) fabrics purchased by the well-to-do were deemed so valuable that fine clothing items were frequently mentioned in inventories, or left in wills to either relatives or the occasional lucky slave.17 From the middle of the sixteenth century to the end of the seventeenth, therefore, the necessity of clothing the large number of African and Creole-born slaves, as well as dressing the poor free population, created and sustained the need for cheap textiles in Brazil which was not satisfied (and probably could not realistically have been met) by Portuguese production. Throughout this period, textiles were created in what were essentially ‘craft’ operations, even while the wealthy in Brazil continued to favour an imported, and perhaps more industrialized, product.

The eighteenth century If the early years of the textile industry in Brazil were characterized by small scale home/fazenda production geared towards internal consumption then, by comparison, the eighteenth century represents a watershed period of development. By the 1700s dedicated textile factories appeared in newly established centres in the north and far west of Brazil, as well as in older coastal centres. A trade in textiles within Brazil now emerged, in conjunction with a parallel trade abroad. The reorganization of Brazilian textile production at this point was a direct result of changes in the Portuguese textile industry at home. Prior to the eighteenth century, Portugal had engaged in three separate phases of the textile industry, as producer, importer and trader of cloth: textiles from Guiné (Africa), Indian fabrics, Portuguese red cloth and French woven goods, for example, all found their way to the Cape Verde islands which functioned as a veritable entrepôt for the Portuguese textile trade.18 However, a dramatic shift in this pattern occurred in 1703, when the Portuguese and English signed the Treaty of Methuen allowing (mostly woollen) British textiles to enter Portugal unrestricted, in return for the entry of Portuguese wines on a tariff par with the French.19 This trade would soon prove to be quite lopsided; Portugal’s economic fortunes waxed and waned after the agreement was implemented, with the ensuing expanded importation of English textiles blamed for its weaknesses.20 16 Heitor Ferreira Lima, Formação Industrial do Brasil (Rio de Janeiro, 1961), p. 150. 17 This was especially the case of favourite slaves and illegitimate children. See examples provided in Eduardo França Paiva, Escravos e Libertos Nas Minas Gerais do Século XVIII: Estratégias de Resistência Através dos Testamentos (São Paulo, 1995), p. 96 and ch. 2 et passim. 18 T. Bentley Duncan, Atlantic Islands: Madeira, the Azores and the Cape Verdes in SeventeenthCentury Commerce and Navigation (Chicago, 1972), pp. 215–216. 19 A.H. de Oliveira Marques, History of Portugal, 2nd edn (New York, 1976), p. 386. 20 Ibid. On the same point, see also Judith Blow Williams, British Commercial Policy and Trade Expansion, 1750–1850 (Oxford, 1972), p. 149. 80

Brazil: the origin of the textile industry The obvious solution to what had rapidly emerged as a situation of economic dependency was to establish textile self-sufficiency. Prodded by their new Prime Minister, the Marques do Pombal, mid-century Portuguese leaders rethought the once-promising Methuen Treaty and now aggressively promoted import substitution at home and in Brazil. This new commercial policy was an intentional rejection of contemporaneous mercantilist strategies with their traditionally defined roles of colonial supplier of raw materials and metropolitan manufacturer. Instead, Pombal insisted upon implementing his unique vision of a state economy in which colony and mother country might function as one economic unit, producing the same product, without competing with each other. This new modality led to the coeval establishment of many new textile centres in Brazil and in Portugal, including those that processed silk as well as cotton, wool and linen.21 As it evolved, the textile industry in eighteenth-century Brazil obeyed a type of geographic determinism, ranging from the production of wool and felted fabrics in the colder south,22 to fine fabrics woven to meet the demands of wealthy gold and diamond miners in central Minas Gerais,23 to the simple cottons needed to clothe indigenous populations and provide uniforms for soldiers in the very warm north and far west. Some of this production was no doubt rationalized by the difficulty and expense of transporting imported fabrics from the coastal ports to interior regions. Minas Gerais, for example, capitalized on its distance from Rio de Janeiro to develop an industry in fine woven goods (including cottons and linens) which was so successful that Mineiro cloth was supplied to hinterland regions where logistics precluded meeting consumer demands from Rio.24 This internal trade was duplicated in other regions as well. Concurrently, workshops in the coastal (and, after 1763, capital) city of Rio de Janeiro turned out fine fabrics, including linens and velvets, some of which boasted gold and silver threads, as well as lace.25 That these offerings were quickly absorbed by the local elite is all the more interesting in light of availability of imported merchandise in that port city. Clearly, we may read this as a confirmation of the high quality of the Brazilian product. It was in the north and the far west, however, that textile production most closely followed a proto-industrial model. In these regions cloth manufacture was heavily promoted and subsidized by the metropolitan government and/or by commercial enterprises which had been organized in Lisbon (for example the Company of Grão Pará and Maranhão). The textile output of this configuration was no longer restricted to satisfying immediate needs of the home or fazenda clientele (that is 21 For a discussion of Pombal’s policies and the creation of textile (especially silk) factories in Portugal and in Brazil, see Oliveira Marques, History of Portugal, pp. 383–384 and Kenneth R. Maxwell, Conflicts and Conspiracies: Brazil and Portugal, 1750–1808 (Cambridge, 1973), pp. 41, 51–56. 22 Caio Prado, Jr., The Colonial Background, p. 260. 23 Maria Beatriz Nizza da Silva, Vila Privada E Quotidiano no Brasil na Época de D. Maria I e D. João VI (Lisbon, 1993), p. 234. 24 Dickenson and Delson, Enterprise under Colonialism, p. 40. 25 Silva Bruno, História do Brasil, vol. 4, p. 64. 81

The Ashgate Companion to the History of Textile Workers self-contained) but was, instead, geared towards creating surpluses for trade. Instead of settlers’ wives or slave labourers producing fabrics, this new mode of production employed numerous wage-earning workers. The factories themselves were located in large cities or in newly developed communities established in rural areas after mid-century. In the scope of its production, the level of organization and its dependency on wage labour, this type of factory manufacture represented a quantum leap forward from the home/estate cloth manufacture associated with the ‘subsistence textiles’ of the sixteenth and seventeenth centuries, which also continued in parallel into the eighteenth. Perhaps the most notable example of the new proto-industrialized configuration was the textile factory created c.1756 in the city of Belém do Pará, the northernmost captaincy.26 The goal of this establishment was to supply cotton cloth in the various colours needed for military uniforms. It was considered so vital to Portuguese interests that the local administrator was ordered to solicit and enlist the aid of the best spinners and weavers in the captaincy.27 Unfortunately, there are no data on the actual number of workers successfully recruited for this venture. More broadly interesting, perhaps, is the scheme that the home government proposed to assure success at this new factory. Drawing on resources from a totally different part of the world, Portuguese administrators proposed to contract expert weavers and calico printers from the Coromandel coast of India to oversee manufacture in the new Brazilian establishment.28 This plan showed an extraordinary, and precocious, awareness of ethnicity for the time, as the government decreed that East Indian workers would be allowed to live in a cultural enclave, where they might preserve their own language and religious practices without interference. The dynamics of such a fascinating social/economic experiment assuredly have implications regarding the growing global thinking of the Portuguese; sadly, there are no data which confirm that the programme was ever implemented. Another variant of this new industrialized textile configuration appeared in territories newly claimed from the Spanish after 1750. In these regions, located mainly in the interior Amazon and the far west, the manufacturing of cloth took on portentous overtones, emerging as a linchpin of Pombal’s mission to ‘civilize’ indigenous populations. Seeking to create ‘an enlightened and idealistic attempt at a new Indian policy [which was] coupled with a coherent plan to settle the Amazon Basin’,29 the Portuguese would relocate and assimilate indigenous peoples, many 26 Maxwell, Conflicts and Conspiracies, p. 41. 27 These instructions are contained in Pedro José Corrêa to Bishop of Para, 24 September 1756 (Lisbon) Biblioteca Nacional do Rio de Janeiro, Códice DCCCII-15-72 in Manoel Barata, Formação Histórica do Pará: obras reunidas (Belém, Pará, 1973), pp. 128–129. 28 Weavers from India are mentioned in ibid. and Ferreira Lima, Formação Industrial do Brasil, p. 141 and in C.R. Boxer, The Golden Age of Brazil, 1695–1750: Growing Pains of a Colonial Society (Berkeley, 1962), p. 300, who additionally notes that such weavers were to be allowed to keep ‘their own dress, manners and customs’. 29 Robin L. Anderson, Colonization as Exploitation in the Amazon Rain Forest, 1758–1911 (Gainesville, Fla., 1999), p. 30. These colonizing projects are also discussed in Roberta Marx Delson, Vilas Novas para o Brasil-Colônia: Planejamento Espacial e Social no século 82

Brazil: the origin of the textile industry of whom had only recently been wrested from Jesuit missions. Under an ambitious governing authority known as the Directorate, local administrators were charged with resettling indigenous charges into new communities, where they would be required to grow the cotton needed to produce clothing in the European mode. While the larger Jesuit missions had long produced cloth for the needs of their residents,30 now even the smallest Directorate community was required to develop communal plots where cotton could be cultivated,31 the objective being to increase the absolute volume of total raw material as well as, ultimately, the total cloth output. Many of the new communities, placed under direct control of Portuguese secular administrators, had specific buildings dedicated to the spinning and weaving of cotton cloth, which are referenced on town plans as ‘factories’. The 1784 map of the new town of Albuquerque in Mato Grosso, for example, is shown with such a textile factory occupying its own, distinct building.32 A visitor to the town in 1786 observed the activities of an already prosperous cotton-weaving industry, the products of which were exchanged with traders in the larger regional city of Cuiabá for luxury goods.33 The proto-industrialization of cloth production in the north was successful to the degree that not only was surplus cotton cloth produced for internal trade, and to clothe the military, but fabrics from the captaincy of Pará were apparently exported back to the mother country!34 Simultaneously, cloth had become so valuable, and so intimately linked to the economy of the northern zones (including the northern captaincy of Maranhão), that it served as a proxy for currency, given as wages to soldiers and indigenous workers on the one hand, and utilized as a medium of exchange for white settlers on the other.35 While payment in cloth actually dates from the seventeenth century, by the mid-1700s resident administrators in the Brazilian north were alleging that the practice had resulted in the scarcity of cloth (presumably because some of it had been exported back to Portugal as well as being absorbed into wages) and the precipitation of an inflationary cycle which saw rises in the price of cloth along with a devaluation of currency equivalents.36 In response to the situation, the local government forbade the export of cloth to Portugal after

XVIII (Brasília, 1997). 30 Colin M. MacLachlan, ‘The Indian Labour Structure in the Portuguese Amazon, 1700–1800’, in Dauril Alden (ed.), The Colonial Roots of Modern Brazil (Berkeley, 1973) pp. 199–230; p. 206 refers to the Jesuit cloth manufacture. 31 Anderson, Colonization as Exploitation, p. 61. 32 For a map of Albuquerque, see Delson, Vilas Novas, p. 77. 33 The observation of Albuquerque was made in September, 1786. See ‘Diário da diligencia da comissão chefiado pelo Engenheiro Ricardo Franco, 1785–1786’ (Rio de Janeiro) Arquivo Histórico do Itamaraty, Lata 266, Maço 1, Pasta 21. 34 Ferreira Lima, Formação Industrial do Brasil, p. 140. 35 Stuart P. Schwartz, ‘Colonial Brazil, c. 1850–c. 1750: Plantations and Peripheries’, in Leslie Bethell (ed.), Cambridge History of Latin American (Cambridge, 1984), vol. II, pp. 423–499, at p. 477. 36 Ferreira Lima, Formação Industrial do Brasil, p. 141. 83

The Ashgate Companion to the History of Textile Workers 1755;37 notwithstanding this prohibition, even as late as the turn of the nineteenth century, spun thread continued to be one of the statistically significant products shipped from Brazil to the mother country.38 By the late eighteenth century, therefore, Brazilian textile production showed two distinct, but coexisting, configurations: home/fazenda production for self-contained use, which continued on much as it had in the earlier centuries, and proto-industrial factory production. The metropolitan government had seemingly taken scarce notice of the pre-industrial mode of production. However, in the factories established in the northern and the far western captaincies of the Brazilian colony, the metropolitan government took on a proactive role, supporting the capitalization of a textile industry. In these settings the ability to manufacture enough product to satisfy local needs as well as to distribute surpluses to internal trade networks (and to Portugal), in conjunction with the level of importance attached to the production by the metropolitan government attest to a remarkably different order of development from the pre-industrial order of the previous two centuries. The differences between these two coexisting modalities are further underscored by the variations in technology utilized in each setting. Home/estate production continued to rely on simple spinning distaffs and looms. By contrast, the state-run factories were associated with more sophisticated, and likely imported, machinery. An illustration of a spinning machine in use in the late eighteenth-century Amazon illustrates this point very well (see Figure 4.1): essentially a variation on the Saxony wheel, this impressive mechanism sported two spools (flyers) to collect the spun thread, as well as a side crank and a foot pedal which could be used by a seated operator. This apparatus is analogous to those used in early English factories where poor workers were obliged to spin as much thread as possible.39 The appearance of a similar machine in the Brazilian interior not only suggests equivalent pressure to produce increased amounts of thread by means of more efficient technology (hence, the two flyers) but, less obviously, implies that the Portuguese considered the indigenous textile workers who utilized the wheel as a type of proletariat. One also suspects that such spinning devices were probably either directly imported from England or forwarded from Portugal,40 rather than being manufactured in Brazil.

37 Ibid. 38 José Jobson de Andrade Arruda, ‘O brasil no Comércia Colonial (1796–1808): Contribuição ao Estudo Quantitativo da Economia Colonial’ (Ph.D., University of São Paulo, 1972), p. 516. 39 Apparently the placement of the two flyers above the wheel was intended as a spacesaving device. On this point, see Laurel Thatcher Ulrich, The Age of Homespun: Objects and Stories in the Creation of an American Myth (New York, 2001), p. 91. 40 Oliveira Marques, History of Portugal, pp. 384 reports that the cotton factory in Portiere, Portugal installed modern equipment from England in 1776, while ‘jennies’ and ‘mulejennies’ arrived in Portugal from England c.1770s and continued throughout the century. 84

Brazil: the origin of the textile industry

Figure 4.1 Spinning wheel for cotton, 1784. ‘Roda de tear algodão’, Plate #51, vol. I from Alexandre Rodrigues Ferreira, Viagem Filosófica pela Capitania do Grão Pará, Rio Negro, Mato Grosso e Cuiabá , 2 vols. Facsimile folio (Rio de Janeiro, 1971) (hereafter, VF, I). Note the presence of two flyers to collect the spun thread, as well as the side crank and foot pedal. In yet another contrast with home/fazenda production, where cotton was cleaned by the tedious labour of plucking seeds from the fibres by hand, in the Amazonian factories a variant of a cotton gin had been put into use at least as early as 1784 (see Figure 4.2). Still hand-powered, it increased efficiency by using two heavy rollers to extract seeds from the cotton fibre. We know nothing about the origin of this gin. However, by 1811, the French botanist Auguste de Saint Hilaire observed the use of 12 water-powered cotton gins in just one locale in western Meia Ponte, suggesting that the earlier mechanisms had been superseded by more modern waterpowered models.41 The use of this relatively sophisticated machinery (as well its rapid obsolescence) suggests extensive capital outlay in the northern and western Brazilian textile factories. Such a profile is consistent with proto-industrialization elsewhere in the colony. 41 As cited in Silva Bruno, História do Brasil, vol. 6, p. 59. 85

The Ashgate Companion to the History of Textile Workers

Figure 4.2

Engine for ginning cotton, 1784. Source: ‘Engenho de Descaroçar Algodão’, VF, I, plate #50. Note that this device is illustrated as being hand-powered, and that the ginned seeds are shown. The origins of this device are obscure.

Figure 4.3 Loom used by Indian women in the village of Monte Alegre, 1785. Source: ‘Prospecto do Tear, com que fazem as suas redes mais delicadas as Indias da villa de Monte-Alegre’, VF, I, plate 53. Note the geometric pattern that is being woven. 86

Brazil: the origin of the textile industry On the other hand, not every aspect of textile production was modernized in the factories. The vertical, indigenous loom (see Figure 4.3) remained in wide usage everywhere, from home/plantation settings to factories.42 No other loom descriptions have been found for Brazil, although in Portugal, as noted above, up-to-date English equipment, including horizontal looms, was widely imported by the late eighteenth century. Statistics on the amount of fabric which was spun and woven in the eighteenth century exhibit similar lacunae. Data obtained on 266 looms registered to individuals (most likely settlers) in the southern captaincy of Santa Catarina provide a figure of 42,181 varas [each c.1.1 m] of cotton cloth produced in the year 1755.43 By contrast, a single factory in the vicinity of Rio de Janeiro (Campos de Goitacases), where approximately a hundred looms were in service, produced nearly 50 thousand varas of cloth in the year 1785.44 This suggests the greater efficiency of the factory settings. A cotton mill established in Rio Negro in the Amazon in the 1790s had 18 looms, run by a labour force of 16 men and women, and another 96 workers who prepared the cotton and spun thread in preparation for weaving. This workforce, composed of resettled Indians, produced 407 rolls of cotton cloth in 1797.45 Unfortunately, there is no way to estimate how many varas of cloth constituted one roll, or even if the size of rolls was consistent throughout the colony. Thus, it is impossible to compare the Rio Negro output with that of the Campo de Goitacases factory. For Minas Gerais, Douglas Libby notes that over a thousand looms were listed in an inventory taken after 1785 but cites no data to confirm how much yardage these devices actually produced.46 One is left with the impression of substantial output, but the data needed for confirmation are lacking. A final category which must be considered in this discussion of the colonial origin of Brazil’s industrialized textile manufacture is labour. Throughout the colonial period, even in the home/fazenda configuration, textile makers were an extraordinarily diverse lot by Brazilian standards.47 The standard divisions between the labour of slaves vs freemen, or Portuguese vs indigenes, dissolve when considering the manufacture of cloth. Douglas Libby goes so far as to suggest that the homely tasks of spinning and weaving were social equalizers; he writes that on Mineiro estates the owner’s wife frequently worked alongside her slaves to spin and weave sufficient cloth for their needs.48 We may also imagine that slave women recently arrived from Africa passed on their past weaving skills to other workers,

42 Ferreira Lima, Formação Industrial do Brasil, p. 161. 43 Dauril Alden cites the report on Santa Catarina in Royal Government, fn. 115, p. 385. 44 Silva Bruno, História do Brasil, vol. 4, p. 92. 45 MacLachlan, ‘The Indian Labour Structure’, pp. 219–220. 46 Libby, ‘Proto-industrialisation in a Slave Society’, p. 24 cites an inventory of 1785. 47 There is some evidence that prisoners were employed in the manufacture of cordage from guaxima but we do not know what their ethnic background was. See Alden, Royal Government, p. 371. 48 Libby, ‘Proto-industrialisation in a Slave Society’, p. 28. 87

The Ashgate Companion to the History of Textile Workers perhaps even to their patrons; as noted earlier, the Portuguese had long been involved in trade with areas in Africa which produced expertly woven fabrics. The labour in home settings was also gender specific. It was almost exclusively women who were engaged in the production of textiles; young women (both free and slave) were taught skills at an early age allowing them to work with more experienced spinners and weavers without real hierarchical distinction. The sheer numbers of women engaged in such occupations is staggering. Libby proposes a tentative figure of 30,000 spinners49 needed to supply the 1,000 looms noted in the 1785 inventory (that is a ratio of 30:1), but he insists that these figures represent home rather than factory organization. Interestingly, Libby suggests that a half century later (by 1831–1840) the ratio had been reduced to 24 spinners to one weaver.50 It is likely that women were also involved in exchanging their homespun in what gradually evolved into a substantial local trade. By contrast, in the proto-industrial setting, where the workforce was recruited from indigenous inhabitants of newly settled villages, work floor organization showed completely different gender tendencies. In the factories, women did not prevail in the labour force as they did in settlers’ homes or fazendas, but shared tasks with men (although to focus their attention on work, they do appear to have been separated by gender in different areas of the factory).51 Significantly, such workers were paid wages. In the northern captaincies of Brazil, wages were set by law, and periodically adjusted across the board for the entire region.52 The wages earned, in turn, reflected divisions in labour defined by assigned task; in such settings definite hierarchies of workers emerged. In the Rio Negro factory, for example, spinners were paid for each roll of thread produced. Master spinners, however, automatically earned a set wage, receiving 1,600 reis a month. Since we do not know how many rolls could be produced monthly there is no way to calculate the difference between ordinary spinners and the master spinner who supervised their work, but a clear categorical distinction is suggested from the data. Other workers, such as weavers, were apparently paid a lower set wage (1,200 reis per month), and were not paid by piecework. Of course, none of the indigenous workers ever received actual currency but were paid the equivalent value in bolts of cloth, which could be used in bartering for other goods. Lastly, the distribution of the output seems to have been overseen by government officials, who kept track of profitability; for the Rio Negro factory, one half of the net profits were deposited in the account of the royal treasury in 1797.53 We lack clear knowledge of the ages of these indigenous factory workers; we can only say with certainty that if anyone was unsatisfactory they were sent back to their villages to do other tasks. It is equally difficult to comment with certainty on the ratio of spinners to weavers in the proto-industrialized factories. Nevertheless, 49 Ibid., p. 24. 50 Ibid., p. 28. 51 MacLachlan, ‘The Indian Labour Structure’, p. 219. 52 Ibid., p. 205. 53 Ibid., pp. 219–220. 88

Brazil: the origin of the textile industry available data from the Rio Negro factory, where the 18 looms operated by men and women were supplied thread spun by 96 Indian (gender unspecified) spinners,54 suggest a ratio of 6:1, considerably lower than that cited by Libby for domestic production in Minas Gerais. What this probably indicates is the greater proficiency of these proto-industrialized indigenous spinners who engaged in cloth production full time, as well as the use of more efficient technology. We noted earlier that the use of the two flyer spinning wheel in English poorhouses implied increased pressure on spinners to produce more thread. It was likely that the same situation prevailed in the Amazon, where the government was essentially the owner of the factories. In any event, we can be certain Indian workers put in long hours; notwithstanding the appearance of a relaxed schedule, factory workers in the Rio Negro worked ten and a half hours a day (five until eight in the morning, followed by a half-hour rest, then from eight-thirty until noon, followed by a two-hour rest and then more work until 6 p.m.).55 Certainly the intensity of production in the factory (where work represented full-time employment) would have differed dramatically from the home/fazenda setting where other tasks likely intervened. All of the indigenous workers, including those engaged in textile production, came under the jurisdiction of the Directorate, which required them to be dressed in quasi-European fashion. Commenting on the Portuguese desire for assimilation through dress, the great eighteenth-century Brazilian naturalist Alexandre Rodrigues Ferreira noted that cloth was not only given as wage payment to native workers but also doled out, free of charge, to recently relocated occupants of indigenous communities. He observed that it was commonplace in such villages to single out the chiefs and to present these with more and finer bolts of fabric than the ordinary tribe members. It was reckoned that this ‘gift’ would encourage the chief to wear European-style dress and inspire others to follow his example, even while such practice simultaneously served to re-enforce his authority and, hence, his obligation to his Portuguese patrons.56 Critics of the Directorate system have suggested that the real motivation behind the clothing mandate was to increase profits by creating a captive indigenous consumer market.57 While it is undoubtedly true that expanded numbers of consumers were a consequence of the legislation (and go far in confirming the existence of a proto-industrialized economic modality), this critique misses the irony inherent in transforming native textile factory workers into a rural proletariat. Not only did indigenous workers wear clothing sewn from the fabrics which they, 54 Ibid., pp. 219. 55 Ibid. 56 This point is made in the ‘Relação’ ordering the settlement of the Muras in Airão, 12 March 1787 and cited in Alexandre Rodrigues Ferreira, Memorias da Viagem Filosófica pela Capitania do Grão Pará, Rio Negro, Mato Grosso e Cuiabá (2 vols, Rio de Janeiro, 1974) (hereafter Memorias), vol. 1, p. 160. 57 This argument is advanced in B.J. Barickman, ‘“Tame Indians”, “Wild Heathens”, and Settlers in Southern Bahia in the Late Eighteenth and Early Nineteenth Centuries’, The Americas 51 (1995): 325–368, at pp. 343–344. 89

The Ashgate Companion to the History of Textile Workers themselves, had produced, but they were even paid in locally-produced cloth.58 Additionally, it was likely that the cotton was grown in their villages. The indigenous spinners and weavers, therefore, were being assimilated, even as, simultaneously, they were the instrument via which the assimilative process took place. The clothing directive exhibited multiple functions: it stimulated the increased output of industrialized cloth production which, in turn, hastened assimilation, economic integration, liquidity, as well as expanded markets. This did not automatically result in a totally ‘Europeanized’ product. Designs reflected local cultural traditions (see Figure 4.4). Fashioned into European styles, the result was a product presumably more palatable to local tastes; the reward for the Portuguese was a clothed population.59 Cloth production was also affected by mid-eighteenth century sumptuary laws, which restricted the working classes from wearing upper class finery. Ostensibly the purpose of these measures was to prevent importation of expensive cloth from other European nations; inversely, they also stimulated local demand for simple fabrics. Given sumptuary laws, Indians were precluded from completely emulating European couture. Thus, it made sense for the Portuguese to allow them leeway in designing their own cloth. Occasionally the sumptuary prohibitions took an ironic twist: elite partisans of the 1789 Inconfidência revolt sported locally produced inexpensive cotton clothing as a proud symbol of their resistance to the Crown.60 Did the state-sponsored Brazilian textile industry impact its Portuguese counterpart? While it is difficult to answer this question with certainty we can observe that the initial enthusiasm that engendered the push for import-substitution on both sides of the Atlantic gave way, first, to scepticism and eventually to alarm concerning Brazilian production. Even while Pombal was still at the helm, an early, but unsuccessful, attempt was made in 1766 to curtail Brazilian textile production, which was now perceived as threatening metropolitan factory output. Justifications for this re-evaluation of the colonial economic scene were often obscured in circuitous, quasi-physiocratic arguments: the captain-general of Minas Gerais argued that local production of cloth had reduced the incentive of miners to dig for gold since with Mineiro cottons readily available they would not need ore to pay for more costly Portuguese manufactures.61 Apparently, if the Minas Gerais residents had simply bought sufficient Portuguese textiles, they would not be facing restrictions. 58 Anderson, Colonization as Exploitation, p. 42. 59 Roberta Marx Delson, ‘Beyond Imperial Domination and Resistance: Extrapolating the Late Colonial Amazonian Cultural Landscape from the Visual Record’, Yearbook, Conference of Latin Americanists Geographers, 26 (2000): 117–130, at p. 124. 60 Judith Blow Williams, British Commercial Policy and Trade Expansion, 1750-1850 (Oxford, 1972), p. 149. (For a discussion of the impact of sumptuary laws in the Atlantic Islands see Abel Soares Fernandes, Angela Freitas Alves and Julieta do Vale Fernandes, O trajo na Madeira: subsidio para o seu estudo (Funchal, 1994), pp. 208-218.) Dickenson and Delson, Enterprise under Colonialism, p. 40. 61 Ferreira Lima, Formação Industrial do Brasil, p. 159; Alden, Royal Government in Colonial Brazil, p. 384. 90

Brazil: the origin of the textile industry

Figure 4.4 Cambeba Indian male with his armament. Source: ‘Índio Cambeba com suas Armas’, VF I, Plate #118, n.d. The Cambeba, or Omagua (as they known by the Spanish), Indians were widely known for wearing ‘European-style’ shirts and trousers sewn from their own woven cotton fabrics. The cloth clearly displayed characteristic geometric motifs, as shown in this illustration. The opportunity to enact further prohibitions against the increasing Brazilian cloth manufacture presented itself when Pombal fell from power in 1777. The very same fabric manufacturers whom he had promoted in Portugal now pleaded with the Crown for relief in the form of totally outlawing Brazilian textile production; Kenneth Maxwell argues that this turnabout may be viewed as a retreat to neo-mercantilism.62 Ultimately, the campaign of these favoured manufacturers met with success, resulting in the Álvara of 1785 which sought not only to prohibit textile production but to curtail all colonial manufacture. In regard to cloth production, the Álvara specifically limited activity to the production of coarse textiles (suitable for the clothing of slaves and the very poor) and sacking. Inadvertently, however, the legislation allows us to confirm the extent to which textile production had burgeoned 62 Maxwell, Conflicts and Conspiracies, p. 78–80. 91

The Ashgate Companion to the History of Textile Workers during the colonial period, by singling out for prohibition ‘the manufacture of not only cotton cloth, but silk, wool, linen and embroidery’.63 Far from delivering a crushing blow to Brazilian cloth manufacture, however, the legislation of 1785 appears to have operated more in the breach than in the observance; available data from the post-1785 period suggest that widespread production and trade in textiles continued unabated in most of the colony. In the area of Vitoria, for example, the principal commerce continued to be cotton, traded both in thread and in cloth. In Minas Gerais production seems to have actually sped up after 1785.64 The Rio Negro factory considered above was first established in 1798 (some 13 years after the publication of the prohibition) while in the far west the Crown openly encouraged the production of fine cotton textiles rather than lesser fabrics even after the Álvara was in effect. A clear example of what is an obvious breach of official policy is provided in the 1787 instructions from the captain-general to the Commandant of the Vila of Borba. The Commandant was to set two Indian couples, who had recently migrated out of Spanish territories wearing apparel of finely woven cotton cloth (see Figure 4.5), the task of weaving such cloth for the Portuguese. Apparently, the government was not restricted by its own prohibitions on manufacture, as the Commandant was even promised new looms!65 Yet another example of an obvious violation of official government policy was the textile factory of the new town of Albuquerque. As noted earlier, when this town was observed in 1786 it was producing substantial amounts of cotton cloth for resale in Cuiabá, even as the prohibition had supposedly taken effect. By the same token, the impact on what appears to have been a longer established major textile producing region, Rio de Janeiro, must also be re-evaluated, as only 16 persons in that city appear to have been constrained from producing luxury cloth.66 Yet another variable in this complex portrait of Brazilian textile manufacture was the increase of exported raw Brazilian cotton to England after the mid-eighteenth century.67 Ironically, this trade began at roughly the same time that cloth production was being encouraged in Brazil. The overlap of these two clearly contradictory trends is problematic. Was this yet another method of controlling the expansion of the competitive Brazilian manufacture or was there simply enough raw cotton to play both ends of the market at the same time? Ultimately, it would seem that the possibility of long-distance trade with the British in Brazilian cotton proved to be at least as attractive as allowing that cotton to be absorbed into local 63 We first made this suggestion in Dickenson and Delson, Enterprise under Colonialism, p. 41. 64 Ibid. 65 Letter of João Perreira Caldas to Commandante, Vila da Borba 25 January 1787 cited in Memorias, vol 1, pp. 94–95. 66 Ferreira Lima, Formação Industrial do Brasil, p. 169. 67 On this point, see Eugene Ware Ridings, Jr., ‘The Bahian Commercial Association, 1840– 1889: A Pressure Group in an Underdeveloped Area’ (Ph.D., University of Florida, 1970), p. 16; Ferreira Lima, Formação Industrial do Brasil, p. 142; and Alden, Royal Government, p. 366. 92

Brazil: the origin of the textile industry manufacture. The impact locally in Brazil, as we have seen, does not seem to have seriously threatened production.

Figure 4.5 Clothed female Indians of an unknown tribe, c.1787. Source: ‘Índias Vestidas de Tribo não Identificada’, VF, I, Plate #107. The women illustrated here (and their husbands) are the couples who arrived in the Village of Borba from Spanish territories. It was the fine quality of the fabric of their dresses, which they had apparently woven themselves and are shown wearing, which resulted in the Portuguese assigning them the task of making cloth in Borba.

Nineteenth century After 1785, ‘semi-official’ sponsorship of Brazilian textile factories, as well as home production, continued unabated until European politics once again dramatically intervened. The Napoleonic invasion of Portugal, which forced the removal of the Portuguese royal family to Rio de Janeiro in 1808, had made it imperative to open Brazilian ports to foreign trade. In a very real sense, the colony was transformed into the metropolitan power; as the ‘new’ Portugal, therefore, Brazil would necessarily 93

The Ashgate Companion to the History of Textile Workers have to lift restrictions on manufacturing (including those on textiles) in order to survive. Thus, the 1785 prohibitions were officially revoked in 1808, in conjunction with the opening of trade to foreign nations. However, as has been shown above, textile production had never really ceased post-1785. It is, therefore, probably more accurate to characterize the post-1808 period as one of expansion of fabric manufacture, rather than its inception. With open encouragement of the now resident-in-Brazil royal government, new textile factories were established soon after manufacturing prohibitions were lifted, appearing in Ouro Preto (Minas Gerais) in 1814, in São Paulo by 1813, in the now Imperial capital of Rio de Janeiro in 1819 and 1826, and in the province of Pernambuco (1822).68 Such new factories produced not only cotton goods, but luxury items such as silks and trimmings of gold and silver.69 Production levels during the early nineteenth century appear to have been substantial; one well-known British source, which reported solely on production in Minas Gerais for the year 1827–1828, suggested that the amount of cloth that province exported, combined with the amount produced for domestic consumption, yielded an astounding figure of more than 8 million yards.70 Yet, in the following decade, Mineiro textile production figures were much lower than those of the 1820s, leading many historians familiar with these statistics to assume that the burgeoning industry had been undercut by outside competition. These historians reason that the preferred nation status gained by British products in the Brazilian market first through an 1810 Treaty, and later re-confirmed by independent Brazil in 1827 (these arrangements were essentially paybacks for British aid in challenging Napoleon in Portugal and helping the Portuguese safely reach Brazilian shores) had finally stymied local Brazilian textile production. According to one view, the stultifying impact of British textiles imports to Brazil began c.1830 and persisted until 1844, when passage of protectionist legislation ushered in a period of reinvigorated Brazilian textile manufacture;71 yet another perspective on this period proposes that the first viable textile manufacture can be confirmed for Brazil only after protection was implemented.72 These positions, however, are strongly contradicted by several indicators. Firstly, there were a number of factories created before 1844, at precisely the moment when the negative impact of British goods is considered to have been most disastrously felt. For example, after 1830 but before 1844, two substantial plants in the vicinity of Bahia, São Antonio de Queimado and Conceição, were already in operation. Then, 68 Dickenson and Delson, Enterprise under Colonialism, p. 41. 69 Nizza da Silva, Vila Privada E Quotidiano no Brasil, p. 234. 70 Libby, ‘Proto-industrialisation in a Slave Society’, p. 27 cites the survey of J.J. Sturz, an Englishman whose Review, Financial, Statistical and Commercial, of the Empire of Brazil and its Resources (London, 1837) suggests that for 1827–1828 ‘Minas has exported some 2,339,605 yards of spun cotton cloth’ while consuming at home ‘roughly 5,800,000 yards of cotton cloth’. 71 Dickenson, Brazil: Studies in Industrial Geography, p. 77. 72 This is the basic premise of Stein, Brazilian Cotton Manufacture. 94

Brazil: the origin of the textile industry in 1845–1846, just a year after the passage of the Protection Act and with scarcely enough time to radically alter investment conditions, the Todos os Santos factory was founded in the nearby Recôncovo area with a yearly output greater ‘than the first two combined’.73 Secondly, as I have demonstrated elsewhere, the British government routinely complained that exported goods (of which cotton textiles were statistically the most important) languished in the Brazilian market, where overstocking was a real concern. Even the removal of sugar duties on Brazilian sugar in the British market in 1846 failed to stimulate the sale of British manufactures in Brazil which could not be sold ‘except at ruinously losing prices’.74 Might one tentatively suggest that this was so because Brazilian factories met local needs? Supporting this conclusion is the evidence that British textiles which entered Brazil frequently bypassed the local consumer and were, instead, transhipped to Africa to pay for slaves marketed in Brazil.75 It is difficult to reconcile the puny showing of British textiles in Brazilian markets with the notion of such fabrics swamping Brazilian production. Given these indicators, the British threat to Brazilian textile production should definitely be re-evaluated. After 1850, the year in which the Brazilian slave trade ended, conditions for textile production improved even further. Eugene Ware Ridings Jr. notes that textile production in the Province of Bahia was favoured by a large resident labour force which could be employed for relatively small sums, as well as by the capital released from the ending of the slave trade.76 Not surprisingly, this capital would be invested in newly created enterprises, with the result that by 1866 Brazil could boast of nine textile factories (five of which were in Bahia). That number jumped to 30 manufactories, nationwide, by 1875.77 Ten years later, Brazil had 48 textile factories; the centre of textile production, however, had now shifted away from the north to the more centrally located Minas Gerais (where there had long been manufacture), and the south-eastern provinces of Rio de Janeiro78 and São Paulo.79 With profitability evident after the 1860s, entrepreneurs began to invest in modernized textile machinery.80 Textile manufacture now spread even further into

73 Ridings, Jr., ‘Bahian Commercial Association, 1840–1889’, p. 259. 74 Hudson to Palmerston, 10 October 1848, National Register of Archives (London) Broadlands Manuscript GC/HU/1–71 as cited in Roberta M. Delson, ‘Sugar Production for the Nineteenth Century British Market: Rethinking the Roles of Brazil and the British West Indies’, in Bill Albert and Adrian Graves (eds), Crisis and Change in the International Sugar Economy, 1860–1914 (Edinburgh, 1984), pp. 59–80, at p. 72. 75 Ibid. 76 Ridings, Jr., ‘Bahian Commercial Association, 1840–1889’, p. 259. 77 Dickenson, Brazil: Studies in Industrial Geography, p. 77. 78 Ridings, Jr., ‘Bahian Commercial Association, 1840–1889’, pp. 260–261. 79 In 1900 there were 17 cotton mills but by 1915 that number had surged to 41. This point is made in Warren Dean, The Industrialization of São Paulo, 1880–1945 (Austin, Tex., 1969), p. 83. 80 Ibid., p. 37. 95

The Ashgate Companion to the History of Textile Workers the south-eastern regions, where abandoned coffee lands were given over to cotton cultivation. The gender of nineteenth-century textile workers varied depending on the region and on the production configuration, that is either home/fazenda or factory. In some areas of Minas Gerais, where the first modality persevered (notwithstanding the creation of textile factories), Douglas Libby has calculated that women so dominated cloth production and, conversely, production dominated their existence, that over 90 per cent of the female population (both free and slave) spun and wove.81 In the factories, by contrast, the workforce included men and women, a direct parallel to the state-sponsored factories of the late eighteenth century. Since the institution of slavery was not abolished in Brazil until 1888, slaves continued to provide labour to the textile industry for the better part of the century. It was not unusual in the nineteenth-century factory for male and female slaves to work alongside free workers82 (much as slave women worked alongside their owners in the home/fazenda configuration). Again, the ratio of free to slave workers varied by geographic region. Brazil’s nineteenth-century textile factories were often financed with foreign capital. Thus it should not be surprising that in these settings, the upper echelon of workers, the technicians, was frequently non-Brazilian; it was rare that Brazilians were promoted beyond the position of assistant. Many of the managerial jobs were held by English workers, who brought with them a kind of global knowledge of the industry, as well as technical expertise to run imported machinery, now allowed in Brazil with the ending of restrictions on imported machinery in 1846 and 1847. The interchange between textile interests of both nations occasionally led to expert English spinners and weavers being relocated in Brazil; this was especially so after 1870 when depression and foreign competition forced large numbers of British cotton mills to close.83 Poor foreign immigrants, recently arrived in Brazil, provided another source of labour for textile factories. The Santo Aleixo factory near Rio de Janeiro counted 100 out of 116 workers as foreigners in 1850; 84 of them had come from the German settlement at Petropolis.84 There is no doubt that child labour was employed in some of these nineteenthcentury mills, when long work hours were standard, despite published work-day standards of 10 hours. In his seminal study of the nineteenth-century Brazilian textile industry, Stanley Stein remarks that when the installation of electric lights made night labour possible in the 1890s, labourers put in 14 to 17 hour days.85 As elsewhere, most of Brazil’s nineteenth-century textile workers were unskilled, illiterate and poor; some were foundlings.86 This is not to minimize the appearance 81 Libby, ‘Proto-industrialisation in a Slave Society’, p. 16. 82 See Soares, ‘A Manufatura na Sociedade Escravista’ for a discussion of slave labour in Brazil’s manufacturing establishments. 83 Stein, Brazilian Cotton Manufacture, p. 52. 84 Soares, ‘A Manufatura na Sociedade Escravista’, p. 33. 85 Stein, Brazilian Cotton Manufacture, p. 62. 86 . Ibid., p. 53. 96

Brazil: the origin of the textile industry of ‘industrial paternalism’ at mid-century which saw some mill owners providing housing, schooling and food. Such enlightened ‘benevolence’, however, was certainly not the norm in Brazilian manufacturing establishments and had a limited lifespan. By the end of the century, worker’s towns (vilas operarias) were replacing dormitory arrangements, with housing units rented out to workers rather than provided as before.87

The twentieth century to the present The political stability achieved with the overthrow of the Brazilian monarchy in 1899 and the establishment of a republican form of government allowed for expansion of the country’s nascent industries. Among these textiles predominated, with production increasing impressively up to the 1930s. Those who have intensively studied this early twentieth-century Brazilian textile industry include John P. Dickenson88 and Stanley Stein,89 both of whom divide the industry’s history into three periods: pre-1930, the Depression and war years, and post-Second World War. With hindsight we can view the time frame from 1900 to 1930 as a kind of ‘golden age’ of textile production in Brazil, as a number of factors, including newly established hydroelectric plants, increased immigration (largely from Europe), formation of investment capital siphoned off from coffee-growing profits and political stability coalesced into what appeared, on the surface, to be a very healthy industry.90 The importance of these factors is demonstrated by the growth in output of textiles, which continued to be largely cotton production (notwithstanding wool production in the far south); whereas in 1885 some 20,595,375 metres of cotton cloth were produced in Brazil, by 1929 the output had increased to over 500 million metres.91 Although most of the output continued to be of coarse and medium grade cotton cloth which was absorbed by internal markets, there was also a certain degree of experimentation with finer types of cloth intended to resemble in weave, finish and width the more highly regarded imported goods.92 Ironically we now can suggest that while colonial production provided a precedent for this type of premium production, as this discussion has shown, the imperatives of the nineteenth century forced the Brazilian textile industry into a pattern of producing lower quality materials. As a result of this re-diversification, which saw early twentieth-century Brazilian textile production characterized by the production of finer goods, foreign producers began to lose their competitive edge in the Brazilian market. The increased 87 Ibid., p. 58. 88 Dickenson, Brazil: Studies in Industrial Geography. 89 Stein, Brazilian Cotton Manufacture. 90 Ibid., p. 98. 91 Ibid., p. 100. 92 Ibid. 97

The Ashgate Companion to the History of Textile Workers Brazilian production, including that of fine cloth, led to an expansion in the number of textile mills as well as an increase in capital expenditure for modernized equipment. Increasing in tandem was the growth in the number of textile workers (three times the number in 1921 as in 1905), a shift which paralleled the overall growth of the industry.93 Not so obviously, the centre of textile production began to shift southwards to Rio de Janeiro and São Paulo, although mills never totally disappeared from the north-east. The overall impression of the industry, especially after the First World War, appears to be one of vigour, a view further bolstered by the profits engendered by the necessity of import substitution production during the war years. In point of fact, North American and European mills only began to re-enter the Brazilian market after 1924.94 Such an impression, however, might be deceptive. Notwithstanding the employment of 115,000 textile workers nationwide by the 1920s,95 and the importation of heavy machinery for the mills in the later part of that decade, the textile industry on the eve of the Second World War was in reality a technologically outdated sector of the economy. The full impact of this failure to modernize would be recognized during (and was, undoubtedly, magnified by) the worldwide Depression of the 1930s. Attempting to turn profits, manufacturers demanded that workers who had previously worked a negotiated eight-hour day now work 10 hours, or more; small mills were increasingly stretched financially to stay profitable. The phenomenon of ‘overproduction’ (and market saturation), perhaps evident to the prescient few during the 1920s boom days, was now correctly understood as inimical to the future of the industry.96 But, amazingly, by late 1939 and into the Second World War period, the textile industry had turned around once again. At the outset of this period, in late 1939, Brazilian manufacturers had begun selling cloth to neighbouring Argentina and Uruguay; two years later the outbreak of war caused European and Japanese textiles to disappear from the world market, again stimulating import substitution in Brazil.97 Probably the most significant factor leading to a modernization of textile production occurred as a result of the demand for cloth by Allied forces in the 1940–1945 period, when Brazil was one of the few nations positioned to fill the void. Modern (post-Second World War) history of the textile industry in Brazil, however, has not continued on this promising course. As elsewhere, modern Brazilian production has been characterized by strong unions and opposing manufacturers’ associations,98 and, typically, there continues to be tension between domestic production and that slated for export. Increased emphasis has been placed on the manufacture of apparel (increasingly for export), as opposed to cloth, 93 Ibid., p. 101. 94 Ibid. p. 110. 95 Dickenson, Brazil: Studies in Industrial Geography, p. 70. 96 Stein, Brazilian Cotton Manufacture, p. 142. 97 Ibid., p. 163. 98 Ibid.,p. 181. 98

Brazil: the origin of the textile industry with fibres other than cotton, including flax, silk and synthetics such as acetate and viscose introduced in the post-war years.99 The strength of the industry came increasingly to depend on government expenditure and protection, especially as equipment became obsolete by the 1960s.100 Special initiatives, such as SUDENDE in the north-east, modernized the industry in localized regions, but the result was a lessening of the need for labour. As elsewhere in Latin America, the 1970s oil shortages, the debt crisis of the 1980s, runaway inflation and shrinking government support adversely impacted manufacturing including the textile industry. By the early 1990s, these combined factors had resulted in a cut of the textile industry’s employment figures of approximately 50 per cent, as a now unprotected industry faced tough competition from Asian producers.101 To rectify the situation, United Nations programmes implemented in the 1990s aimed to strengthen Brazil’s textile industry by emphasizing and funding technological modernization, especially in the application of CAD/CAM technologies to the apparel industry, and in the establishment of degree programmes in textile science.102 The results have been largely positive; as of 2003–2004 some 30,000 textile companies of all sizes were operating in Brazil with capital expenditures of more than US  $6 billion in modernization programmes, resulting in increased productivity. The industry then supported 1.5 million jobs and was expected to generate more.103 Again, in another ironic reinvention of the colonial past, particular emphasis was placed on the supply of higher quality cotton produced in Mato Grosso104 which, as has been shown above, furnished extraordinary quality cottons for eighteenthcentury looms.

Conclusions This discussion has examined the issues concerning textile production in pre-modern Brazil from a revisionist perspective. During the first two centuries of colonization, the pre-industrial pattern of cloth production included simple home manufacture and estate production. In both instances, the output of such endeavours was intended for local consumption, and operations were essentially self-contained. By the eighteenth century the evidence suggests a dramatically 99 Dickenson, Brazil: Studies in Industrial Geography, pp. 82–83. 100 Ibid, p. 82. 101 United Nations Industrial Development Organization (UNIDO), http://www.unido,org/ doc/3616. 102 Ibid. 103 Denise Barbosa, ‘Textile Sector Overview’, in International Market Insight, US Commercial Service 12/03/2003. 104 Ibid. See also Todd Benson, ‘Brazil’s Big Stake in Cotton Likely to Become Bigger’, New York Times, Tuesday, 29 June 2004, W 1. 99

The Ashgate Companion to the History of Textile Workers altered production scale, with factory settings now providing a substantial portion of Brazilian textile manufacture. Moreover, fabrics produced in urban factories or in mills located in rural Indian communities contributed to what had become by the eighteenth century an internal textile trade, which apparently continued unabated into the 1800s. To the extent that fabrics were shipped to Portugal, that trade may also be considered an international one. The proto-industrialized textile factories of the 1700s were staffed by wageearning workers who were employed full time for the purpose of producing cloth. The pay scale, at least for weavers, apparently corresponded to the hours of work spent in the factory, rather than to pieces produced. There is no evidence from the factory data that ‘finishing’ work was done at home, or that there was auxiliary ‘putting out’ to home producers. Rather, factory labourers appear to have put in long hours in these state-run enterprises to produce a finished product in situ. Such enterprise suggests modernization as well as profit-orientation. The limited but available statistics on ratios of spinners to weavers in factories definitely confirms far greater efficiency than in home/fazenda settings. Neither the factories, nor home/estate production, nor even the internal trade, appear to have been eclipsed by the prohibitions of 1785. Indeed, the only impact of the legislation seems to have been to refocus the industry to the production of inexpensive textiles (although some fine fabrics continued to be produced at government request). While it is clear that Britain did absorb increasing amounts of Brazilian raw cotton through the late eighteenth century, emerging as a global giant in the textile world in the 1800s, there is no clear evidence that its production ever totally inhibited Brazilian textile production either in the eighteenth or nineteenth century. Beatriz Nizza da Silva has argued that Brazil was on the cusp of becoming a manufacturing presence before the prohibitions of 1785.105 This chapter has presented the evidence for a rather different interpretation as regards the textile industry. It appears that not only had cloth manufacture resulted in true industry by the eighteenth century, but its parameters were defined by the government, which encouraged its survival despite legal restrictions. In their scale and level of organization the proto-industrialized textile factories of late colonial Brazil suggest clear analogies to those of contemporaneous Portugal or England.106 Perhaps the only notable difference is that Brazilian textiles were never intended to supply global markets (the temporary trade with Portugal is a clear exception). The point of textile production in colonial and nineteenth-century Brazil was to provide needed fabrics to a very large and geographically far-flung internal market, and all indicators suggest that the goal was successfully met. While it is obvious that the momentum of textile production was by no means even, it is equally the case, as this discussion has shown, that it never really stopped altogether. Rejecting previous perspectives on Brazil’s cloth industry, this chapter concludes that the origins of today’s enormous textile output may be directly traced to the late colonial period 105 Dickenson and Delson, Enterprise under Colonialism, p. 39. 106 Ibid., p. 40. 100

Brazil: the origin of the textile industry and to its continuation unabated through the nineteenth century when home/ fazenda manufacture was finally superseded by the factory mode.

101

This page has been left blank intentionally

5 China Robert Cliver

Textiles have a very long history in China, as garments, currency, tax goods and commodities. Chinese domesticated silkworms no later than 3000 bce, and archaeologists have discovered cocoons at Neolithic sites. The ancient Chinese also wove ramie and hemp, which were worn by the common people. However, within just a few centuries of its transmission to China around 200 bce, cotton became the staple cloth of ordinary Chinese. In ancient China, weaving came to define women’s social, economic and moral role. Beginning as early as the Song Dynasty, however, the gender identity of textile work slowly changed, as the Chinese economy grew more commercialized. While imperial silk weaving workshops mainly employed men, rural women working at home produced most of China’s cotton cloth. Production of cotton and silk developed rapidly under state control during the Mongol Yuan Dynasty (1271–1368) and the Ming Dynasty (1368–1644), which maintained a caste of hereditary weavers producing for the needs of the state and the market. By the eighteenth century, the hereditary weavers had all but vanished, and most Chinese households wove cotton cloth for their own consumption or for market distribution. With the arrival of European warships, however, changes began which would radically and fundamentally transform Chinese textile production. 

This introduction to Chinese textile production before 1650 is based on relevant chapters of Kang Chao (Zhao Gang), The Development of Cotton Textile Production in China (Cambridge, Mass., 1977); Lillian M. Li, China’s Silk Trade: Traditional Industry in the Modern World, 1842–1937 (Cambridge, Mass., 1981); Yan Zhongping, Zhongguo mianfangzhi shigao 1280–1937: cong mianfangzhi gongyeshi kan Zhongguo zibenzhuyi de fasheng yu fazhan guocheng [Historical Outline of Cotton Textiles in China, 1280–1937: Looking at the Process of the Emergence and Development of Chinese Capitalism from the Point of View of the History of the Cotton Textile Industry] (Beijing, 1955); Xu Xinwu, Yapian zhanzheng qian Zhongguo mianfangzhi shougongye de shangpin shengchan yu zibenzhuyi mengya wenti [Commodity Production and Sprouts of Capitalism in China’s Cotton Textile Handicrafts Industry before the Opium War] (Nanjing, 1981); Zhu Xinyu, Zhejiang sichou shi [A History of Zhejiang Silk] (Hangzhou, 1985).  On silk weaving between 1000 and 1800, see Francesca Bray, Technology and Gender: Fabrics of Power in Late Imperial China (Berkeley, 1997).

Map 5.1

China

China The following pages explore the changes and continuities in Chinese textile production over the last 350 years, focusing on the dramatic and often painful changes of the eighteenth, nineteenth and twentieth centuries. The discussion includes government policies, the role of private capital, technological change, market development and foreign imperialism. But the changing roles of the women and men who spun and wove remain the primary focus throughout.

Commercialization and expansion of textiles during the Qing Dynasty (1644–1911) In the Qing Dynasty, the late Ming trend towards commercial textile production accelerated. In an edict promulgated 19 May 1645, the Shunzhi Emperor released all remaining hereditary artisans from bondage, making them ordinary subjects with no special obligation to the imperial state. Although state textile production was several times greater by the end of the seventeenth century than it had been during the Ming, imperial production declined as a proportion of total textile production. By the end of the seventeenth century, the Qing government had completed the conversion of taxes in kind to a consolidated land tax paid in silver. The state met its need for cotton and silk cloth mainly through market purchases. As a result of these policies, as well as a marked increase in population, markets and commercial textile production expanded rapidly, reflecting the broader commercialization of the Qing economy during the eighteenth century. As farmers in Guangdong began to produce new cash crops, such as sugar, they continued to produce cloth for their own consumption using cotton imported from Zhejiang and Jiangsu. Records indicate huge shipments of sugar from the south, which Guangdong merchants exchanged for Yangzi cotton. Merchant guilds and associations, called gongsuo, bang or hang, managed regional and long-distance trade in both raw materials and cloth. These merchants purchased raw cotton, thread and cloth produced by farming families in the Yangzi Delta and north China, marketing these goods throughout the empire. Although export statistics are scarce, there is documentary evidence that Chinese merchants exported cotton textiles to Japan as early as the sixteenth century. Chinese ‘nankeens’ were also exported to Europe and the Americas by the 1730s. Before 1831, Great Britain imported more ‘nankeens’ than it exported cotton cloth to China. Zhang Yan, Qingdai jingji jianshi [A Brief History of the Economy of the Qing Dynasty] (Beijing, 1998), pp. 419–466, 609–615.  Sucheta Mazumdar, Sugar and Society in China: Peasants, Technology, and the World Market (Cambridge, Mass., 1998). Also Zhang Yan, Qingdai jingji jianshi, p. 431.  Albert Feuerwerker, ‘Handicraft and Manufactured Cotton Textiles in China, 

105

The Ashgate Companion to the History of Textile Workers From the early Qing, imperial silk-weaving factories increasingly contracted work out to private weavers. This trend was accompanied by a flourishing of private commercial silk weaving in the Jiangnan region. In cities like Suzhou, male silk weavers gathered at bridges under guild auspices to await work in the shops of the master weavers or in the imperial factories. Although hiring practices were conditioned by the needs of the imperial workshops, as well as by guild traditions, these weavers’ lives were as insecure as those of modern proletarians. A late seventeenth-century source describes their existence in this way: ‘Like vagrants, they gather together, and after having rice gruel, they scatter and return [whence they came]. If the master-weavers should reduce their output, this group will have no sources of food and clothing.’ Despite a similar commercialization and flourishing of cotton production and trade in the eighteenth century, however, there emerged almost nothing resembling a factory system in cotton textiles. This is surprising in light of the appearance of handicraft workshops and wage labour engaged in calendering and dyeing, as well as silk weaving. The absence of workshop spinning and weaving is also surprising given that eighteenth-century cotton merchants were clearly able to amass large amounts of capital. Nor does technology seem to have been a significant impediment. By the early fourteenth century, China had developed large-scale, water-powered machinery for producing silk and hemp thread. The only technical innovation required was a drawbar to replace the spinner’s fingers in pulling the shorter filaments of cotton. A drawing of a five-spindle spinning wheel from the Ming Dynasty, which shows a drawbar, reveals that the technology was available, but never found widespread use. Explanations as to why China did not produce a capitalist industrial revolution have pointed to many factors, including cultural values, the impact of foreign competition, investment preferences, family values and the roles of the state, legal system and social structure.10 Other explanations have looked to population and China’s success in improving agricultural production as factors inhibiting

1871–1910’, Journal of Economic History 30 (2) (1970): 338–378.  Jiangnan, literally ‘south of the river’, refers to the region south of the lower reaches of the Yangzi River, including southern Jiangsu and northern Zhejiang.  Quoted in Li, China’s Silk Trade, p. 49.  Workshops processing cotton cloth multiplied steadily during the eighteenth and nineteenth centuries, and these workers proved sufficiently organized and conscious of their status to engage in strikes, as did silk weavers. Zhang, Qingdai jingji jianshi, pp. 455–464 mentions a number of these, including a 1715 strike by calendering workers.  Zhao Gang (Kang Chao) and Chen Zhongyi, Zhongguo mianyeshi [A History of China’s Cotton Textile Industry] (Taibei, 1977), pp. 81–99. 10 It is important to distinguish, following R. Bin Wong, between economic development and capitalist production. Clearly, England’s remarkable transformation is the exception, rather than the rule. But Chinese historians must nonetheless ask why this did not happen in China. 106

China industrialization.11 R. Bin Wong, whose work benefits from comparing specific regions of Europe and China rather than entire continents, demonstrates that Chinese farmers were more secure on the land than their European counterparts, a factor that inhibited the proletarianization of handicraft production. Working in a similar comparative framework, Kenneth Pomeranz has put forward an explanation for China’s late industrialization based on ecological constraints on land use and fuel.12 As we shall see in the next section, a series of catastrophic events in the nineteenth century also impacted China’s economic development just as Europe’s economy moved towards industrialization. In order to comprehend the particular history of cotton textiles in Qing China, however, it is necessary first to understand the nature of the rural household economy and its relationship to commercial markets and merchant capital. In this regard, the contrast between silk and cotton is instructive. The most important difference is that farming households in Qing China that grew cotton, and many that did not, produced cotton cloth for personal consumption as well as for market distribution. This was not generally true of silk. Most analyses of the family economy in this period hold that cotton weaving was an activity subsidiary to agriculture utilizing ‘surplus’ labour. Of course, rural women’s labour was by no means ‘surplus’, and to assume that there was no opportunity cost to the labour of the most skilled weavers is clearly absurd. Such women were responsible for an almost infinite variety of tasks on the farm and their labour contributed greatly to their families’ prosperity. However, especially in the late eighteenth century, cotton weaving proved attractive enough that large numbers of rural women, and many men, engaged in weaving for the market as the best means of benefiting the family economy. It is true, however, that cotton spinning often relied on labour with very low opportunity costs, such as that of older women, or girls as young as 9. As Kang Chao (Zhao Gang) notes, this is one of the main reasons why one does not find large-scale, complex machinery used in production of cotton textiles. Given that spinning and weaving were household activities, it was necessary that the equipment be light, cheap and capable of being operated by one person. By the late eighteenth century, nearly half of all farming families engaged in some aspect of cotton production, whether cultivation, spinning or weaving. However, not all families engaged in all three. Spinners and weavers in Guangdong or Fujian who did not themselves cultivate cotton required supplies of raw cotton. Weavers in Shaanxi required cotton thread, and many families that might grow and spin

11 For example, see Mark Elvin’s ‘high level equilibrium trap’ and Philip Huang’s ‘involution’ thesis. Mark Elvin, The Pattern of the Chinese Past (Stanford, 1973); Philip C.C. Huang, The Peasant Family and Rural Development in the Yangzi Delta, 1350–1988 (Stanford, 1990). 12 R. Bin Wong, China Transformed: Historical Change and the Limits of European Experience (Ithaca, 1997); Kenneth Pomeranz, The Great Divergence: China, Europe, and the Making of the Modern World Economy (Princeton, 2000). 107

The Ashgate Companion to the History of Textile Workers cotton commercially, but did not engage in weaving, had to purchase their cloth on the market, or even hire weavers on a seasonal basis.13 The relationships between cotton merchants and rural households were diverse and complex. In the eighteenth century, there emerged a kind of ‘putting-out’ system, called baomaizhuzhi in Chinese. The integration of cultivation, spinning and weaving in the domestic production system, however, often made ‘putting-out’ arrangements unnecessary. In cotton-producing regions such as the Yangzi Delta, rural weavers might trade cloth for warp. But in general they grew, spun and wove their own cloth, marketing their surplus through the ubiquitous cloth merchants. In regions that engaged in cotton textile production but did not cultivate cotton, the ‘putting-out’ system was a natural development. This was, however, a commercial relationship rather than a wage contract.14 Commercial spinning and weaving proved attractive to a great many farming households, for there was no other activity that would earn more for the family. Spinners in Central China and the Yangzi Delta could earn 30–50 per cent over the cost of raw materials. Depending on prices, annual earnings from cotton spinning would have been enough to support at least a child, and at most a grown woman and a child. Weaving could be even more lucrative. Given stable price ratios, spinners/weavers who purchased ginned cotton could expect to make a profit of 300 per cent, while those who grew, spun and wove their own cotton could earn as much as four times the market price of raw cotton. The data are incomplete, and the debate over household earnings from handicraft textiles in Qing China continues. The most recent work, however, indicates that, under favourable market conditions such as obtained in the mid-to-late eighteenth century, a weaver’s earnings could be sufficient to support 1.4 to 1.9 adults. Given that women and children consumed less than men engaged in farming, a woman weaving for the market could potentially earn enough to support herself, two children and perhaps another adult.15 We see no evidence, however, of capitalist relations of production developing in cotton weaving in Qing China. Household-based production remained the norm. This situation contrasts with silk weaving which, as mentioned above, was practised not only in rural households, but also in urban workshops employing wage workers. Furthermore, as we shall see, when handicraft cotton weaving revived in the late nineteenth century, it was again on the basis of the rural household. Why did household production of cotton textiles prove so resilient and long-lived in China? Aside from conditions that applied to the Qing economy more generally, 13 Fang Xing, ‘Lun Qingdai qianqi mianfangzhi de shehui fen gong’ [‘On the Social Division of Labour in Cotton Textile Production in the Early Qing Dynasty’], Zhongguo jingjishi yanjiu [Economic History Research in China] 2 (1) (1987): 83–89. 14 For more on the early ‘putting-out’ system in Chinese textiles, see Xu Xinwu, Yapian zhanzheng qian Zhongguo mian fangzhi shougongye de shangpin shengchan yu zibenzhuyi mengya wenti and Yan Zhongping, Zhongguo mienfangzhi shigao 1280–1937. 15 Fang, ‘Lun Qingdai’, pp. 92–94. For a detailed discussion of the data and analysis, see Pomeranz, The Great Divergence, Appendix E, pp. 316–26. 108

China the most important reason for the durability of household cotton weaving is that farming families preferred to weave cloth to meet their own needs if this was at all possible. Even when markets declined in the nineteenth century, and many merchants left the cotton trade, farming families continued to spin and weave for their own consumption. Household-based textile production enjoyed other advantages as well. Kang Chao asserts that, in contrast with capitalist manufacturers, farming families were inclined to maximize output rather than profits, and would accept returns which could fall below the level at which wage labour became unprofitable. In effect, households treated labour as a fixed factor of production rather than as a variable input. When combined with the fact that cotton production was necessary to meet families’ own needs, and the labour devoted to textile production enjoyed relatively low opportunity costs, the domestic production system proved highly resistant to efforts to develop factory-based cotton weaving well into the twentieth century.16 Chao’s analysis of the economics of household textile production is essentially correct, even if it was not true that the earnings of all cotton weavers were below subsistence levels at all times. Although eighteenth-century weavers were clearly able to earn enough to make their work economically profitable, the very nature of the Chinese household economy produced a robust system that survived even when earnings from cotton textiles fell below subsistence levels, as they seem to have done during the tumultuous nineteenth century.

Nineteenth-century changes During the reigns of Jiaqing (1796–1820) and Daoguang (1821–1850), earnings from cotton textile production declined, especially in the core Yangzi Delta region. Several factors combined to depress cloth prices. Beginning in the early nineteenth century, the price of raw cotton began to increase relative to prices for cloth, which strongly affected the earnings of both weavers and spinners. As profits fell and markets shrank, merchants began to withdraw from the cotton trade. Rural households wove less for the market, but generally continued to weave for their own consumption. The development of cotton weaving in provinces outside of the core cotton regions affected cloth prices throughout China. As population and markets expanded during the eighteenth century, rural families in Guizhou, Shaanxi and the Middle Yangzi region (Hubei and Hunan Provinces), which had previously imported cloth from the Yangzi Delta, began to produce their own textiles, distributing their products to neighbouring regions and taking advantage of lower transportation

16 Chao, Development of Cotton Textile Production, pp. 30–47; Zhao and Chen, Zhongguo mianyeshi, pp. 58–65. 109

The Ashgate Companion to the History of Textile Workers costs. The diffusion of household textile production reduced markets for producers in core regions like the Yangzi Delta.17 The military conflicts and violent uprisings that plagued the Qing in the nineteenth century disrupted trade and reduced agricultural productivity. The Taiping Rebellion, which lasted from 1851 to 1864 and resulted in an estimated 20 million deaths, devastated the cotton-producing regions along the Yangzi River. In addition to internal upheavals, the Qing also faced the threat of European navies beginning with the First Opium War of 1841–1842. This resulted in the opening of the Qing Empire to trade and direct foreign intervention. In 1860, the Qing government was forced to cede control of the imperial customs administration to foreign nationals. Initially restricted to a few treaty ports, within a generation these foreigners were travelling, trading and proselytizing throughout the empire. Foreign goods, ideas and technologies flooded into China. The specific effects of European imperialism and trade are debatable, but the overall impact of China’s encounter with the West was clearly epoch-making. For textile merchants, the most immediate effect was that the outflow of silver through the opium trade affected the exchange rate between China’s two currencies – copper cash and silver ingots. As silver became more expensive relative to cash, merchants who were previously able to pay silver in advance for supplies of cloth, found themselves less able to make purchases, which affected circulation and aggravated falling cloth prices. The increased price of silver also amounted to a revaluation of the Chinese currency, strengthening Chinese purchasing power and making imports less expensive.18 Perhaps more significant in the long run, imports of European textiles, especially cheap, high-quality, machine-spun thread, began a fundamental and initially very painful transformation of cotton textile production in China. Although imported thread comprised only about 0.12 per cent of the total consumed in China, imports were perceived as a threat in some regions. As early as 1831, peasant producers in Guangdong Province seized and burned stores of imported thread.19 After the Treaty of Nanjing (1842), Europeans gained the right to import cotton cloth and thread freely. Despite some initial setbacks, from the 1870s, British textile exports to China steadily increased, as did exports from India, Japan and America. British cotton exporters, however, found that forcibly opening China’s markets was not enough to guarantee consumer acceptance of British goods. Residents of north China wanted heavy cloth that provided protection from dust and cold and held up under strenuous wearing. Clothing made from imported British cloth might last only one year, whereas Chinese hand-woven cloth would last three years. One response was to adulterate lighter cloth by filling it with clay and chalk. While this gave the appearance of heavier hand-woven cloth, these qualities disappeared 17 Fang, ‘Lun Qingdai’, pp. 92–93. 18 Shou-eng Koo, Tariff and the Development of the Cotton Industry in China, 1842–1937 (New York, 1982), pp. 74–79. 19 Xu Xinwu and Byung-Kun Min, ‘The Struggle of the Handicraft Cotton Industry against Machine Textiles in China’, Modern China 14 (1) (Jan, 1988): 31–49, at p. 33. 110

China after the first washing. Naturally, this further disinclined Chinese consumers to purchase British cloth.20 Imported cloth did find favour with consumers in some parts of China, however. As early as the 1850s, 90 per cent of consumers in Fujian Province purchased foreign cloth. By the 1890s, coarser and heavier American and Japanese cloth competed successfully with Chinese domestic hand-woven, while Britain continued to dominate the market for lighter, high-end cloth.21 Nonetheless, imported cloth only comprised a small proportion of cotton cloth consumed in Qing China – little more than 3 per cent in 1860 and about 13 per cent in 1894.22 Imported machine-spun thread enjoyed more success in Chinese markets and had a greater impact on Chinese textile production, but change was slow nonetheless. All goods transported within China were required to pay internal transit taxes called likin, or lijin, which were applied to both imported and domestic products.23 When Europeans took over China’s customs administration in the 1860s, they established a system whereby, in addition to the standard 5 per cent ad valorem tariff, importers could pay 2.5 per cent in lieu of all likin taxes. As domestic thread enjoyed no such provision, it was often more profitable for Chinese merchants to ship raw cotton overseas, to Japan or India, than to transport cotton to inland regions of China such as Sichuan. As a result of this situation, in provinces that did not produce cotton, imported thread was often cheaper than Chinese raw cotton. With the development of mechanized spinning in Japan and India, which enjoyed similar factor endowments of cheap labour and locally grown cotton, the cost of imported thread fell even further. China’s average annual yarn imports increased 24 times between 1871 and the first decade of the twentieth century. Imports of cotton piece goods during the same period merely doubled.24 Hand-spinning continued to survive in China, however, and as late as 1894, imported thread comprised only about 23 per cent of the total consumed in China.25 Nonetheless, imports did affect some producers, especially in the coastal areas. In the 1860s, spinners in Guangzhou rioted in protest, and by the 1870s, as foreign thread penetrated inland markets, similar protests began to appear in other parts of the empire. Importation of Indian thread at a cost equal to or lower than Chinese raw cotton dictated that commercial hand-spinning would eventually disappear. Hand-spinning survived only because many cotton cultivators were willing to spin thread for their own use, or for the shrinking market for hand-spun yarn. By the 1930s, hand-spinners, who by that time produced only one-quarter of China’s 20 Koo, Tariff and the Development of the Cotton Industry in China, pp. 57–61, 85. 21 Fang, ‘Lun Qingdai’, p. 93; Chao, Development of Cotton Textile Production, pp. 169–171. 22 Xu and Min, ‘Struggle’, pp. 38–39. 23 The likin was established in the 1850s, during the Taiping Rebellion. Originally designed to impose a minimal burden while increasing central government revenues, within a decade collecting stations had proliferated, and the policy met with strong and even violent opposition from merchants, farmers and artisans. Susan Mann, Local Merchants and the Chinese Bureaucracy, 1750–1950 (Palo Alto, 1987), ch. 6. 24 Feuerwerker, ‘Handicraft,’ p. 342. 25 Xu and Min, ‘Struggle’, p. 34. 111

The Ashgate Companion to the History of Textile Workers thread supply, earned little more than two cents a day, a mere fraction of the wages of a worker in a modern thread mill.26 In contrast, the influence of the world economy on Chinese silk production was initially very different. Raw silk27 exports developed rapidly from the 1830s, fuelled by demand from European weavers. When French and Italian sericulture collapsed due to an epidemic of pebrine silkworm disease in the 1850s and 1860s, foreign demand for Chinese silk increased dramatically. Export demand stimulated the spread of sericulture outside the core Jiangnan region. By the 1870s, the export trade in raw silk was thriving. However, whereas Jiangnan sericulturalists tended to plant at least half of their land to rice, farmers in the new sericulture regions, especially Guangdong in the south, grew increasingly vulnerable to fluctuations in world markets and exchange rates.28 Due to these differences, a comparison of cotton and silk production can illuminate both the varied impact of global capitalism and the influence of recent Chinese history on the processes of textile industrialization. This offers an opportunity to examine the diverse effects of China’s incorporation into the capitalist world system, as well as the specific characteristics of the linked processes of mechanization and proletarianization in the Chinese context. The following section outlines the varied trajectories taken by cotton and silk, thread and cloth, from the late nineteenth century to the Second World War.

The mechanization of thread production and the revival of household cotton weaving By the late nineteenth century, most leaders in government and society recognized that China would have to establish its own industry in order to compete in the modern world. Responding to export demand for higher quality thread, silk filatures utilizing steam reeling began to appear from the 1870s.29 Without tariff control, however, it was difficult to protect China’s nascent cotton industry from 26 For a fuller description see Yan, Zhongguo, pp. 59–77. 27 Raw silk refers to silk thread that has been reeled from silkworm cocoons, but has not yet undergone any of the many other processes that can be performed on silk thread. It is rather stiff and coarse to the touch, almost like flax but composed of much longer continuous fibres. 28 For an elaboration of this thesis, see Alvin Y. So, The South China Silk District: Local Historical Transformation and World-System Theory (Albany, 1986). 29 In the mid-nineteenth century, Chinese silk was usually reeled by hand in sericulturalists’ homes. The thread thus produced was often of poor quality, uneven and discoloured by smoke from boiling the cocoons. As American demand began to increase with the development of the silk and ribbon industries in places like Paterson, New Jersey, US importers sought to improve supplies of Chinese silk. This led to the establishment of steam filatures in South China, and later in Shanghai and the Jiangnan region. 112

China foreign imports. Both India and Japan established cotton mills in the 1850s and 1860s. China, however, only established its first large-scale mechanized mill in 1890, and did not have a million spindles until 1915. India had surpassed that amount in 1875. As a result, by 1900 China was the world’s largest importer of cotton yarn, and the second largest importer of cloth in the world.30 Why did the process of mechanizing cotton textiles prove so difficult, and what were the effects of foreign influence and Chinese conditions for cotton and silk respectively? From the late nineteenth century, Japanese silk producers, assisted by government development of national banks and shipping companies, managed to take control of the export trade and its high profits. In contrast, until the Second World War, foreign firms, mainly British, French and American, dominated Chinese silk exports. This contributed to a separation of production and sales that inhibited efforts to improve quality.31 In order to stimulate production of high-quality silk of even width and colour, foreign exporters attempted to set up steam filatures in Shanghai and Canton in the 1860s, but these usually failed, unable to obtain a sufficient supply of cocoons. The first successful filature was established by Chen Jiyuan, who had studied French steam reeling techniques in Vietnam.32 Although his first effort was stymied by artisan opposition, once the local gentry in his native Nanhai realized that machine-reeled silk brought prices 30–60 per cent higher than hand-reeled silk, they quickly invited him back to set up a filature. Recruiting local women as workers, and offering high prices for cocoons to ensure a stable supply, Chen’s rural factory soon employed some 300 workers. Within 10 years there were 10 such filatures in Guangdong Province, with a total of 2,400 basins. By century’s end, virtually all of the province’s raw silk exports were produced by steam filatures. In the Yangzi Delta, filatures were successfully established in Shanghai from the 1880s, and by 1901 there were more than 20. Three years later, local elites in Wuxi, who had already gained a high degree of control over cocoon supplies, began to establish their own filatures, competing with Shanghai. With both government and foreign support, however, Shanghai’s silk reeling industry continued to develop, and by 1930 there were more than one hundred filatures there operating some 25,000 basins.

30 Sung Jae Koh, Stages of Industrial Development in Asia: A Comparative History of the Cotton Industry in Japan, India, China, and Korea (Philadelphia, 1966), pp. 88–90, 190–200; Chao, Development of Cotton Textile Production, pp. 106–167. 31 Hu Minjian, ‘Waishang yanghang kongzhi de huasi chukou’ [‘The Export of Chinese Silk Controlled by Foreign Merchants’], in Lu Jianxin and Wanyan Zhaoyuan (eds), Ershi shiji Shanghai wenshi ziliao wenku [A Collection of Essays on the History of Shanghai in the Twentieth Century], vol. 4 (Shanghai, 1999), pp. 45–62. 32 The following account is mostly based on Li and So, who sometimes do not agree. When there is a conflict, I have looked to see which scholar is closest to the primary source for the information. For example, Li states that Chen studied steam reeling in France, while So, who cites Chen’s autobiography, states that he studied in Vietnam. 113

The Ashgate Companion to the History of Textile Workers Silk reeling was very much an industry owned and operated by Chinese businessmen.33 In both South China and Jiangnan, local elites provided social, political and financial support. Local gentry legitimized filatures’ hiring of local girls through their control of clan and lineage organizations based on male bloodlines and organized around the ancestral temple. They also gained control of cocoon supplies, relying on their wealth, social status and connections with county governments. Usually considered bastions of conservatism, in the case of filatures, the local gentry led the introduction of modern technology into silk textiles. If steam filature technology was imported, however, the organization of the silk reeling industry was strongly influenced by Chinese conditions. The silk trade was fraught with risk. The filatures were dependent on foreign markets, but separated from them by a long chain of intermediaries. Raw materials supplies were highly unstable, and prices could vary dramatically. The work was seasonal and factories often lay idle. In response to these uncertainties, all players in the silk industry sought to reduce risk to a minimum. Filature owners, who could organize the financing to build the factory and purchase the equipment, usually rented the facilities to an operator. In order to purchase cocoons, by far the largest portion of production costs, the operator would have to obtain a loan. The lender would usually take as much as half of the cocoons as collateral. The finished product was then purchased by another middleman, who might store the raw silk until international markets became more favourable.34 Gentry domination of cocoon hang facilitated the cheap and relatively stable supply of cocoons to filatures, but also inhibited efforts to improve sericulture under the Nationalist government in the early 1930s. Gentry corruption of state sericulture improvement programmes for profit led farmers to mistrust governmentsponsored efforts to eradicate silkworm disease and improve breeding techniques. For example, local leaders’ replacement of improved silkworm eggs with inferior strains resulted in riots that destroyed sericulture research stations, egg breederies and even schools and training institutes.35 Increased exports may also have had an adverse effect on Chinese silk weavers. Weavers in Canton, 10,000 of whom participated in a martial arts society and weavers guild, violently protested steam filatures in 1881. Silk weavers’ guilds were very active both in the south and in Jiangnan, protesting layoffs and declining 33 There were a handful of European and American filatures in Shanghai. In Shandong and Manchuria there were also many Japanese filatures specializing in tussah silk. Japanese also established filatures in the Jiangnan region, paying very high prices to secure supplies of cocoons. 34 Lynda S. Bell, One Industry, Two China’s: Silk Filatures and Peasant-Family Production in Wuxi County, 1865–1937 (Stanford, 1999) describes the organization of the Wuxi silk industry in fascinating detail, especially pp. 89–108; See also Li, China’s Silk Trade, pp. 163–173. 35 Bell, One Industry, pp. 145–153; Also Terry M. Weidner, ‘Local Political Work under the Nationalists: The 1930’s Silk Reform Campaign’, in Association for Asian Studies Conference, Chicago, 1982, Essays in the History of the Chinese Republic (ChampagneUrbana, 1983). 114

China wages from the 1890s through the 1910s. However, there is also evidence that demand for hand-reeled silk from the weaving industry may have held back the further development of steam filatures in the Jiangnan region.36 The poor quality of Chinese silk, as well as the separation of production and marketing, contributed to China’s losses in competition with Japan. By the 1920s, Japan controlled 90 per cent of the silk market in America, the world’s largest importer. In contrast with Japan, China lacked a strong central government that could implement technical and financial policies to benefit the industry. Even during the Nanjing Decade (1927–1937), the Nationalist Party’s programmes usually foundered on the rocks of peasant mistrust and elite co-optation. It proved impossible to eliminate silkworm disease, the main cause of unstable and poor quality supplies. At the same time, the government taxed imports of rayon, which Chinese weavers might have used to supplement problematic raw silk supplies. Despite these difficulties, however, China’s silk reeling industry continued to expand and develop, peaking in the late 1920s before depression and war drove down production and sales. Although both countries’ silk industries suffered from the depression, between 1930 and 1935, Chinese exports dropped by 50 per cent, while Japan’s only fell by 15 per cent. Domestic production dropped accordingly.37 Production revived somewhat in 1936 and 1937, but the Japanese invasion and subsequent occupation drove the industry nearly to extinction. Even during the depression, however, some managed to profit. Wuxi’s ‘silk magnate’, Xue Shouxuan, established both vertical integration and a near monopoly on filature production through a combination of coercive mergers, control of cocoon hang and exploitation of government programmes that inadvertently provided an opportunity for Xue to increase control over the prices paid to sericulturalists for their products. Silk reeling thus presents an example of early export-driven industrialization. China’s silk filatures overcame the impediments to industrialization traditionally posited by scholars as explanations for Chinese backwardness. Given that silk reeling overcame these constraints, we must ask what factors delayed a similar industrialization of cotton production in late Qing China. James Morrell describes the ‘excruciating deadlock’ created by the conflict between the imperialist treaty port system and Qing government policies.38 According to this thesis, during the period from 1870 to about 1890, efforts to establish cotton mills failed due to resistance and obstruction on both sides. In 1877, Nanjing Governor General Shen Baozhen successfully opposed the establishment of the Anglo-Chinese Steam Cotton Mill, raising the spectre of riots by handweavers. The first Chinese attempt at establishing the Shanghai Cotton Cloth Mill

36 Contrast Li, China’s Silk Trade, pp. 118–129, and So, The South China Silk District, p. 111, on this issue. Also Bell, One Industry, p. 79. 37 Ding Kangshi, ‘Woguo cansi chanxiao gaikuang,’ [‘Situation regarding China’s Silk Production and Sales’], Jinrong [Finance], 18 July 1947. 38 James Robert Morrell, ‘Origins of the Cotton Textile Industry in China, 1865–1915’ (unpublished Ph.D. thesis, Harvard University, 1989), pp. 88–90, 190–200. 115

The Ashgate Companion to the History of Textile Workers under official auspices failed in 1882 when the project lost backing after the United States sent a gunboat to Shanghai to support American cotton interests. From 1890 to 1896, the Qing government actively promoted mechanized mills, adding 417 thousand spindles in six years. However, foreign interference, as well as direct competition, continued. After several years of gathering capital, about half of which was supplied by the government, the Shanghai Cotton Cloth Mill reopened in 1890, marking the first successful establishment of a mechanized cotton mill in China. However, when the mill caught fire in 1893, the British-controlled fire department refused to put out the blaze, resulting in loss and damage from which the mill never recovered. Qing officials also continued to hinder the establishment of foreign-operated mills, impounding British shipments of production equipment and intimidating workers who sought employment at foreign mills. The Treaty of Shimonoseki, which ended the Sino-Japanese War of 1894–1895, broke the deadlock by resolving the issue of whether or not foreigners were legally permitted to establish manufacturing enterprises in Chinese cities. Although the Japanese did not pursue this option until 1902, a few British, American and German mills appeared beginning in 1897. During this same period, some Qing officials, such as Li Hongzhang, Shen Baozhen and Zhang Zhidong, actively encouraged investment in cotton manufacturing.39 Zhang Jian, an official and native of Nantong who established the Dasheng Cotton Mill there in 1899, strongly emphasized the importance of cotton textiles to the nation’s industrialization and overall economic development. His vision included linkages between regional and national development, as well as among agriculture, industry and education. Although his own efforts focused mainly on Nantong (and his own business interests) Zhang also called for a policy of national economic development based on cotton and iron.40 China’s spindles increased from 35,000 in 1890 to more than 1 million in 1915, and total production increased from 22,000 to 645,000 bales. However, China’s cotton textile industry did not really begin to flourish until the First World War. By 1922, China had more than 3.6 million spindles and output had increased to almost 2 million bales. The curtailment of imports during the European war also led to a brief resurgence in hand-spinning, but from the 1920s, both handicraft and imported thread lost ground to domestic machine-spun yarn. From 1925, China became a net exporter, and by the early 1930s, was largely meeting demand for cotton thread with domestic production. After a stagnant period during the global depression, capacity continued to increase, and by 1936 China’s textile industry boasted more than 5.6 million spindles.41 When China’s cotton industry began to take off during the First World War, the majority of cotton mills were Chinese. But during the post-war depression, 39 Richard C. Bush, The Politics of Cotton Textiles in Kuomintang China, 1927–1937 (New York, 1982), p. 48. 40 Elisabeth Köll, From Cotton Mill to Business Empire: The Emergence of Regional Enterprises in Modern China (Cambridge, 2003). 41 Zhao and Chen, Zhongguo mianyeshi, pp. 285–287, 294–297. 116

China Japanese companies came to dominate textile production in the treaty ports. Increased Chinese tariffs on high-count thread and factory legislation regulating night shift employment in Japan encouraged Japanese textile companies to set up production in China. By 1925, 32 out of 58 cotton mills in Shanghai, where about half of China’s cotton mills were located, were owned by 11 Japanese companies as compared to just 22 mills owned by 18 Chinese companies.42 Four other Shanghai mills were British. Japanese mills operated almost half of all cotton spindles in Shanghai, and 37 per cent of the total for China. Before 1945, Japanese mills in Shanghai consistently operated more spindles than Chinese firms, and from the 1930s possessed more than 1.5 times as many power-looms. However, outside of the treaty ports, more than half of cotton textile production remained in the hands of Chinese businessmen and local governments. This supply of domestically produced machine-spun thread provided the basis for a revitalization of rural handicraft weaving and the establishment of both putting-out arrangements and handicraft workshops. Before the Second World War, handicraft weavers continued to produce about half of China’s cotton cloth. There were several reasons for this. Whereas mechanized spinning quickly replaced hand-spinning due to a gain in efficiency of some 32 times, power-loom weaving was only four times more efficient than hand-loom weaving, and required much greater investment in machinery, training, maintenance and industrial plant.43 Furthermore, markets for Chinese hand-woven cloth remained strong despite the growth of power-loom production and foreign imports. Chinese handicraft products successfully competed with machine-woven cloth both in Shanghai and in international markets. Between 1871 and 1936, China’s exports of handicraft cotton textiles grew from 3,903 to 201,486 dan,44 an increase of 319 times. Nonetheless, by 1936, machine weaving supplied 45 per cent of China’s cotton cloth, while handicraft weaving still supplied about 43 per cent.45 The organization of handicraft weaving also changed during the interwar period. Increased supplies of machine-spun thread enabled the emergence of ‘putting-out’ arrangements, especially in the new textile-producing regions of North China. Direct ties between urban commercial firms and weaving households enabled commercial capitalists to set the size and style of the cloth produced according to shifting market demand. The contractors usually supplied weavers with iron gear looms, which by the 1920s were three times faster than traditional wooden 42 Of the Japanese mills in Shanghai, 11 were owned by the huge Naigai Wata conglomerate. Japanese firms were generally much larger and more heavily capitalized than their Chinese counterparts. 43 Only the largest, most heavily capitalized cotton mills were able to invest in complex and expensive cotton-weaving machinery. Japanese firms owned half the power-looms in China, and only the largest Chinese mills attempted to integrate spinning and weaving. Zhao and Chen, Zhongguo mianyeshi, pp. 229, 285–287; Bush, Politics of Cotton Textiles, pp. 23–25, 31–32. 44 One dan is about 133 pounds. 45 Xu and Min, ‘Struggle’, p. 41. 117

The Ashgate Companion to the History of Textile Workers looms, and were relatively small, inexpensive and easy to use.46 Cotton weaving quickly became an attractive way to recover from a flood or simply a means of increasing the family income. In the new cotton centres such as Baodi and Gaoyang, 60–70 per cent of household weavers worked for large cotton companies that controlled thousands of looms.47 This system allowed merchant capitalists to reduce risk and provided them with a high degree of flexibility. As cotton firms shifted from a system of payment in cash to payment in surplus thread, even more of the risk of price fluctuations was displaced onto weavers. The degree to which rural weaving workshops employing wage labour developed in this period is less certain. There are accounts of thriving handicraft weaving industries both in North China and in the Yangzi Delta. The former were often owned by farmer/weavers who employed family members and sometimes hired male weavers locally. The latter were often established by textiles merchants and hired women weavers from the surrounding area.48 These rural weaving workshops enjoyed many advantages over the larger, mechanized mills. They employed the same cheap, reliable technology adopted in household weaving and drew upon local sources of capital and labour, which offered lower costs, flexibility and rapid turnover of capital. When demand lagged, local weavers could return to farming until conditions improved. Their small scale and minimal capitalization enabled rural workshops to adjust product mix to meet shifting market demand, and these enterprises came to dominate the markets for higher value-added specialty textiles. These small-scale, rural workshops also benefited from policies to promote native industries and were generally able to avoid paying the taxes levied on their larger, urban competitors.49 Although there are reports of handicraft factories with as many as 1,000 looms in the late nineteenth century, they did not survive for long. Smaller workshops, with an average of 12 looms, were better able to reduce management expenses, maintain flexibility and lower costs. At their peak in the 1920s, rural and suburban textiles workshops probably numbered several thousand, but a 1933 survey found only 2,281 such workshops. Some reports indicate that handicraft weaving workshops had revived by 1937, and employed tens of thousands of workers.50 Despite the strength of handicraft weaving workshops in some regions, by the 1930s, production of cotton cloth in China was, for the most part, dominated by 46 Xu Xinwu also states that the newer looms could produce wider bolts of coth as well as interwoven piece goods. Xu and Min, ‘Struggle’, pp. 44–46. 47 On complex production relations within the Gaoyang weaving district, see Linda Grove, A Chinese Economic Revolution: Rural Entrepreneurship in the Twentieth Century (Lanham, Md., 2006). 48 See, for example, Xu Xinwu (ed.), Jiangnan tubu shi [A History of Native Cloth in the Yangtze Delta Region] (Shanghai, 1992); also Linda Grove, ‘Rural Manufacture in China’s Cotton Industry, 1890–1900’, in Douglas A. Farnie and David J. Jeremy (eds), The Fibre that Changed the World: The Cotton Industry in International Perspective, 1600–1990s (Oxford, 2004), pp. 431–459. 49 See Part I of Grove, Chinese Economic Revolution. 50 Grove, Chinese Economic Revolution, pp. 121–146. 118

China modern integrated mills and dispersed household weavers.51 According to rough estimates, handicraft production of cotton cloth increased from 1.8 billion square yards in 1909 to a peak of about 2.8 billion in the 1920s when it accounted for 86 per cent of all cotton cloth produced in China. Although production dropped to 2.3 billion square yards in the mid-1930s, handicraft weavers still produced about one-half of domestic production.52 In the silk weaving industry, although tens of thousands of Jiangnan men (and some women) continued to earn a living through household handicraft silk weaving, silk manufactories employing wage workers flourished from the late nineteenth century. Despite competition with exporters for access to raw silk, silk weavers managed to obtain supplies of raw materials, mainly relying on handreeled silk while filature silk was exported.53 Exporters of silk fabrics faced stiff competition from Japan in international markets. However, China’s large domestic demand supported a thriving weaving industry. Despite some initial resistance, Chinese weavers also began using synthetics such as rayon, and later nylon, which were easy to use and almost as strong as silk, although they did not take dyes with the same depth and brilliance. By the 1930s, the majority of Chinese silk products included at least some synthetic fibres. As earnings from independent handicrafts began to decline in the twentieth century, household silk weavers migrated to urban factories. During the First World War, silk production shifted from the old centres of the imperial silk weaving industry, such as Nanjing, to cities like Suzhou, Huzhou and Hangzhou, where the silk weaving industry consumed more than 80 per cent of local cocoons. Usually small, with less than ten looms, silk workshops utilized the most modern patternsetting looms. In the 1920s, many factories purchased electric-powered looms, which were much smaller and cheaper than power-looms for cotton weaving. From the late 1920s, Shanghai became the major centre for power-loom silk weaving. By 1931 there were 500 weaving workshops in Shanghai operating about 6,000 powerlooms. By 1936, when there was a bumper cocoon harvest, there were more than 7,000 power-looms in Shanghai and more than 10,000 in other Jiangnan cities.54 The paths to industrialization of cotton and silk were thus very different, as were the histories of thread and cloth. The men and women recruited to work in China’s cotton mills, silk filatures and weaving workshops exhibited similar diversity. Work in textile industries was differentiated by gender, place of origin, skill and market. Size, organization and degree of integration into the world economy created different conditions for workers in the various textile industries. The following two sections examine the nature and origins of these textile workers, 51 Zhao and Chen, Zhongguo mianyeshi, pp. 230–238. 52 Ibid., pp. 216–221, 249–250; Xu and Min, ‘Struggle’, p. 41. 53 This would account for the early decline of handicraft silk weaving in Canton, while silk weaving workshops flourished in older sericulture regions with fewer filatures such as the Lake Tai area. 54 Shanghai Silk Gazetteer Editorial Committee (ed.), Shanghai sichou zhi [Shanghai Silk Gazetteer] (Shanghai, 1998), p. 172. 119

The Ashgate Companion to the History of Textile Workers as well as their working lives, their politics and the ways in which they reacted to changes in their industries.

China’s industrial textile workers The number of cotton workers in Shanghai, the treaty port metropolis where about half of all Chinese industry was located, increased from just over 33,000 in 1921 to a peak of 131,000 in 1930 before declining to roughly 100,000 in 1935. About half of Shanghai’s industrial workforce was employed in cotton textiles. In Tianjin, the cotton textiles workforce peaked at around 17,000 in 1929. In that year, over 50,000 Shanghai workers were employed in silk reeling, with another 4,000 in silk weaving. The lives of even the most privileged of these workers were fraught with insecurity and hardship. But for some, life in the textile industries offered opportunities that were not to be found in the countryside, especially as earnings from handicrafts began to decline after the First World War.55 Cotton mills sorted workers into different jobs according to gender and place of origin, as well as skill and experience. Before the 1920s, about half of all Shanghai cotton workers were men. A series of strikes between 1922 and 1925, however, convinced mill owners that a female workforce would be more docile and stable. The process of replacing male workers with women was difficult and fraught with conflict, but by 1929, women comprised the majority of Shanghai textile workers. In Tianjin, three-quarters of textile workers were still men in 1929. But from the mid-1930s, mill owners began to hire more women with an eye to reducing costs. In general, men worked in machine repair and maintenance, as well as in jobs requiring physical strength. Women comprised about 80 per cent of the spinning and weaving workforce, while children under 13 made up only 3 per cent. The economic depression of the early 1930s, however, led mill owners in both Shanghai and Tianjin to hire women to perform more of the tasks that had previously been done by men such as packing, blowing, stripping and even carrying heavy loads. While wages generally decreased for women cotton workers during the depression, those men still employed in the cotton mills actually saw their wages increase. Furthermore, women were mainly paid piece-rates, while men monopolized the time rate jobs. Thus, as the gender labelling of certain jobs changed so did the designation of those jobs as ‘skilled’ or ‘unskilled’, as well as the rate of pay associated with such work.

55 We are fortunate to have some excellent books on the lives of modern Chinese textile workers. The following account is based mainly on Emily Honig, Sisters and Strangers: Women in the Shanghai Cotton Mills, 1919–1949 (Stanford, 1986); Gail Hershatter, The Workers of Tianjin, 1900–1949 (Stanford, 1986); Elizabeth Perry, Shanghai on Strike: The Politics of Chinese Labor (Stanford, 1993); also Köll, From Cotton Mill to Business Empire, ch. 4. 120

China The silk filature workforce was 95 per cent female, with men employed as boilermakers, machinists and supervisors. Filature workers were also on average much younger than cotton mill workers. In 1929, about 24 per cent of Shanghai’s filature workers were children, and the proportion was even higher in provincial filatures. In contrast with Japan and Europe, Chinese silk filatures hired girls as young as 8 to perform the painful task of beating the boiling cocoons to release the thread. This work caused their hands to come in frequent contact with boiling water, and sometimes led to fatal or disfiguring accidents. A foreign observer described their condition, stating: Tiny children stood for an eleven-hour day, soaked to the skin in a steamy atmosphere hot even in winter, their fingers blanched to the knuckles and their little bodies swaying from one tired foot to the other, kept at their task by a firm overseer who did not hesitate to beat those whose attention wandered.56 The number of young girls working in steam filatures gradually declined during the Second World War, however. In contrast, silk weaving was dominated by adult men, who were very much a ‘labour aristocracy’ – well paid, relatively independent and possessing valued skills. As silk weaving became mechanized and proletarianized in the twentieth century, however, weaving factories hired more women. At first employed mainly in spooling and other low-paying prep work, women also began to return to commercial silk weaving, reversing almost a thousand years of masculinization. By the 1950s, women comprised about half of those employed in Shanghai’s silk weaving industry, although most were still employed in preparatory work rather than weaving. The earliest textile workers were hired from among the residents of the treaty ports, or from among the local farmers in the case of provincial silk filatures and Zhang Jian’s Dasheng mill.57 As the textile industries expanded in the 1920s, however, an increasing proportion of the workforce was recruited from a greater distance. In Shanghai’s cotton mills, for example, there were workers from three main regions, each with their own language, customs and social status. Women hired from the suburban areas in which the first mills were built usually had a background in handicrafts. By the 1930s, as earnings from handicrafts production dwindled to half of a mill worker’s wages, nearly all these women went to work in the mills.

56 Eleanor Hinder, Social and Industrial Problems of Shanghai, with Special Reference to the Administrative and Regulatory Work of the Shanghai Municipal Council (New York, 1942), p. 18. 57 Köll emphasizes that the mill’s unskilled workers were hired from the countryside surrounding the rural town of Tangjiazha, where the mill was located, while skilled workers were brought in from other parts of the Jiangnan region such as Ningbo, Wuxi, Shaoxing, Shanghai and Nanjing. Köll, From Cotton Mill to Business Empire, pp. 81–92. 121

The Ashgate Companion to the History of Textile Workers Over time, however, mill owners hired workers from more distant rural areas. In Shanghai, such workers came from two main regions, Jiangnan and Subei. The decline of handicrafts forced many Jiangnan women to look for work in the cities. Leaving their husbands to farm, these women soon found that they could earn enough to contribute half or more of their families’ incomes. Subei refers to the northern part of Jiangsu Province, south of the Huai River. Subei women spoke a different dialect, and generally possessed fewer handicraft skills than their Jiangnan counterparts. Whereas Jiangnan women traditionally produced silk and cotton, Subei women worked in agriculture alongside their husbands, and were considered rough, coarse and dirty by their southern counterparts. Furthermore, while women from rural Jiangnan might leave their families to find work in the urban mills, most Subei women migrated with their entire families, husband and wife seeking a new life in the city together. Fleeing floods and warfare, Subei women joined Shanghai’s textile workforce from the early 1920s. Looked down upon by their Shanghai and Jiangnan bosses and co-workers, they were generally assigned the dirtiest and most backbreaking work. In contrast, most of the machine weavers, among the highest-paid jobs, were women from Shanghai or Jiangnan. Some Subei women preferred to work in Japanese mills where they would not face this kind of prejudice, which further lowered the reputation of Subei workers in the eyes of the women from Jiangnan and Shanghai. The relationships among workers from different parts of China were often characterized by prejudice, contempt and hostility.58 Tianjin workers were similarly divided, with those from Tangshan playing the role of despised outsider, similar to Subei people in Shanghai.59 We see a similar pattern in the silk reeling industry. The earliest filatures in rural South China hired young women from the surrounding villages. The situation was similar in the Jiangnan region, where filature owners sought to take advantage of a cheap, flexible supply of local labour. Wages in Wuxi were generally lower than in Shanghai, except in 1929 when a strike forced the filature operators to increase wages to Shanghai levels. Provincial capitalists’ efforts to maintain seasonal flexibility in their workforce prevented these rural workers from improving their skills to the level of Shanghai filature workers.60 As in cotton textiles, filatures only hired Subei workers for the most difficult, unpleasant and poorly paid jobs such as peeling, selecting and beating the cocoons. Silk weavers came almost exclusively from the Jiangnan region. Two types of silk weavers could be found in Shanghai workshops in the 1920s. On the one hand, there were older artisans – household weavers forced into factory jobs by declining 58 See Honig, Sisters and Strangers, pp. 57–78; also Emily Honig, ‘Burning Incense, Pledging Sisterhood: Communities of Women Workers in the Shanghai Cotton Mills, 1919–1949’, Signs: Journal of Women in Culture and Society 10 (4) (1985): 700–714. 59 Hershatter, The Workers of Tianjin, pp. 144–150 recounts that when Tianjin cotton mills wanted to hire more women, they initially encountered difficulties attracting local women, and instead recruited experienced spinners from Shanghai. 60 Bell, One Industry, pp. 102–104. 122

China wages in handicrafts. Hailing from Zhejiang weaving towns, these men were usually poorly educated and mostly illiterate. On the other hand, the more modern silk factories, such as the Meiya Company, established in 1920 with electric-powered looms, hired younger men, and some women. These workers usually had at least a primary school education. They were highly literate (95 per cent of the men and 35 per cent of the women could read) and received training as apprentices by the factories that hired them. They looked upon factory life as a step up in the world, and modelled their dress, customs and politics on urban students their own age. Older artisans often viewed factory work as a loss of independence and were more deeply rooted in traditional, rural culture.61 As we shall see, these groups produced very different forms of labour organization and action in the 1920s and 1930s. There were three main paths to employment in the Chinese textile industries – making contact with a foreman through a relative or acquaintance, recruitment by labour contractors or recruitment and training through a mill’s own apprentice training system (yangchenggong). When most girls first went out to work, they followed their mothers or older sisters to work at the same mill, learning the skills and making contacts through their relatives. Many girls who had worked in silk reeling from the age of 8 or 9 went to work in the cotton mills when they reached their early teens, seeking higher pay and better working conditions. Most cotton mill workers found employment by making contact with a ‘Number One’. First appearing in British mills in Shanghai, the Number Ones were women, generally from Shanghai, Wuxi or Changzhou, who had come up through the ranks to achieve the most privileged and powerful position available to a woman worker. Based as it was on personal ties reinforced by ritual and mutual obligation, the Number One’s relationship with the workers, while exploitative, was also very complex. The Number Ones collected a portion of the workers’ wages, and thus had a strong interest in maintaining their monopoly on hiring. They also commanded such authority that they could mobilize thousands of workers to oppose mill owners’ efforts to dislodge them.62 As the supervisors controlled nearly every aspect of life in the mills, it was important to maintain the relationship through gifts on birthdays and holidays. Some workers even pledged loyalty to their Number One as a ‘godmother’ in a simple ritual. As a ‘godchild’, the worker could expect full protection and favours, but was responsible for sending appropriate gifts. This arrangement was also common in Shanghai silk filatures. However, as mentioned above, Jiangnan and South China filature workers came from the surrounding countryside, and employers were able to exercise more direct control over the hiring process. 61 Perry, Shanghai on Strike, pp. 185–188; on the Meiya Company, see Xu Xinwu (ed.), Jindai Jiangnan sizhi gongye shi [A History of Modern Jiangnan Silk Weaving Industry] (Shanghai, 1991). 62 Honig, Sisters and Strangers, pp. 79–87. Köll describes the early twentieth-century baogongzhi system in the Dasheng Cotton Mill, in which gongtou, male and female foremen, contracted to produce a certain amount of thread, and were responsible for hiring unskilled workers and maintaining the mill’s strict disciplinary rules. Köll, From Cotton Mill to Business Empire, pp. 95–105. 123

The Ashgate Companion to the History of Textile Workers Many mill hands joined the textile workforce through labour contractors.63 The business of buying and selling young women, known as ‘plucking mulberry leaves’, was often controlled by the criminal underworld, in Shanghai centred on the notorious ‘Green Gang’ (Qingbang in Chinese). The contractors only operated in certain regions, relatively distant from Shanghai, where there were desperate families willing to sell their daughters, at least temporarily. These areas also seem to correspond to early centres of Green Gang activity. The families received an ‘allowance’, negotiated at the signing of the contract and usually paid out in three increasing instalments. A fee of 30 yuan for three years’ work, as well as food, clothing and housing for the worker, was common, although higher fees might be offered for older or more experienced women. After bringing the contracted women to the city, the contractor would then seek employment for them by either approaching the mill’s hiring office directly or working through one or more of the Number Ones. Those for whom work could not be found might end up working as servants in the contractor’s home or dormitory. Some might be sent to work in restaurants or teahouses, while others ended up as prostitutes or worse. Although contract workers were treated much as other workers inside the mills, their lives outside the mills were quite different. They were entirely under the control of the men who held their contracts. Not allowed to keep their own wages, they had to deliver their pay to their bosses. Contract workers lived in spare, crowded rooms and enjoyed no possessions to speak of. Two workers, one on the night shift and one on the day shift, would share the same sleeping space. The contract guaranteed that the workers would be provided with meals. However, with an eye to their profit margins, the contractors provided no more food than was necessary for subsistence. As a result of these conditions, many contract workers became ill. Most would try to keep working, however, as any days missed would have to be made up double or triple in order to complete their contract. Escape was difficult, and even if successful, the consequences for both the girl and her family could be disastrous. The family would have to repay the contract fee, as well as other expenses, and the girl would most likely be blacklisted and prevented from finding work in another mill.64 Mill owners recognized that the contract labour system provided them with weak, tired and apathetic workers. Social outcry against the conditions under which contract workers lived, which not only affected their health but also made them vulnerable to sexual and physical abuse, also encouraged the mills to find alternatives. However, it proved extremely difficult to overcome the combination of Number One patronage and contract labour that dominated hiring in the cotton industry. In almost all of the industrial cities, both of these paths to employment were linked to organized crime. The mill owners were themselves partially to blame for the domination of Shanghai’s labour market by criminal gangs. Fearful of the militant, Communist63 It is not clear how common labour contractors were in the silk reeling industry, but they were very prominent in cotton textiles. 64 Honig, Sisters and Strangers, pp. 96–111. 124

China led labour movement of the 1920s, the mill owners acquiesced in Chiang Kai-shek’s ‘White Terror’ of April 1927. Chiang, with the support of Shanghai’s capitalists, turned on his erstwhile working-class allies, using his contacts in the Green Gang to round up and murder thousands of labour leaders and Communists. With violence, extortion and government complicity, the Green Gang gradually solidified its control over Shanghai’s labour markets during the 1930s. Green Gang leaders sometimes supported strikes by the unions they controlled, if their members stood to gain. Number Ones were often gang members, and as they received a portion of their workers’ wages, they had an incentive to support strikes in pursuit of wage increases. The Green Gang both suppressed and, at times, cooperated with more radical elements in the labour movement, demonstrating the canny pragmatism of criminals everywhere. Through threats and violence, they also protected the rights of Number Ones and labour contractors to decide hiring issues.65 Employers naturally sought to achieve greater control over hiring practices, but there was little they could do in the face of such lawlessness. As early as the 1920s, Japanese cotton mills in Shanghai had attempted to institute an apprentice training system, which came to be known as ‘yangchenggong’. For the most part, these initiatives were successfully resisted by the Number Ones and contract bosses, who mobilized the workers under their control to defend the existing system. It was not until after the Second World War that cotton mill owners were able to make significant progress towards regularizing hiring practices. The apprentice system was most successful in the large, government-run mills taken over from Japanese owners after the war. Admission to the apprenticeship programme was based on an examination, including a literacy test. Apprentices were trained through a combination of hands-on experience and classroom instruction. After two months’ training the mill administered a test. If the worker passed, she would become an apprentice working full time. After two years, or sometimes less, she would become a worker earning full wages. However, the contract system and the power of the Number Ones persisted in many private mills even after the Communists came to power in 1949.66

Working life Chinese textiles workers exhibited a wide range of incomes and standards of living, with young, female filature workers and cotton mill hands receiving the lowest pay, and male, adult silk weavers enjoying the pay and privileges of ‘labour aristocrats’. Whether they lived in their own homes or in rooms owned by contract bosses, most mill hands worked 11 or 12 hours, alternating between the day and night shifts 65 Ibid., pp. 120–131; Perry, Shanghai on Strike, pp. 92–103; Bush, Politics of Cotton Textiles, pp. 300–306. 66 Honig, Sisters and Strangers, pp. 87–93. 125

The Ashgate Companion to the History of Textile Workers every two weeks. They had to dress smartly and keep themselves pretty in order to satisfy managers’ expectations. If they had families to care for, they might also have to spend time fetching water and preparing meals. What time was left was spent sleeping. The weekly rest day was usually devoted to laundry and bathing, but some workers also found time to visit friends, shop or enjoy a performance. The working lives of these women ranged from tolerable to miserable. The Chinese and British cotton mills were described as darker and dirtier than the Japanese mills, but all cotton mills were noisy places, filled with dust and fluff. Working long hours on day or night shifts, thread workers had to pay close attention to the spinning frames, repairing breaks as soon as they appeared. An experienced spinner could repair a break every ten seconds. Workers might be responsible for as many as 25 to 30 staffs of six spindles each. With only a half hour for lunch to rest, workers collaborated with their Number Ones to slack off when managers were absent, known as ‘soaking mushrooms’. Otherwise, a young trainee could take over for a spell. Sitting or squatting to alleviate the strain, turning off the machine on some pretext or going to the bathroom were ways to gain a respite. Use of the toilet was carefully restricted and required permission from the supervisor. In addition to the usual surveillance and supervision, Tianjin mills were also highly militarized, with a strong police or army presence.67 Nearly all cotton mills imposed bodily searches on workers leaving the factory. The mill’s exit was set up as a maze-like barricade that forced workers to pass slowly through a narrow area where they could be searched for cotton, thread, cloth or bobbins. There might be a box with a sign reading, ‘Have you something in your pocket? Ask your heart. You do not want to steal. If the answer is Yes, put it at once into this box.’68 Finally, on the way home from work in the evening, women mill hands faced harassment from male workers, as well as the more serious threat of neighbourhood gangs, whose predations ranged from sexual harassment, theft and extortion to rape and kidnapping. Women in Chinese silk filatures worked similar hours, but due to their youth, usually had fewer responsibilities outside of work. As described above, some filature work was unpleasant and painful, and required careful attention. A silk reeler had to repair breaks in the filaments quickly to maintain a constant thickness, and watch the basins to remove unravelled cocoons before the chrysalis could stain the water. Male supervisors tended to be much more authoritarian and less prone to collaborate with workers who wanted to take a few minutes’ rest. In contrast, silk weavers were relatively independent. However, they also worked long hours, and their work was often complex and painstaking. The best weavers, who earned the highest wages, paid careful attention to their looms, watching for defects or malfunctions in order to produce the highest quality silk.

67 Hershatter concludes that this constant threat of violence was one factor that made Tianjin workers less rebellious than their Shanghai counterparts. 68 Honig, Sisters and Strangers, p. 148. 126

China Although the working conditions were often harsh, the wages that textile workers could earn were fairly attractive compared with other forms of employment. The average daily wage for a woman worker in a Shanghai cotton mill in the early 1930s was 0.45 yuan, and for a power-loom weaver it was 0.56 yuan. Before 1946, powerloom weavers in the cotton mills received somewhat higher pay than workers in the roving and spinning departments. After the war, however, weavers earned slightly less than spinners. Silk reelers and weavers occupied opposite ends of the wage spectrum. Between 1930 and 1934, Shanghai filature workers earned an average of 0.41 yuan for an 11-hour day. Silk weavers earned more than twice as much, about 1.08 yuan per day.69 With these wages, women who worked in textiles, even young girls in the silk filatures, provided a major contribution to their families’ economies. In some cases, a single worker might support her entire family. Some women mill hands were able to achieve a degree of independence unheard of in rural China, even leaving home to move in with other young mill workers. Such younger, unmarried women might wear fancy coiffures or embroidered shoes, and could afford to go shopping or to see an opera with friends. For some filature workers in Guangdong, employment and wages provided the means to avoid married life, some even purchasing a concubine to take over their wifely duties. Other filature workers ‘cut their own hair’ (zishufa), refusing to marry at all. These joined a sisterhood of silk workers, similar to religious communities of Buddhist nuns, who dedicated themselves to the silk, enjoying more pride and independence than most could have found in a rural marriage.70 Although handicraft textiles declined in Jiangnan from the end of the nineteenth century, such work provided needed income to many rural families in North China in the early twentieth century. In the predominant ‘putting-out’ system of the 1920s, the prices of raw materials and finished products were fixed in advance. The producers, both men and women, earned a fixed piece-rate based on the market price of cloth rather than the labour market. Household weavers were apparently able to make a decent living for themselves and their families, relative to earnings from other cottage industries. In the early twentieth century, cotton weavers in Gaoyang, in Hebei Province, could earn several times the annual wage of an agricultural labourer. In 1932, poorer households that devoted more than 1,000 hours to textile production annually could earn 33–35 yuan per year, while average earnings were around 22.5 yuan per year. This was not much, but every yuan earned contributed to the family economy. Income from household weaving gradually declined, however. By 1935, weavers in Baodi, Hebei Province could only earn enough from weaving to meet one-third of their subsistence needs.71 69 Ibid., pp. 176–177; Perry, Shanghai on Strike, p. 191. 70 Marjorie Topley, ‘Marriage Resistance in Rural Kwangtung’, in Margery Wolf and Roxane Witke (eds), Women in Chinese Society (Stanford, 1975); So, The South China Silk District, pp. 122–134. For a very moving fictional account, see Gail Tsukiyama, Women of the Silk (New York, 1991). 71 Zhao and Chen, Zhongguo mianyeshi, pp. 205–207, 212–213, 227–229. 127

The Ashgate Companion to the History of Textile Workers Worker turnover was fairly high in the textile industries, at least before the Great Depression. Most cotton workers stayed less than three years at a given mill, and some changed jobs after less than a year. Wages varied widely from mill to mill, even within the same company. Other considerations such as better working conditions, superior equipment or better raw materials motivated workers to seek out work at some mills and avoid others. While workers avoided the Shenxin Number Nine Mill, the Yong’an mills enjoyed a reputation as a good place to work. Turnover was even higher in the silk reeling industry, exacerbated by the seasonal nature of the work. Mill work usually ended, at least temporarily, with marriage and childbirth. In the more progressive mills, such as Yuyuan in Tianjin (which also provided dining halls, schools, a cooperative shop, bathhouse, athletic field, a martial arts society and melons in the summer), women were granted paid maternity leave.72 In 1946, Shanghai workers nominally won the right to 56 days maternity leave. But in the 1930s, pregnant women were usually fired. Most cotton mill owners did not want old women, women with bound feet or pregnant women to work in their mills. Efforts to hide pregnancies often resulted in miscarriages and stillbirths. Shanghai’s Yong’an was more progressive in retaining pregnant women, and allowing them a few days off to give birth. It is likely that only about 30–40 per cent of these children survived, however.73 Like proletarians everywhere, Chinese textile workers organized and mobilized in a variety of ways to protect their interests. In the 1920s and 30s, silk reelers and cotton mill workers were more likely to launch spontaneous protests in response to specific incidents or simply to gain some relief from the tedium of factory life. Such workers were sometimes mobilized for strikes and protests by their contractors, supervisors or the criminal elements which came to dominate the Shanghai labour movement. When filature hands organized the Shanghai Silk and Cotton Women Workers’ Association in 1924, they chose a woman with strong gangland connections named Mu Zhiying to lead it. Although the Association was able to extract some concessions, the Cocoon Guild soon co-opted the corruptible Mu, turning the workers’ association into a control mechanism to ensure that strikes did not take place.74 Silk weavers were often better organized and more independent than their thread mill counterparts. As mentioned above, nineteenth-century handicraft weavers possessed strong guild organizations, which, while unable to preserve their artisan lifestyle, were very active around the turn of the century. During the height of the labour movement in 1926–1927, far more cotton workers participated in strikes than did silk weavers. However, male silk weavers had far more strikes per capita. Perry records that, between 1918 and 1940, silk weavers launched 20.8 strikes per 1,000 workers, as compared with just 2.7 per 1,000 cotton mill hands and 72 Hershatter, The Workers of Tianjin, p. 165. 73 Honig, Sisters and Strangers, pp. 178–193. 74 Perry, Shanghai on Strike, pp. 170–176; Xu Xinyu (ed.), Zhongguo jindai saosi gongye shi [A History of China’s Modern Silk Reeling Industry] (Shanghai, 1990), pp. 583–589. 128

China 2.0 per 1,000 filature workers. However, cotton strikes involved an average of 2,286 participants, while silk weavers’ strikes averaged only 212 participants.75 Gender and age also influenced strike activity. Male cotton workers were more likely to join large-scale sworn brotherhoods, or even criminal gangs, than were women workers. The two types of twentieth-century silk weavers described above exhibited different preferences for organizations and politics. The younger, more educated weavers organized modern-style unions, often with a radical political content inspired by the students they emulated in their dress and habits. Older craftsmen preferred to establish more traditional brotherhoods, and were more attracted to secret societies and criminal gangs.76 Female cotton and silk workers throughout China organized ‘sisterhoods’ (jiemeihui) for mutual support and protection. Small groups of women, perhaps six to ten, usually from the same locality, swore loyalty to each other, sometimes with a simple banquet or a visit to a Buddhist temple to light incense together. ‘Sisters’ protected each other from abuse by hoodlums and supervisors, and pooled resources for important events such as weddings and funerals. Often a group of younger women would pledge loyalty to an older worker or Number One, who might also be affiliated with a criminal gang. Such informal groups could be just as powerful in mobilizing workers to strike as more formal unions. If one ‘sister’ turned off her machine, all of the others would follow.77 The role of the Chinese Communist Party in the Shanghai labour movement is difficult to assess. In the 1920s, radical politics were more popular among male machinists and artisans than among textile workers. It was not until 1934 that underground party activists,78 attracted by the Meiya silk workers’ classconsciousness and militancy, began to influence the silk weavers’ organizations. One notable aspect of Shanghai silk weavers’ strikes in the mid-1930s was active participation by women weavers. Whereas the artisan weavers excluded women from their ‘brotherhoods’, the younger, more radically minded workers included demands such as equal pay for men and women in their strike demands.79 As the global depression began to take its toll on the industry, factory owners and the government intensified their suppression of the silk weavers. Meiya employees lost many of their former privileges, and the gap between the large, modern factories and the smaller workshops diminished. The citywide silk weavers’ 75 Perry, Shanghai on Strike, p. 168. 76 Perry provides an excellent analysis of the varied political activities of Shanghai workers based on gender, occupation and place of origin. 77 Honig, Sisters and Strangers, pp. 202–217. 78 Open organizing by Communists in Shanghai ended with the April 1927 ‘White Terror’ in which thousands of Communists and labour organizers who had cooperated with the Nationalists were murdered by their erstwhile allies in collusion with criminal gangs. On the underground party organization after 1927, see Patricia Stranahan, Underground: The Shanghai Communist Party and the Politics of Survival, 1927–1937 (New York, 1998). 79 D.K. Lieu, The Silk Industry of China (Shanghai, 1941), pp. 233–239; Perry, Shanghai on Strike, p. 195. 129

The Ashgate Companion to the History of Textile Workers strikes in 1936 and 1937 were less bold in their demands, involved far fewer women (except at Meiya) and were organized mainly by criminal elements and Nationalist Party agents. Later strikes were often violent, but rarely successful. It was not until after the war that Communists again came to play a leading role in the silk weavers unions. In a context of civil war and economic crisis, theirs was a radical voice calling for political change and economic responsibility. Wartime repression and post-war inflation eliminated weavers’ privileged position, leading to greater solidarity with other workers. Silk weavers continued to lead the economic and political struggle in late 1940s Shanghai. Communist-influenced weavers were at the forefront of protests to end the wage freeze of 1947, and played a leading role in welcoming and facilitating the Communist takeover of Shanghai in May 1949. In contrast, the Communists initially had little success in organizing women textile workers. This was mainly because the party had very few women activists, and made little effort to unite organizing work with the day-to-day issues that concerned women mill hands. In Shanghai, most women workers’ first introduction to radical politics was through YWCA educational initiatives. From 1928, the YWCA began to train women workers to become labour activists in an effort ‘to instill in women a radical understanding of their position as women and as workers’.80 These women actively educated their fellow workers, and increasingly took a leading role in strike actions. Beginning in the refugee camps during the Second World War, the party made inroads among textile workers, primed to receive radical politics by their YWCA education. Through this contact, underground labour organizers developed a greater sensitivity to the needs and values of women workers. As a result, the party came to view the textile workers’ traditional sisterhoods, even those associated with criminal gangs, as an effective means of organizing and educating women workers. Women party members sought to organize ‘red’ sisterhoods under the cover of traditional forms and rituals. By means of these tactics, the party carried out education work, organized resistance to the Japanese occupation, and mobilized urban workers to support the party’s rural army. After the war, the party maintained these tactics, organizing textile workers to demand a living wage, maternity leave and an end to the government’s disastrous economic policies. For the first time, women labour leaders and issues pertaining specifically to women workers came to the forefront of the Communist-led labour movement.81

80 Quoted in Honig, Sisters and Strangers, p. 218. 81 Ibid., pp. 224–243. On gender conflict, and the persistence of the party’s male orientation in the 1948 Shenxin Number Nine Cotton Mill strike, see Perry, Shanghai on Strike, pp. 125, 213–214. 130

China

Textiles in the People’s Republic of China Japan’s 1937 invasion of China aggravated the already violent and unstable situation there. Agriculture and sericulture declined dramatically, and some 90 per cent of textile production in cities like Shanghai was destroyed outright. The early years of the Japanese occupation saw something of a revival in production, profits and employment. However, after 1941, Chinese textiles came to be dominated by Japanese capital and occupation monopolies such as the Central China Sericulture Company. Due to the ravages of war and the exploitative policies of the occupiers, textiles production fell by one-half or more by the end of the war. In many rural areas, iron gear looms were melted down for use as tools or arms. However, in Communist-held areas, such as the Shaan-Gan-Ning Border Region, there was an impressive revival of rural handicraft based mainly on household and cooperative forms of organization. Under the Communists’ production drive, household production of cotton cloth increased from about 14,000 bolts in 1941 to more than 45,000 bolts in 1943.82 The establishment of state monopolies, and the seizure of the most advanced Chinese firms by the occupation government, enabled the post-war Nationalist government to achieve control over a sizeable share of the urban textile industries through the China Textile Company and the China Sericulture Company. As a result, by 1947, the Chinese government controlled almost 40 per cent of China’s cotton spindles and nearly 60 per cent of its power-looms, producing 43.7 per cent of the yarn and 72.6 per cent of the cotton cloth in China.83 Although fraught with problems, the Nationalist Party’s administration of these mills introduced some progressive measures including labour protection, a rational hiring system and welfare benefits. Cotton cultivation had declined dramatically during the war. But through an agreement with the American government, China was able to import large quantities of cheap cotton. Shanghai’s thread mills relied on US cotton for 90 per cent of supplies. Even with US imports, however, by 1947 there was a severe shortage of cotton, as well as unstable electricity supplies. Between the pressures of short supplies and rampant post-war inflation, as well as competition from government mills, 12 of Shanghai’s 55 cotton mills were forced out of business by 1949. The end of the war also saw a revival of small-scale handicrafts textiles workshops, based on a new generation of rural entrepreneurs, which lasted into the 1950s. The post-war years also witnessed the growth of urban workshops. In 1936, only about 50,000 cotton weaving looms (about half of them power-looms) were concentrated in urban factories. By 1947, weaving workshops operated more than 130,000 looms, and produced 56 per cent more cloth than 10 years earlier. By 1949, weaving workshops in Shanghai employed more than 60,000 workers.

82 Mark Selden, The Yenan Way in Revolutionary China (Cambridge, 1971), pp. 256–260. 83 Albert Feuerwerker, Economic Trends in the Republic of China, 1912–1949 (Ann Arbor, 1977) pp. 24–25. 131

The Ashgate Companion to the History of Textile Workers However, about two-thirds of production in Shanghai, and more than threequarters in Tianjin, still relied on foot or hand-powered looms.84 Several factors combined to threaten China’s silk industry in the post-war years. Aside from rampant inflation and economic chaos (which increased production costs and inhibited the growth of exports), the government’s efforts to supply filatures with capital and cocoons proved counterproductive. Inflation quickly outstripped government loans for cocoon purchases, leaving the filatures in debt and short of cocoons. Efforts by the China Sericulture Company to gain control of the trade in raw silk resulted in prices that were too low to allow the silk filatures to continue production. Behaving much like other speculators, government agencies and banks hoarded silk thread, aggravating weaving factories’ high production costs. Although the Communists seized power on the basis of a military-agrarian movement, the party’s ideology envisioned a partnership with the urban proletariat. With the establishment of the People’s Republic in October 1949, the party faced numerous social and economic problems. Inflation and unemployment worsened in the first year of Communist rule. The Nationalist blockade stifled exports, and air raids knocked out Shanghai’s main power plant on 6 February 1950, affecting the livelihoods of thousands of workers. However, by 1953, the new government had made impressive progress in restoring production and controlling inflation. Supplies recovered and employment increased as industry and trade revived. The pre-war cotton harvest peaked at about 1.3 million tons in 1936. Production of cotton thread peaked at almost 500,000 tons in 1933 and cotton cloth at 2.8 billion metres in 1936. Although by 1949, production had fallen by about one-third, by 1952, the cotton harvest had recovered to the pre-war peak, and production of cotton thread and cloth had reached 656,000 tons and 3.8 billion metres respectively. Sericulture and silk production were slower to recover. By 1957, silk production was still less than half the pre-war peak, and it would not reach that level again until the early 1970s.85 The party’s policy towards loyal Chinese businessmen was ‘utilize, restrict and transform’. The party recognized that it needed to utilize the capitalists’ skills and authority in order to restore production. However, the Communists also hoped (eventually) to incorporate private factories into a state-managed, socialist planned economy. The post-1949 government’s early efforts to control raw materials supplies and sales channels served the twin purposes of restoring economic health and restricting the scope of capitalist enterprise. The party also led the establishment of labour-capital consultative conferences in private factories, which included representatives from among the employees, as well as from the party and unions. 84 Shanghai Textile Industry Gazetteer Editorial Committee (ed.), Shanghai fangzhi gongye zhi [Shanghai Textile Industry Gazetteer] (Shanghai, 1998), pp. 61–64, 570; Xu Dixin and Wu Chengming, Zhongguo zibenzhuyi fazhan shi, disan juan, xinminzhuzhuyi geming shiqi de zhongguo zibenzhuyi [History of the Development of Chinese Capitalism, Vol. III: Chinese Capitalism during the New Democratic Revolution] (Beijing, 2003), pp. 663–669. 85 Wang Zhuangmu, Sichou biji [Silk Notes] (Beijing, 2001), pp. 1–14. 132

China In the end, however, unable to control capitalists’ illegal activities, stimulated as they were by the dual nature of the 1950s economy, the party moved quickly to merge private textile mills with the state administration. By 1956, this had resulted in the nationalization of the country’s entire textile industry.86 Despite the growth of weaving factories in the post-war period, by 1949 half of all cotton cloth in China was still produced by household handicraft weavers. Although the party had promoted handicraft weaving in the revolutionary base areas during the war, its industrial policy strongly favoured modern, large-scale, mechanized production. Handicraft weaving was viewed as at best ‘backward’ and at worst ‘petite bourgeois’. Furthermore, the newly resurgent rural textiles entrepreneurs of North China were successfully competing with state firms. Over the course of the 1950s, as markets for major commodities came under state control, the proportion of cloth produced by household weavers declined rapidly. In 1954, the new government established a state monopoly in cotton textiles, shutting down the small rural workshops and banning household weaving.87 By 1956, more than 90 per cent of handicraft textile production was concentrated in cooperative weaving workshops participating in something resembling a socialist ‘putting-out’ system. Weaving cooperatives received thread from the state and produced cloth in return for what amounted to a piece-rate wage.88 Under Communist leadership, China’s cotton industry both expanded and underwent a technological revolution. During China’s First Five-Year Plan (1953– 1957), investment in cotton mills shifted away from the coastal cities to establish factories in inland cotton-growing regions such as Hebei and Henan Provinces. Between 1952 and 1956, the proportion of yarn produced in the coastal cities (Shanghai, Tianjin and Qingdao) fell from 82 to 72 per cent, and cloth production fell from 88 to 80 per cent.89 Although cotton cultivation declined during the disastrous ‘Great Leap Forward’ (1958–1961) leaving much of the new plant idle, production recovered quickly, increasing 50 per cent in 1963 and 37 per cent in 1964. By 1965 the textile industry was operating at full capacity. That year saw the construction of 37 new cotton mills. Spinning capacity increased by 1.4 million spindles in 1959 and by the same amount again in 1965.90

86 For an informative discussion of the process of socialization and the relationship between textile capitalists and the Communist Party, see Kathryn Sears, ‘Shanghai’s Textile Capitalists and the State: The Nationalization Process in China’ (unpublished Ph.D., University of Michigan, 1985). 87 Grove, Chinese Economic Revolution, pp. 182–185. 88 Statistics reveal, however, that handicraft production declined rapidly with cooperativization, from 311 million metres in 1954 to just 67 million metres by 1957. Kang Chao, The Rate and Pattern of Industrial Growth in Communist China (Ann Arbor, 1965) p. 141. 89 Nai-Ruenn Chen and Walter Galenson, The Chinese Economy under Communism (Chicago, 1969), p. 73. 90 Alexander Eckstein, Walter Galenson, and Ta-Ching Liu (eds), Economic Trends in Communist China (Chicago, 1968), pp. 584, 589. 133

The Ashgate Companion to the History of Textile Workers This growth was made possible by the rapid expansion of China’s textile machine industry. Whereas China had previously imported most of its textiles machinery, by the early 1960s production was increasing rapidly, and China had become a net exporter of textiles machinery as well as cotton cloth.91 By that time, almost all cotton cloth in China was produced with mechanized looms, and handicraft weaving had all but disappeared. By 1979, more than 80 per cent of cotton looms were fully automatic, greatly increasing productivity. Although the number of cotton spindles only doubled between 1949 and 1979, thread production increased by 6.5 times, and almost half of China’s cotton spindles were located in inland provinces. Reflecting the shift from handicraft to machine weaving, cotton cloth production increased almost fivefold.92 Textile production increased slowly relative to producer goods industries such as steel and machinery, reflecting the Communist Party’s emphasis on heavy industry at the expense of consumer goods. Nonetheless, average annual per capita consumption of cotton cloth increased from 4.3 metres in 1936 to 7.2 metres in 1955, although consumption of cotton cloth quickly declined thereafter. At the same time, China became a net exporter of textiles, which helped to fund imports of capital equipment for the development of heavy industry.93 In even more dire circumstances by 1949, the silk industry also witnessed an impressive revival followed by mergers and nationalization. With price stability, the recovery of sericulture and state prices for raw silk sufficient to maintain filatures’ operations, the silk reeling industry slowly recovered. With exports restricted by the US embargo, most of this thread went to supply the domestic weaving industry. By 1953, the majority of silk weaving factories were engaged solely in contract production for the state. Under this system, the China Sericulture Company (under new management) supplied weaving factories with thread, and purchased their products at a guaranteed price. Although this policy brought the industry back from death’s door, private silk weaving became entirely dependent on the state. Between 1954 and 1956, these factories entered into joint state–private mergers resulting in the nationalization of the entire industry. While the cotton industry enjoyed a stable domestic market, exports to the Soviet Union and Eastern Europe were crucial to the restoration of silk weaving. Paying prices three or four times higher than world markets, socialist countries imported large amounts of Chinese silk as part of an effort to support China’s economic recovery and industrial development. Although the Soviets exported much of this in exchange for hard currency, it is unlikely that China could have obtained similar material and technical support from any capitalist country at that time. During the Korean War, silk comprised more than 90 per cent of China’s total textile exports, and this luxury cloth remained an important export product through to the end 91 Alexander Eckstein, China’s Economic Development: The Interplay of Security and Ideology (Ann Arbor, 1975), p. 167. 92 Xin zhonguo fangzhi gongye sanshi nian [Thirty Years of New China’s Textile industry] vol. I (Beijing, 1980), pp. 151–155, 300. 93 Chu-Yuan Cheng, Communist China’s Economy, 1949–1962: Structural Changes and Crises (Seton Hall, 1963), p. 160. 134

China of the decade. Silk workers were motivated to produce more, higher quality cloth with the reminder that two bolts of silk could import a ton of Soviet steel for the nation’s economic development. Foreign exchange earnings from cotton exports increased nine times between 1953 and 1978. Over the course of the 1950s, Chinese industrial workers gained impressive benefits such as an unprecedented degree of job security, improved working conditions and labour relations, labour insurance, welfare and education. As Gail Hershatter puts it: When the CCP [Chinese Communist Party] took power in 1949, it was able to deliver what the [Nationalist Party] had promised but could not provide: stable jobs, a wide range of welfare facilities, safety from daily violence in the workplace, and cooperation between workers and those who controlled production … At last, workers had the patronage, protection, and stability that had been so elusive in pre-revolutionary Tianjin.94 Not all workers enjoyed such benefits, however. During the 1960s, a strong division developed between permanent workers, who enjoyed all of the benefits of state employment, and temporary or contracted workers who did not. This division became an important source of conflict during the Cultural Revolution. Women workers also gained some specific benefits, not least of which was the right to be paid the same as a man for the same work, but also including excellent maternity benefits and childcare facilities. However, the principle of equal pay for equal work was often violated in the twin processes of ‘feminization’ and ‘de-skilling’, such as occurred in the silk weaving industry. Just as Shanghai cotton mills began to employ more women workers in the 1920s, so silk weaving gradually came to be dominated by women in the 1960s. In a very similar process, as silk weaving came to be defined as ‘female’, wages for those jobs declined. By the 1980s, the men working in silk weaving factories were generally highly paid machinists and drivers, while the rest of the jobs, performed by women, were devalued. However, the fact that the party promoted industrial labour as glorious, and honoured industrial workers as the agents of history in the modern age, was very important in shaping women workers’ identities and sense of worth. Women workers went from being viewed as ‘broken shoes’, or low women, to being heroines of a new historical narrative, one which placed labour, especially industrial labour, at the heart of historical change and development.95 With the 1949 revolution, workers in private textile factories gained an unprecedented degree of influence over their employers. Silk weavers especially were able to participate in management and production decisions in factories that had established labour–capital consultative conferences. However, with the 94 Hershatter, The Workers of Tianjin, p. 180. 95 For the memories and experiences of this generation of silk workers, see Lisa Rofel, Other Modernities: Gendered Yearnings in China after Socialism (Berkeley, 1999), pp. 41–149. 135

The Ashgate Companion to the History of Textile Workers socialization of industry and commerce in 1956, the situation changed suddenly. Although production workers were often appointed as the state-side representatives in newly formed joint state–private enterprises, workers as a whole generally found their influence diminished once a factory came under state management. The haste with which these factories were ‘socialized’, the inequalities that developed when several factories merged and workers’ diminished influence in factory affairs contributed to the strike wave that erupted in Chinese cities in 1956–1957. More than 90 per cent of these incidents occurred in newly socialized factories. In some cases, worker protests resulted in violence and even death. Frustration with the weak party-led unions and the attitudes of state administrators also contributed to this, the second largest strike wave in Chinese history. The movement was crushed, however, and under the repressive political climate of the Anti-Rightist Campaign, and the frenzied drive for production in the ‘Great Leap Forward’, workers found it increasingly difficult to express their interests either individually or collectively.96 On the whole, workers in China’s urban textile industries benefited greatly from the Communist-led revolution, enjoying an enviable degree of job security and considerable benefits. However, they also lost the ability to organize independently and to respond autonomously to changes in their working lives. The party’s cooptation of the labour movement in the name of economic development and stability has had lasting consequences, and has strongly conditioned the nature of China’s industrial working class today. Chinese industry and agriculture recovered slowly from the ‘Great Leap Forward’, which witnessed impressive infrastructure and capital construction development at the cost of declining productivity and one of the worst famines in human history. Production in the 1960s improved only gradually. The souring of Sino-Soviet relations after 1960 also affected textile production, forcing state cotton and silk exporters to find new markets in the capitalist world. However, technological innovations, many introduced by textile workers themselves, continued to drive textile industrialization in the 1960s. The Cultural Revolution of 1966–1976 periodically interrupted production with political campaigns. But with the exception of silk, which was branded a ‘bourgeois’ cloth and a luxury product unsuited to a socialist society, textile production continued to increase by an average of 6.1 per cent annually. Production of cotton thread increased from 656 thousand tons in 1952, to 1.5 million tons in 1959, on the eve of the Great Leap Forward. Although production declined during the ‘Leap’, by 1966 it had recovered, and in 1970, China produced more than 2 million tons of cotton thread. Following Mao’s death in 1976, production increased even more rapidly, reaching 3.27 million tons in 1983. Cotton cloth saw similar progress, increasing from 3.8 billion metres in 1952, to 7.6 billion in 1959, and more than 9 billion metres in 1970, overtaking Japan as the world’s largest producer 96 On the 1957 strike wave, see Elizabeth Perry, ‘Shanghai’s Strike Wave of 1957’, China Quarterly 137 (1994): 1–27; See also Jackie Sheehan, Chinese Workers: A New History (London, 1998), pp. 47–84; and F. Gipouloux, Les cent fleurs a l’usine: Agitation ouvriere et crise du modele Sovietique en Chine, 1956–1957 (Paris, 1986). 136

China of cotton cloth. By the 1990s, China was producing more than 15 billion metres of cloth annually, and production has continued to increase up to the present day.97 By the end of the Maoist era, China’s cotton textile industry had not only increased in size, but had also undergone a fundamental transformation. State cotton mills, built on a huge scale, dominated textile production. Handicraft weaving was replaced by power-loom weaving. The silk industry had long since exceeded prewar production and quality levels. Despite ongoing marketing difficulties, by 1978, China had overtaken Japan, becoming the world’s largest producer of silk for the first time since the First World War. In 1994, China produced more than 70 per cent of the world’s raw silk. China furthermore began producing synthetic fibres in 1957. Production slowly but steadily increased, reaching 155 thousand tons in 1975. After 1978, synthetics grew even more rapidly. In 1983, China produced 540 thousand tons and in 1992, overtook Japan as the largest producer of synthetic fibres in Asia.98 China’s economy has undergone an equally far-reaching transformation under Mao’s successors. During the 1980s, and increasingly in the 1990s, the Chinese Communist Party radically re-imagined itself and its social and economic role. Private enterprise re-emerged and flourished. In the old handicraft textiles areas, small-scale rural factories again emerged, purchasing materials and second-hand equipment on the emerging markets, and utilizing local labour and capital. Some of these rural factories were collectively owned by the villages, while others were family-run businesses, but all profited from soaring demand for consumer textiles. Low production costs and taxes allowed these enterprises to compete effectively with the state sector in the 1980s. Drawing on many of the same advantages they had enjoyed in the 1920s, by 1993, rural workshops were producing 26 per cent of Chinese cotton cloth, 75 per cent of silk thread and almost half of all silk cloth.99 The 1990s thus saw the re-industrialization of rural China, and by 1993, China employed 40 per cent of the world’s textile labour force.100 The Chinese economy has also become more integrated into the world economy, becoming one of the most important targets of foreign investment and a major provider of consumer goods, especially textile products. By the mid-1990s, textiles production had expanded such that China had become the world’s largest producer of cotton textiles and a net importer of textiles machinery.101 China’s accession to the WTO in 2001 resulted in greatly expanded exports and further increases in production. China’s emergence as the ‘workshop of the world’, however, has been 97 For a thorough analysis of Chinese textile production in the mid-1990s, see James Glasse, Textiles and Clothing in China: Competitive Threat or Investment Opportunity? (London, 1995). 98 State Statistics Bureau (ed.), Zhongguo gongye de fazhan, tongji ziliao 1949–1984 [The Development of Chinese Industry, Statistical Materials 1949–1984] (Beijing, 1985), pp. 25–27. 99 Grove, ‘Rural Manufacture’, pp. 450–458. 100 Farnie and Jeremy (eds), The Fibre that Changed the World, p. 583. 101 Glasse, Textiles, pp. 72–77, 106–111. 137

The Ashgate Companion to the History of Textile Workers achieved through the dismantling of the socialist economy and the abandonment of many of the benefits won by industrial workers in the 1949 revolution. Economic growth has been achieved on the basis of a rapidly expanding labour market, combined with massive layoffs in the state sector. China’s new industrial workers enjoy few of the benefits of the socialist era, and are consistently denied even legal rights such as a 40-hour working week, overtime pay, labour insurance and union representation. Under these reforms, the organization of both silk and cotton textiles has changed dramatically. Market reforms in the 1980s led to rising prices for raw cotton and cocoons. The 1987 ‘cocoon war’ resulted in the dismantling of Shanghai’s silk industry. The last remaining silk weaving factory in that city moved to Songjiang at the end of 2004. With the liberalization and privatization of textile production, intense competition has developed, in which all firms are forced to keep costs down by any means available. Eliminating expensive benefits and laying off workers from previously secure positions has reduced production costs considerably, but has also contributed to rising unemployment. State workers enjoying the full benefits of socialism have gradually been replaced with younger, more docile rural migrants who are willing to work longer hours for low pay.102 China’s textile industries continue to face considerable challenges today. In China’s rapidly changing economy, prices for raw materials can vary considerably, sometimes as a result of speculation, as in the case of the 50 per cent increase in prices for raw cotton in 2003.103 With the recent elimination of country quotas for US and European textiles imports, Chinese producers look forward to a world market open to their products.104 But textile workers today in China can do little to protect themselves from ever-intensifying exploitation as owners seek to reduce costs and maintain competitiveness in the jungle of emerging markets.

Conclusions In the course of its long history, Chinese textile production has undergone many changes. Commercialization, mechanization and the establishment of capitalist, and then socialist relations of production, have accompanied the industry’s growth. However, historical changes have had different effects on different kinds of textile products and producers. Between 1000 and 1800, silk weaving came to be identified as masculine work, appropriate to the highly commercialized and 102 On the contemporary silk industry, see Feng Xu, Women Migrant Workers in China’s Economic Reform (New York, 2000); See also Rofel, Other Modernities, pp. 217–256. 103 For earlier price wars, see Glasse, Textiles; also Zhang Xiaohe et al., The ‘Wool War’ and the ‘Cotton Chaos’: Fibre Marketing in China (Adelaide, 1991). 104 David Barboza and Elizabeth Becker, ‘Free of Quota, China Textiles Flood the U.S.’, New York Times, 10 March 2005; Richard McGregor and Edward Alden, ‘China Pledges to Curb Clothing Exports to U.S.’, Financial Times, 11 March 2005. 138

China workshop-oriented production system of the Ming and Qing Dynasties. The growth of cotton cloth markets, whether in the context of eighteenth-century commercial networks or twentieth-century ‘putting-out’, did not produce similar changes, however. Despite the emergence of rural workshops and urban factories, most women, and some men, continued to weave at home as they had always done, well into the twentieth century. The impact of foreign imperialism and China’s incorporation into the world system from the nineteenth century had very different effects on cotton and silk thread production. Whereas the treaty port system and cheap imports both destroyed handicraft spinning and hindered the growth of Chinese cotton mills, foreign demand for silk fuelled the emergence of steam filatures owned and operated by Chinese. In the face of foreign imports and mechanized weaving, handicraft cloth production proved competitive, and in the 1920s developed new forms of enterprise based on the advantages offered by China’s rural economy. Ultimately, however, all of these types of textile production were mechanized, and the men and women engaged in textile production were transformed from domestic workers and artisans to proletarian wage labourers. In the case of silk weaving, this process was accompanied by increased employment of women, resulting in a reversal of the trend of the previous thousand years. At the dawn of the twenty-first century, Chinese textiles dominate world production. China is now by far the world’s largest producer of silk, cotton and hemp cloth. Although much of this is consumed domestically, China also exports to markets all over the world, having reversed the position of a century ago. The same is true of the needle trades, as China produces and exports almost every kind of clothing imaginable, from towels and underwear to wedding dresses and Japanese kimonos. It is impossible to say what directions Chinese textile production will take in the next century. But one can be certain that, whatever developments occur, textiles will continue to develop according to consumers’ needs, the pursuit of profit and the skill and determination of the millions employed in this ancient and vital industry.

139

This page has been left blank intentionally

6 Denmark: the textile industry and the formation of modern industrial relations Lars K. Christensen

Seen from a global point of view, Denmark has always been a minor player in the field of textile production. However, from the perspective of industrial relations, the Danish case has relevance in an international comparison. The textile industry provides a good example of how industrial relations based on strong unions and nationwide collective bargaining were formed in Denmark. These industrial relations, created from below, later became the foundation of the modern welfare state. This chapter will begin with a short presentation of the history of textile production in Denmark and will then focus specifically on the issue of industrial relations.

Textile production from pre-industry to the present Geographically Denmark consists of the peninsula of Jutland and approximately 100 inhabited islands, including Zealand with the capital of Copenhagen. In the late eighteenth century, the population was a little smaller than one million. Presently it is 5.4 million. The country is fertile, and for centuries agriculture was the main source of income, supplemented by trade, fishing and shipping. From ancient times, wool and flax were spun and woven into fabrics in peasants’ homes. This domestic manufacturing of textiles was undoubtedly of great importance in

 For a discussion of the concept of modernity, especially in relation to the labour movement and industrial relations, see Lars K. Christensen, Det moderne arbejde – kulturelle og institutionelle forandringer af arbejdet i den danske tekstilindustri 1895–1940 (Ph.D., University of Copenhagen, 1998), pp. 283ff.

Map 6.1 Denmark

Denmark pre-industrial Denmark. A coarse woollen fabric was used for everyday clothing, while the more precious linen could be used for undergarments or bedclothes. The last half of the eighteenth century was a period of peace and general prosperity in Denmark. Agricultural reforms gradually improved the daily life of farmers. In the same period, there was a change in the demand for textiles. Fabrics with a linen warp and woollen weft, often with bright coloured stripes or patterns, became popular. Common people used these fabrics for finer clothing and for interior decoration. In the rural areas, textiles were also produced by artisans: weavers, employed by an estate, or working in their own shop in a village. Normally, artisans were only allowed to set up shops in towns. However, weavers formed one of the few exceptions, and in some areas there could be two or three weavers in one village. The workshop of a village weaver was usually small, comprising only the weaver himself and maybe an apprentice. Spinning, winding and so forth was often carried out by the weaver’s wife or children. While peasants in some areas ceased to weave their own textiles and instead started to bring their homespun yarn to the village weaver, peasants in other parts of the country saw an opportunity to supplement their income by selling their domestic products. In certain areas, the domestic manufacturing of textiles grew considerably, especially between 1750 and 1850. Overall, it seems safe to conclude that in the eighteenth and early nineteenth centuries domestic production and rural artisans provided the majority of textile products.

Manufactories The manufacturing of textiles became a centrepiece in the mercantilist endeavours of the Danish monarchy. Agriculture provided wool and flax, and cotton was imported from the colonies. The missing factors were the technology, know-how and capital needed for large-scale manufacture. However, the absolutist state had a relatively effective system of tax collection at its disposal. With this as a foundation, capital was accumulated and transferred to state subsidies for privately owned manufactories. The monarchy also commanded a well-developed administrative system, which was used to secure the necessary supply of raw materials and skilled labour for industrial production, and to further the sale of products.

 Known as vadmel in Danish.  Known as hvergarn in Danish.  E. Andersen and E. Budde-Lund, Folkelig vævning i Danmark (Copenhagen, 1941), pp. 7ff.  Even though Denmark was a minor colonial power, cotton was produced in the Danish West Indies.  Richard Willerslev, ‘Den danske tekstilindustris udviklingsforløb 1730–1960’, in Helge Bjørn and E. Zinklar Zinglersen (eds), De danske tekstilerhverv (Copenhagen, 1965), pp. 112f. 143

The Ashgate Companion to the History of Textile Workers Since medieval times, the trades had been organized in guilds. The guilds were a mechanism for internal control within the trades, in order to avoid competition and to control prices. They were also instruments of state intervention, since guild regulations had to be officially approved. Membership was compulsory for most artisans. In Copenhagen and other major cities, guilds existed for cloth-makers, weavers, silk weavers and hosiery-makers. New factory-guilds were organized, mimicking this well-known form of organization. These guilds provided the owners of the manufactories with quite extensive power over their workers, as they prescribed fixed maximum wages and gave employers the option to refuse labourers the right to leave the factory. A subsidized General warehouse (Generalvaremagasinet) was established, for the purpose of securing imported raw materials at favourable prices, providing a guaranteed outlet for products and supplying the textile merchants with credit. Furthermore, the import of foreign textiles was banned – even though the ban seems not to have been completely effective. A vast amount of direct and indirect state subsidies were paid and a number of master artisans were attracted from abroad, especially from Germany, who founded so-called textile manufactories, mainly in Copenhagen. Most of these manufactories were quite small, and a significant number were organized according to the putting-out system, where a number of domestic workers or small artisan shops worked under the supervision of a master artisan, or sometimes a merchant. One of the exceptions to this rule was the Royal Privileged Cotton Manufacture – publicly known as the Manchester Factory – established just outside Copenhagen in 1779. This manufactory marked an important first step in the mechanization of textile production: it was equipped with both water frames and mules for spinning as well as carding machines. There was no water power though; instead all machines were either hand- or horse-driven. The size of the complex, as well as the introduction of spinning machines, made this the first Danish establishment deserving the designation factory. Due to the Tools Act, which prevented export of certain machines from England, technology could not be imported, but had to be copied locally. The Swedish-born  In Danish: Dugmagere, tøjmagere, silkevævere and strømpevævere. The words dugmager and tøjmager can be confusing, since Dug and Tøj are both derivations from the German word Tuch, which originally had the broad meaning of tool, equipment or clothing. In German, the combination Tuchmacher has the specific meaning of cloth-maker, which has been transferred to Danish as Dugmager, designating a person who is producing fulled fabrics from wool. The word Tøj is also a derivation from Tuch, which originally had the same broad meaning (for example equipment), but has acquired the more specific meaning of clothing. Thus, Tøjmager means a weaver who is producing light, non-fulled fabrics from wool, flax or (later) cotton.  Axel Nielsen, Industriens historie i Danmark, vol. II (Copenhagen, 1943–1944), pp. 83–84; Willerslev, ‘Den danske tekstilindustris udviklingsforløb’, pp. 113–114.  Nielsen, Industriens historie i Danmark, II, pp. 174ff.; Dan Ch. Christensen, Det moderne projekt: Teknik & kultur i Danmark–Norge 1750–(1814)–1850 (Copenhagen, 1996), pp. 269ff. 144

Denmark manager of the Manchester Factory, Ch.A. Nordberg, later left the Manchester Factory to establish a machine shop, which supplied several textile mills with water frames and other machinery. In an irony of history, the Manchester Factory was destroyed in 1807, when Copenhagen was bombarded by English troops, because of Denmark’s involvement in the Napoleonic wars. Because it caused the rapid devaluation of Danish currency, the war was immediately advantageous for the textile manufacturers. For a short period, the export of textile goods was possible. This attracted investment capital, and the production of cloth quadrupled between 1801 and 1814. In the smaller cotton sector, growth was relatively even higher. The growth of production, however, was not the result of mechanization; it was rather based on well-known technological and organizational forms. Despite these immediate advantages, wartime had a catastrophic impact on the country as a whole. In 1813 the state was in effect bankrupt, and the following year saw the separation of Norway from the monarchy. State subsidies for textile production were radically lowered, import restrictions were abolished and the economy entered a long period of deflation. The result was the worst ever economic crisis for Danish textile producers. Pre-war cotton production was almost wiped out by the renewed competition from England. However, the demand for cotton fabrics was continually growing, and it soon became possible to make a business out of weaving cotton from yarn imported from England. New weaving shops were established in several provincial towns – partly by Copenhagen manufacturers, who wanted to take advantage of the cheaper labour costs outside the capital. The number of workers in cloth production dropped by more than two-thirds between 1814 and 1830. Nevertheless, production only dropped by one-third during this period.10 This increased productivity reflected the fact that those cloth producers who survived the crisis primarily did so by relocating from Copenhagen to areas where water power was available. The most prominent example was I.C. Modeweg, who had established himself as a cloth manufacturer in the centre of Copenhagen in 1809. In 1831, he relocated his enterprise to Brede, in the countryside north of Copenhagen. This new cloth mill was to become one of the leading textile factories in the nineteenth and early twentieth centuries.11 Thus, in the 1830s the majority of cloth production had become concentrated in three to four large mills, all based on water power.12 The larger cloth mills had all mechanized the carding, spinning and finishing processes. Weaving, however, continued to be carried out by hand. 10 Willerslev, ‘Den danske tekstilindustris udviklingsforløb’, p. 121. 11 I.C. Modeweg’s cloth mill in Brede was finally closed down in 1956. Shortly after that, the complex was acquired by the Danish National Museum. It has been the subject of a number of studies – most recently in Jeppe Tønsberg, Den danske klædeindustri i international belysning: virksomheden I.C. Modeweg & Søn A/S (Brede Klædefabrik) 1810–1956 og dens baggrund (Copenhagen, 2003), but also in Lykke L. Pedersen et al., Industriens vugge: Brede – et fabrikssamfund ved Mølleåen 1800–1965 (Copenhagen, 1993) and several others. 12 Nielsen, Industriens historie i Danmark, II, pp. 78–79. 145

The Ashgate Companion to the History of Textile Workers

Industrialization In 1848, the absolutist monarchy was abolished and a liberal government was instituted, which carried out reforms in favour of free trade, and from 1862 the guilds lost their right to control the market. Since a growing part of the population lived in towns, both the supply of labour and the demand for textile goods increased. Following a civil war in 1864, the duchies of Schleswig and Holstein were lost by the Danish crown, and became part of Germany. Consequently, the cloth mills in the Danish part of the monarchy were relieved of major internal competition from the well-developed cloth industry in Holstein. Based on a general increase in the demand for agricultural products, nationally as well as on the world market, this was a period of growth and modernization in the countryside. Consequently, the domestic production of textiles became relatively less attractive to farmers and smallholders than agriculture. Moreover, the export of agricultural products brought capital into the country, which in turn could be invested in industrial development. Between 1840 and 1865, modern industry gained a foothold in Denmark.13 In 1843 two mills were equipped with power-looms and steam power respectively, and in 1846 Modeweg’s cloth mill in Brede was the first to adopt both new technologies together.14 The replacement of handlooms by power-looms did not proceed rapidly, though. The cloth mills typically used carded yarn, which was relatively fragile. The only way to prevent the weft from breaking frequently was by reducing the speed of the loom – which hampered the productivity gain of power-looms.15 Nevertheless, as power-looms improved, they gradually became dominant. Spinning technology was also modernized, and in the mid-1860s self-actors were introduced, at least at the cloth mill in Brede.16 The number of cloth mills grew, reaching a peak around 1875. Many smaller mills did not have the resources to take part in the technological advancements of this period and were put out of business. However, the number of workers kept increasing – a sign of the concentration of production in larger units.17 The technology employed in the industrialization of cloth production seems to have been of mixed origin. Locally made machines – often copied from foreign models – were still in use. From around 1870, German

13 Ole Hyldtoft, ‘Den teknologiske udvikling i Danmark’, in Flemming Mikkelsen (ed.), Produktion & arbejdskraft i Danmark gennem 200 år (Copenhagen, 1990), pp. 35–56. According to Hyldtoft, industrialization in Denmark generally took off in a first phase of in-depth change in 1840–1865, became stabilized in 1865–1896 and entered a new phase of in-depth technological development in 1896–1933. See also Ole Hyldtoft, Københavns industrialisering 1840–1914 (Herning, 1984), especially pp. 23–45. 14 Tønsberg, Den danske klædeindustri, pp. 50–51. 15 Ole Hyldtoft, Teknologiske forandringer i dansk industri 1870–96 (Viborg, 1996), p. 159. Hyldtoft also assumes that there were cultural barriers against the power-looms among the skilled weavers. 16 Ibid, pp. 158, 163. 17 Willerslev, ‘Den danske tekstilindustris udviklingsforløb’, p. 127. 146

Denmark machines became dominant in the wool industry.18 In addition, many of the masters and overseers in the Danish cloth mills were of German origin, or at least educated in Germany. The demand for cotton fabrics kept on growing. Mechanical looms were introduced in cotton weaving in 1843, the same year as in the cloth mills. In 1854 an enterprise was established in Copenhagen, which, under the name of I.H. Ruben, would become the largest Danish manufacturer of cotton textiles during the nineteenth century.19 English technology was more strongly represented in the cotton industry than in the cloth mills. But there was also a German influence, not only via the supply of machinery, but also via the education of masters and overseers. A few steam-powered factories were also established in provincial towns. It seems as if power-looms spread more quickly in the cotton industry than in the cloth mills. Nevertheless, industrialization as such progressed more slowly in this sector. In 1872 there were still slightly more workers employed in non-mechanized cotton and linen weaving in Copenhagen than in steam-powered cotton factories – but the production of the latter was more than two times that of the former.20 The final deathblow to non-mechanized cotton weaving in Copenhagen was a hand-weavers’ strike in 1884, since some of the major operators simply reacted by closing down their hand-weaving shops.21

Industrial prime time From the mid-1890s, the process of industrialization entered a new phase. New types and methods of production characterized the growing industry.22 In the late nineteenth and early twentieth century, the cloth mills continued the process of mechanization. Some of the smaller mills were shut down, while the larger ones grew even bigger. Some of the largest mills were converted into limited companies, reflecting the need for investment capital. Between 1897 and 1914, the use of machine power more than tripled, while the number of workers rose only 16 per cent, causing increased productivity.23 The same tendencies were visible in the cotton industry. Attempts were made to raise productivity by having one weaver operate several looms simultaneously. 18 Hyldtoft, Teknologiske forandringer i dansk industri, pp. 167ff; Tønsberg, Den danske klædeindustri, pp. 63–64. 19 Nielsen, Industriens historie i Danmark, III.2, p. 100; Hyldtoft, Københavns industrialisering, p. 81, 191. 20 Hyldtoft, Københavns industrialisering, p. 188. 21 J.J. Møller, Dansk Tekstilarbejderforbunds historie. Tekstilindustriens udvikling i Danmark gennem Aarhundreder samt Dansk Tekstilarbejderforbunds Historie fra Forbundets stiftelse til 1918 (Copenhagen, 1924), pp. 106–107; Willerslev, ‘Den danske tekstilindustris udviklingsforløb’, p. 125. 22 Hyldtoft, ‘Den teknologiske udvikling i Danmark’, pp. 42ff. 23 Hans Chr. Johansen, Industriens vækst og vilkår 1870–1973 (Ordense, 1988), p. 110. 147

The Ashgate Companion to the History of Textile Workers The multiple loom system was introduced gradually from around 1900, and its progress in the cotton industry is illustrated in Table 6.1. Another significant development was the re-establishment of spinning within the cotton industry. In 1892, a spinning plant was established in Vejle on the east coast of Jutland, using ring spinning machines for the first time in Denmark. Another spinning plant was established in the same town a few years later, and together with some major cottonweaving plants, this established the city’s reputation as the Danish ‘Manchester’.24 A few other plants were established in the following years, among them a large one in Valby, Copenhagen. While cloth mills integrated spinning, weaving and fulling in a single plant, cotton production would remain divided in separate spinning and weaving mills. Table 6.1

Cotton weavers operating one or several looms, 1906, 1913 and 1937 (%) Year

One loom Two looms Three–four Six looms looms or more

1906

20

67

12

1

1913

13

51

35

2

1937

0

24

60

16

Source: Danmarks Statistik, Statistiske Meddelelser: produktionsstatistikken 4.30.II, pp. 15, 26; 4.50.I, pp. 19, 21; Textilfabrikantforeningen: Gennemsnitsfortjenesten i September 1937. Dansk Tekstilarbejderforbunds arkiv, Arbejderbevægelsens Bibliotek og Arkiv, folio no. 217.

Entering the twentieth century, the Danish textile industry was well developed, but it was also a fragile industry, since it experienced heavy competition from imported goods on its own home market. Textile factories were generally larger than the average Danish factory. This was especially true in spinning, were the four to six plants typically had 200–350 workers each in the period 1906–1935, compared to the average in industrial plants of 30–40 workers.25 Measured by the number of

24 Hyldtoft, Teknologiske forandringer i dansk industri 1870–96, p. 186; Bomuldsbyen. Tekstilarbejder i Vejle gennem 100 år (Vejle, 1989); John J. Hansen, Vejle – et tekstileventyr. Vejles tekstilindustri gennem 200 år og Tekstilarbejdernes Fagforening i Vejle gennem 100 år (Vejle, 1996). 25 Figures for the industry in general: Svend Aa. Hansen, Økonomisk vækst i Danmark (2 vols, Copenhagen, 1976), I, p. 287 and II, p. 64. Figures for the textile industry: calculated from Danmarks Statistik, Statisk Tabelværk – see Christensen, Det moderne arbejde, Appendix 1 for details. 148

Denmark workers employed, textile production was one of the important branches of Danish industry at the beginning of the twentieth century (Table 6.2).26 Table 6.2 Industrial textile workers, 1872–1993 Number of textile workers

Percentage of industrial workforce

1872

3,542

10

1897

5,784

6

1914

6,723

5

1935

8,225

4

1951

14,129

5

1965

8,172

3

1972

4,620

1

1993

3,498

1

Source: Johansen, Industriens vækst og vilkår 1870–1973, various tables (1872–1972); Danmarks Statistik, Statistikbanken (2003), table RAS8 (http://www.statistikbanken.dk) (1993). Only blue-collar workers in cloth, cotton and linen production are included.

Denmark did not take part in the First World War, but during the last years of the war the cotton industry was hampered by the lack of raw materials. The cloth mills could substitute imported wool for domestic wool, as well as shoddy, and thus they were not severely affected by the war itself. However, the general post-war crisis in 1920–1921 dealt a severe blow to both branches of the industry. Many of the larger textile plants, which were established during the nineteenth century, experienced difficulties in the interwar period, and some were closed down. It was a turbulent period with several instances of economic recession and high unemployment. In turn, this of course affected the consumption of textile goods. Moreover, the textile industry felt the competition from major textile-producing countries – primarily England and Germany. Prior to the First World War, the Danish industry possessed approximately 50 per cent of the home market. In the 1920s competition grew even stronger, and the market share of national textile production fell below 33 per cent. During the recession in the 1930s, a Social Democratic government introduced a 26 Cloth mills and cotton weaving and spinning occupied 6,723 in 1914. This can be compared to 9,843 in machine-building, 7,949 in the tobacco industry and 6,230 in shipyards. 149

The Ashgate Companion to the History of Textile Workers more protectionist trade policy, resulting in the market share of national production rising again to around 50 per cent.27 Aside from the general political and economical developments of the period, part of the problems can be ascribed to the industry’s inability to adapt to changing consumer demands. While there was a change in demand from heavy to lighter woollen fabrics, the industry was in fact producing relatively more heavy fabrics until the mid-1920s. The woollen industry finally adjusted after 1925, when its production shifted markedly towards light fabrics.28 The cotton industry, however, still lagged behind. As late as 1935 printed cotton fabrics – which were high in demand – were not produced at all. In fact, even the employers’ association admitted in 1939 that the industry was failing to keep up with the ongoing changes in fashion.29

Working faster There are only limited documentary sources on the productivity of the Danish textile industry prior to the First World War but the sources we have indicate an average growth of approximately 1.2 per cent per annum from 1897 to 1905 and 2.4 per cent per annum from 1905 to 1913.30 These figures are pretty much in line with the increase in productivity for the industrial sector as a whole during this period. From 1916 onwards we have annual figures,31 which indicate a fluctuation in yearly production, especially during the First World War and the first part of the 1920s. This is not surprising, since this period was generally characterized by rapid changes in the economy, falling supplies of raw materials during the war years and a wave of strikes and lockouts in the 1920s. Despite these short-term fluctuations, there was a clear long-term tendency towards productivity growth from 1916 until around 1930, with an average yearly increase of more than 3 per cent. From 1930 onwards, the growth stopped, and productivity was more or less constant until the Second World War. This tendency is quite consistent with industrial development

27 Johansen, Industriens vækst, pp. 111, 161–162. 28 Danmarks Statistik, Statistiske Meddelelser, IV.75.7, pp. 51–52, 64; IV.103.6, p. 48. 29 Johansen, Industriens vækst, pp. 162–163; Otto Jensen, Dansk Tekstilarbejderforbund gennem et halvt Aarhundrede 1885–1935 (Copenhagen, 1935), p. 265; Danmarks Statistik, Statistiske Meddelelser, IV.103.6, p. 51. 30 Calculated from figures in Danmarks Statistik, Statistisk Tabelværk (1897), V.A.1, table 12 and Statistiske Meddelelser (1905), 4.30.2 and (1913), 4.50.1. Productivity is calculated as the ratio between the number of workers employed and the output of the production, measured either in quantity or in fixed prices. Both these and the following numbers are based on source material which is not always complete and includes some factors which have had to be estimated. Thus, the numbers given should be considered as approximations. For a further discussion of the source material and methods of calculation, see Christensen, Det moderne arbejde, pp. 84–89 and Appendix 2, p. 330. 31 Based on Danmarks Statistik, Statistiske Meddelelser for each year. 150

Denmark in general, even though overall decline of industrial growth seems to have started a few years earlier. The statistics are not very detailed, but it seems that most of the rise in productivity between 1897 and 1913 was achieved in cotton spinning, while productivity in cotton weaving and cloth-making remained almost constant. During the interwar period, the situation was reversed: apart from some heavy fluctuations in the early 1920s,32 spinning did not experience productivity growth, while cotton weaving and cloth-making experienced a very steep rise in productivity following the end of the First World War, reaching a peak around 1935 with a productivity approximately 2.5 times that of the last ‘normal’ years prior to the outbreak of the war. The rise in productivity coincided with a further significant increase in the multiple loom system employed, particularly in the cotton industry, as can be seen in Table 6.1. The multiple loom system did not spread as rapidly in the cloth mills, probably because the woollen weft was more fragile than that of cotton. The development of the system was facilitated by the introduction of automatic looms just after 1900. Some of the early attempts at introducing this new technology were made with looms from the Swiss company Rüti, while the English Northrop-type also seems to have been in widespread use.33 Because of its more complicated nature, an automatic loom would run at a slower pace than an ordinary loom. An English experiment, referred to in a widely used Danish textbook for textile engineers, showed that all things being equal, automatic looms in a four-loom system produced between 7 and 16 per cent less per week.34 A clear advantage of the automatic loom was that the weaver did not have to change the shuttles or cops, and so he or she could attend more looms at the same time. Depending on the wages paid for attending more looms, the loss in efficiency for each loom could be more than compensated by the decrease in the cost of labour. However, automatic looms were in no way a prerequisite for the multiple loom system. In fact, the initial expansion of this system did not require much change in technology. We have no precise numbers for the spread of automatic looms, but different sources suggest that it was a gradual process, and that at least many of the three- and four-loom systems were based on ordinary power looms. The industry may also have applied other technical advancements, such as machinery for cutting and preparing the warp. But all in all it was not technical advancements as such, but rather new ways of organizing the labour process which caused the rise of productivity in the interwar period. Textile workers simply had to tend more machines and/or work faster. One female weaver recollects: 32 These fluctuations might be the result of flaws in the statistics. 33 Nationalmuseets Industri-, Haandværker og Arbejdererindringer (The National Museums Collection of Industrialists, Artisans and Workers Memoirs) (hereafter NIHA), no. 1197, pp. 9–10; no. 1872, pp. 13 and 26; and no. 2571, p. 5; H.J. Hannover, Tekstilindustri (3 vols, Copenhagen 1924–1938), III, pp. 904ff. 34 ‘Automatic versus Ordinary Looms. Official Report concerning a Test of Automatic Looms, etc., made in 1931 by the Lancashire Cotton Corporation Ltd.’, International Cotton Bulletin, April (1932). 151

The Ashgate Companion to the History of Textile Workers The textile industry underwent widespread rationalization. Automaticlooms were introduced, so that instead of tending two or four looms, we now had to tend to eight, and work in double shifts from 6am to 3pm and 3pm to 12pm. That was in 1929 and I worked this way for 12 years. My nerves were strained and I had difficulties falling asleep.35 Between 1933 and 1935, a few companies even took this a step further, when they introduced the Bedaux system. This was a system of rationalization, inspired by Taylor’s Scientific Management and based on time studies, but adapted to the particular requirements of the textile industry. However, the Bedaux system never came into widespread use.

The post-war climax During the Second World War, Denmark was occupied by Germany. Compared to most other Nazi-occupied countries, the direct impact of the war was relatively insignificant. One of the impacts was a shortage of raw materials, which also affected the textile industry. However, the war also prevented much of the competition from abroad, and a few plants were actually expanded during these years. The lack of competition continued in the immediate post-war period, and around 1950 the industry reached a historic peak in the number of factories and the size of the workforce.36 It was at this point, in 1950, that a newly elected liberal government abolished the protectionist trade policies that had favoured the national industry. The renewed competition from abroad was a hard blow to some parts of the textile industry. The traditional production of fabrics for clothing declined. Following worker and employer protests, the duty on imported textiles was raised in 1956. However, this only provided temporary relief, and several mills were forced to close during the 1950s and 1960s. Large companies, such as Mogensen & Dessau in Odense, which had built the largest single cotton weaving shed in northern Europe in 1949, and I.C. Modeweg & Søn A/S, the oldest cloth mill still in existence, closed down in the early 1950s. The industry reacted with intensified attempts to rationalize production. From 1948 to 1954 productivity rose almost 5 per cent per annum.37 The number of persons employed in cotton weaving and cloth production, as well as the total number of factories, diminished by 50 per cent during the 1950s, even though total production only fell by approximately 15 per cent.38 The increase in productivity and decline in employment were characterizing features of the production of yarn and fabrics for clothing until the beginning of the 1970s at least. This situation was sustained by new technologies, such as more efficient ring spinning machines and the new rotor 35 NIHA, no. 1379, pp. 27–28. 36 Johansen, Industriens vækst, pp. 223–225. 37 Jacob Christensen and Tonny Jensen, Textil og skotøj i 100 år (Copenhagen, 1985), p. 42. 38 Johansen, Industriens vækst, p. 225. The numbers are for the period 1950–1958. 152

Denmark spinning machines, introduced in the mid-1960s. At the same time, new types of shuttleless looms, capable of weaving at even higher speeds, were introduced.39 Another new technology taken up by the Danish textile industry was that of carpet production using mechanized tufting. From a modest beginning, this production increased during the 1960s, spurred on by the general increase in housing standards of that decade. This was also one of the few examples of the Danish textile industry able to export to the world market. In contrast to other sectors of the textile industry, employment within the carpet industry rose steadily until the beginning of the 1970s – but not enough to compensate for the overall decline of employment.

Post-industry? While the 1960s had been a period of gradual decline, the early 1970s were a period of explicit crisis for the Danish textile industry. One reason was the general economic recession. The sale of long-term consumer goods was especially affected by the recession, which meant that employment levels were decreasing, even within the carpet industry.40 Another contributing factor was the transition to free trade within the European market, which was a consequence of Denmark joining the EEC in 1972. In addition, new countries were becoming active textile producers and exporters, able to compete on the world market, based on their comparatively lower wages.41 Finally, changes in public consumption towards synthetic fabrics caused problems for the traditional cotton- and wool-based industry, of a more long-term nature. The companies that best survived the crisis seem to be those that specialized their product range, in terms of design, flexibility and superior quality. Egetæpper, a successful carpet factory, was among the first in Europe to invest in computercontrolled dyeing technology, which made it possible to apply colours to the fibres of an already finished carpet, thus enabling rapid changes in design.42 The last survivor of the major cloth mills, Kjærs Mølle in Ålborg, gave up producing textiles for clothing in favour of high quality furniture fabrics, under the trade name of Gabriel. Another characteristic feature has been a number of mergers, takeovers and other changes in ownership.

39 John Thøgersen, Omstilling i tekstil– og beklædningsindustrien (Aalborg, 1986), pp. 118ff. 40 Ibid., pp. 130–131. 41 Johansen, Industriens vækst, p. 291. 42 Thøgersen, Omstilling i tekstil– og beklædningsindustrien, p. 134. 153

The Ashgate Companion to the History of Textile Workers

Labour, organization and industrial relations The guilds were the predominant instrument of social organization of labour in the pre-industrial trades. They took care of the social needs of their members – masters as well as journeymen – and provided a framework for the cultural praxis of the artisans. To be regarded as a skilled and worthy journeyman, one had to complete an apprenticeship, followed by a period of travelling from workshop to workshop, picking up new experiences. Danish artisans would often travel in central Europe, which is one of the reasons why many expressions used within the trades are German in origin. One of them is zünft: the concept used to designate the rituals that travelling artisans had to know and practise in order to be accepted by fellow artisans. After travelling for some years, a journeyman would normally settle down, and become master of his own workshop. In the late eighteenth century, the gradual development of a capitalist economy began to undermine the guilds. The guilds proved unable to handle the new types of disputes and contradictions that arose as journeymen became wage-labourers and masters became employers. Furthermore, from the point of view of the proponents of liberalism, the guilds were harnessing free trade and constituted an obstacle to economic development. The liberal government, constituted in 1848, removed the guilds’ rights to control trade, and made membership voluntary in 1862. Journeymen turned wage-labourers and unskilled workers made redundant by the structural changes in agriculture formed a growing proletariat in the larger cities – especially Copenhagen.43 In the 1860s, the ‘worker issue’ was discussed in a variety of forums. Attempts were made to revitalize the guilds, while others tried to form workers’ associations on a liberal platform. However, in the end, the most viable new form of social organization of labour would be the socialist trade unions and the modern labour movement. The first Danish labour movement was formed in 1871. It was called the International Workers’ Association – the International in short – and was part of the First International. It included a special weavers’ section. The International was quite successful in organizing the, mainly Copenhagen, working class during the first half of the 1870s. Soon the organization was banned and its leaders jailed. Furthermore, a period of recession led to rising unemployment. The labour movement declined, and had to be reconstituted during the 1880s. The reconstitution took place based on a decision taken in 1878 to split the labour movement into two parts: a Social Democratic political party and a trade union movement. A division of labour was established that would characterize the Danish labour movement for generations to come.

43 Knud Knudsen, ‘Arbejdets historie i Danmark. Nogle hovedlinjer i det kapitalistiske arbejdes opståen og udvikling i Danmark frem til 2. verdenskrig’, in Mikkelsen, Produktion & arbejdskraft i Danmark, pp. 99ff. 154

Denmark The quest for recognition Following the ban of the International, a new weavers’ union was formed in Copenhagen in 1873.44 From the beginning, only artisans working as hand-weavers were accepted as members. Women, who tended mechanized looms in the cotton factories, were generally regarded as inferior, since they were not skilled and did not practise the zünft. In 1874 an attempt to organize female weavers in a separate organization failed. The first strike organized by the union in 1873 was a success. However, the decline of hand-weaving made life hard for the union in the following years, and in 1884 the statutes of the union were changed, making it formally possible for both male and female industrial weavers to become members.45 In the following years, new unions in the two provincial towns of Horsens and Odense joined the Copenhagen union, and in November 1885 the Danish Weavers’ Federation was officially proclaimed. Membership increased very slowly during the first 10 years of its existence, but grew suddenly in the second half of the 1890s. The most important factor behind this was probably the changes in industrial relations within the textile industry, which again were a result of the work carried out by the union itself. In the first years of the union’s existence, strikes were undertaken on a more or less ad hoc basis, sometimes with very little preparation. The strength of the workers in such a conflict would depend on the union’s ability to collect funds among those members who were not on strike – not to mention the often difficult task of keeping strike-breakers out. This way of fighting for improvements in working conditions was costly, both for the Federation and for the participants. Furthermore, the results were fragile: the employers could lower wages again, as soon as they found an opportunity.46 Out of these experiences, a twofold strategy evolved. On the one hand was the quest for negotiated wage schedules. The Federation demanded that wages should be determined through negotiations between the employer and the union and stipulated in writing. On the other hand was the quest to strengthen internal discipline among the workers, in order to ensure that strikes would be better prepared and organized. In 1892 the first small steps were taken to build up a strike fund. In the mid-1890s, the first wage schedules were negotiated. In 1895 the union’s name was changed from the Weavers’ to the Textile Workers’ Federation, to signal that its principal aim was to organize all workers within the textile industry, regardless of their level of skill. The increased membership in the second half of the 1890s coincided with increased strike activity, coordinated by the central leadership of the Federation. In this way, the Federation succeeded in ‘covering’ several provincial towns, by striking – or threatening to strike – against 44 Under the Danish name Vævernes Velfærdsforening (The Weavers’ Welfare Association), which was changed the following year to Vævernes Fagforening (The Weavers’ Trade Union). 45 Møller, Dansk Tekstilarbejderforbunds historie, pp. 107, 112. 46 Christensen, Det moderne arbejde, pp. 207ff. 155

The Ashgate Companion to the History of Textile Workers one employer at a time. In 1898 the Federation decided to take on the employers in Copenhagen in the same fashion. But the Copenhagen employers reacted in an unforeseen manner: they joined the provincial employers in a national Textile Manufacturers’ Federation (Dansk Tekstilfabrikantforening). The strategy of the employers was to obtain centralized control over wage negotiations. Responding to the demand by the union for a common wage schedule for Copenhagen, the Manufacturers’ Federation demanded negotiations for a nationwide agreement. After some hesitation – based on an unwillingness to contribute to the strengthening of the employers’ federation – the Textile Workers’ Federation accepted. The outcome was the first national agreement for the textile industry. The national agreement stipulated the general conditions of employment, such as conditions for paying piece-rates and so on. Furthermore, the agreement stipulated that only ‘lawful’ strikes and lockouts were acceptable.47 The concept of ‘lawful’ was not defined, but the course of events seems to indicate that both parties took it quite seriously. The national agreement was to be in effect for three years, after which it was to be re-negotiated. It was still to be supplemented by local wage schedules – henceforward referred to as local agreements. Such agreements were concluded with a number of companies. For the textile industry, 1898 marked the breakthrough to industrial relations based on the collective bargaining of wages and working conditions. In the same year the National Confederation of Trade Unions (CTU) was formed. The trade union federations that comprised the CTU represented approximately one-quarter of all workers, including those in rural areas. However, among male workers in larger towns the rate of organization was as high as threequarters or more. These numbers made the Danish working class one of the best organized in the world.48 Agreements and collective bargaining also evolved in other branches of industry. Eventually, most employers accepted the unions as a permanent phenomenon, best dealt with through negotiated agreements. The National Danish Employers’ Confederation was formed in 1898, the same year as the CTU. While the different federations that made up the CTU maintained much autonomy, the strategy of the employers’ confederation was one of centralization. They consequently strove to make the national confederations of the two parties the main bodies of bargaining. This strategy, which the employers would pursue for most of the twentieth century, was supposed to discourage local strikes and diminish pressure for higher wages. Ultimately, the employers would threaten to declare a national lockout.49 The following year, the Employers’ Confederation felt ready to test this strategy, and launched a nationwide lockout. The Textile Manufacturers’ Federation 47 Quoted from Jensen, Dansk Tekstilarbejderforbund, p. 177. 48 Knud Knudsen et al., Kampen for en bedre tilværelse. Arbejdernes historie i Danmark fra 1800–tallet til 1990 (Copenhagen, 1991), pp. 72–74. 49 F. Ibsen and H. Jørgensen, Fagbevægelse og stat (2 vols, Copenhagen, 1979), I, p. 121. J. Due et al., Den danske model. En historisk sociologisk analyse af det kollektive aftalesystem (Gentofte, 1993), p. 76. 156

Denmark apparently felt bound by the agreement with the Textile Workers’ Federation, and refrained from taking part in the lockout in 1899. Thus the textile workers were not directly involved in this conflict. The conflict, however, assumed principal importance for the whole industrial labour market. The employers claimed that the time had come when it was necessary for them to prove that they were ‘the master in their own house’. The answer from the labour movement was that this was an obsolete point of view: under modern, industrial conditions of production, the factory could not be equated with the employer’s house. Thus, the terms of employment had to be considered a contractual relation between two parties, neither of which could have an exclusive right to decide.50 After more than three months of fierce conflict, which affected 40,000 workers and their families, a settlement was finally agreed upon. The main points of this September Compromise were that the employers recognized the workers’ right to organize and negotiate collectively, while the trade union movement in turn recognized the employers’ exclusive right to manage and distribute work at the workplace. Furthermore, certain formal procedures had to be followed before a strike or lockout could be declared. These principles were to be valid for all member organizations of the CTU, including the Textile Workers’ Federation. The September Compromise was to become the basis of industrial relations in Denmark throughout the twentieth century.51

Consolidation As industrial relations based on centralized, collective bargaining developed, the trade union federations had to accept the formal responsibility to uphold agreements and prevent their members from violating them by striking. During the following decades, the Textile Manufacturers’ Federation as well as the Danish Employers’ Confederation continuously strove to centralize negotiations and the right to close agreements. This development was contrary to the tradition of autonomy within the labour movement, and a potential source for conflict. Due to these developments, the relationship between labour and capital became more formalized and rationalized. The class struggle became enshrouded in technical deliberations. The 1901 national agreement between the Textile Workers’ Federation and the manufacturers, which replaced the first agreement of 1898, is an example of this. During negotiations the question of piece-rate scales was brought up. Most companies used their own scale, stipulating payment for many different types of fabric separately, while others used the so-called English method of payment, based on picks per inch. Both sides found that there was a need for a more uniform model. Finally it was agreed that all piece-rate scales should be based on the English model – even to such an extent, that all measures should be 50 Knudsen et al., Kampen for en bedre tilværelse, p. 75. 51 The settlement remained in force without changes until 1960. That year, it was replaced by a revised general agreement, built on the same principles. 157

The Ashgate Companion to the History of Textile Workers made in English inches. Since an English inch is 3 per cent shorter than a Danish inch, this meant a proportional reduction of payment, if the rates were not adjusted accordingly. Apparently, when local negotiations of piece-rate scales started, based on the general agreement, the effects of these very technical deliberations were not clear to all union activists. This resulted in some branches making agreements that led to a decrease in wages, due to a lack of understanding of the implications of the new system.52 In Odense, which had the largest local branch outside Copenhagen, some members went on strike against the wage decrease, but since this was considered an invalid action, the Federation refused to support them and they were eventually expelled. Several others followed them, resigning from membership in protest. Other local union activists took the challenge upon themselves. One example is the workers at Bloch & Andresen’s cotton mill in Jutland, who decided to produce their own proposal for a new piece-rate scale, based on picks per inch. One of the union members later recalled: ‘It took us a long time, before we reached any result … We held a meeting once a week and had everything put down and written out and send to the Federation … It took a huge amount of effort before it was completed.’53 Adapting to the new industrial relations clearly was a challenge, involving everyone from rank and file union members to the Federation leadership. During the same period there was a mild recession in the textile industry, and for the first time since its foundation the Federation started to lose members. But the crisis was soon overcome: at the congress of the Textile Workers’ Federation in 1906 a resolution expressing confidence in the Federation leadership was passed unanimously. The severe disagreements in the years around 1900 were replaced by an almost demonstrative display of unity. This was due to one single event: the great strike of 1905. When the national agreement was to be renewed in 1905, the employers made a proposal, aimed at accelerating the introduction of multiple loom systems. The disputes over multiple loom systems will be presented below, and the proposal referred to in more detail. What is important here is that it was turned down by the Federation, and the result was a strike, which was to be the biggest conflict ever in the history of the Textile Workers’ Federation: almost half of the total number of workdays that year were lost.54 The symbolic significance of the 1905 strike to the textile workers was in many ways similar to the significance of the Great Conflict of 1899 to the labour movement as a whole. The strike experience gave a boost to the local organizations. In Vejle, a city with a major cotton industry, the union had suffered severe problems and was in 52 Møller, Dansk Tekstilarbejderforbunds historie, pp. 183–184; Jensen, Dansk Tekstilarbejderforbund, p. 188; Jens Petersen et al., Dansk Tekstilarbejder–Forbunds Afdeling i Odense. Et tilbageblik paa fagforeningens virksomhed i 50 aar, 1885–1935 (Odense, 1935), p. 28. 53 NIHA, no. 1872, pp. 11–12. 54 Danmarks Statistik, Statistiske Meddelelser, IV.30.2; Jensen, Dansk Tekstilarbejderforbund, pp. 195ff. 158

Denmark effect split in two in 1903–1904. But when the strike came, the workers reunited. One of the local activists recalls: ‘organizing sort of gained momentum again. The workers learned for themselves the importance of being organized.’55 As a direct result of the strike, the membership rose markedly, and there was a lasting effect of 40 per cent more members than in the down period of 1902–1903. Approximately a quarter of all textile workers were now organized – and the number grew steadily from that point to the approximate 80 per cent that was to be the normal level after the First World War. The year 1905 was the breakthrough for the Textile Workers’ Federation as a mass organization.

Bargaining, agreements and arbitration After the Great Conflict in 1899 a set of industrial relations gradually evolved that incorporated the trade unions into a modern, capitalist mode of production. The employers and the state accepted the unions and their federations as the legitimate representative body of the workers. Furthermore, the employers accepted that wages and working conditions had to be laid down in agreements between the union federations and the corresponding employer organizations. The use of strikes was not banned, but strongly regulated, and to be used under certain conditions only. The federations were charged with the responsibility of maintaining discipline among their own members, and preventing ‘unlawful’ strikes. The advantage of this system, seen from the point of view of the labour movement, was the introduction of an instrument that could be used to reinforce the agreed upon rights of labour, without the need to resort to strike. The idea of arbitration was originally promoted by the unions. Already in 1887, the Textile Workers’ Federation had taken a principle stand in favour of arbitration. As could be expected, the employers were more sceptical. To them, the notion of legally binding agreements and arbitration was an infringement of their right to run their own business as they saw fit.56 Nevertheless, arbitration gradually became a central element of industrial relations. One of the reasons for this might be that the employers discovered that it could also be a tool for making the unions responsible for maintaining discipline in their own ranks. Arbitration became generally recognized in the general agreement of the textile industry in 1902. It was to be used when new products were introduced to a factory for which proper piece-rates could not be agreed upon between workers and employers. If a local conciliation was unsuccessful, the question could be brought all the way up to a central board of arbitration.57 In the following years there were ongoing discussions between the union federation and the employers’ federation, resulting in gradual changes of the general agreements. It was especially the precise distinction between those types of disagreements where strikes and lockouts were allowed, and those where they were not, that was the subject of discussion. This 55 Dansk Tekstilarbejderforbund, Vejle afdeling 1896–1946 (1946), p. 34. 56 Ibsen and Jørgensen, Fagbevægelse og stat, p. 266. 57 Dansk Textilarbejder-Forbunds Fagblad (1904), no. 11. 159

The Ashgate Companion to the History of Textile Workers principal distinction was adapted to real life through the rulings of the so-called Permanent Court of Arbitration in a series of principal cases. Thus, it was not until 1926 that it was finally made clear that workers in fact had the right to strike if agreement on the piece-rate wage paid for new types of products could not be reached.58 This ruling was one of the victories won by the Textile Workers’ Federation in the Permanent Court of Arbitration. The Federation had learned how to manoeuvre in this new land of modern, industrial relations, and how to use the legal system in the interest of labour. However, the employers had more than their fair share of victories too – they did in fact win most of the cases in the same period.59

Coping with changing conditions of work One of the central challenges to the Textile Workers’ Federation was that of the changing conditions of work. In 1902, the chairman expressed the general attitude of the Federation in this way: We should … go as far as possible, when attempts are made to reach the level of other countries, with the help of new systems and more rational operations, since this is our textile industry’s only hope of salvation from destruction. It is our task to safeguard our own interests, so that new systems are not introduced at the expense of our wages or by prolonging our hours of work. Yet, we, at the same time, should never hinder the evolution, which is necessary to create a new era for our industry.60 The ‘new systems’ that the chairman refers to is probably the multiple loom system, which was discussed quite intensively among textile workers at this time. What seemed to be one of the most decisive factors in the advancement of the system was the problem of agreeing on adequate piece-rate scales. Since weavers were almost exclusively paid piece-rate wages, there would be no real benefit for the employer if a weaver were just paid the normal rate per loom. The employers’ answer to this was to deduct a certain percentage of the piece-rate paid when working in a multiple loom system. For example, as a starting point, the wage of a weaver operating four looms would be calculated as the total sum of what was earned at each loom, according to the piece-rate scale. From that total sum, a certain percentage would then be deducted. In 1905 this deduction would typically be 20 per cent when operating four looms. Furthermore, since a weaver would not be able to operate each of the four looms as effectively as if he was operating only one loom, the total sum paid would not be four times the rate earned at one loom minus 20 per cent, but maybe 2.5–3.5 times the rate of one loom minus 20 per cent. 58 Dansk Arbejdsgiverforening, Den faste Voldgiftsrets Kendelser (1926), case no. 952. 59 Judged from yearly issues of Den fast Voldgiftsrets Kendelser 1916–26. 60 Dansk Textilarbejder-Forbunds Fagblad (1902), no. 9. 160

Denmark It is very difficult to calculate precisely how much more a weaver would earn when operating four looms compared to one, since there are so many variables involved, including technical issues as well as the personal skills of the weaver. Consequently, apart from the level of deduction from the piece-rate, there were also other issues involved in the discussion of the multiple loom system. One issue was the conditions under which the implementation of the system was acceptable to the workers. There were a number of complaints from weavers, who felt that the looms and yarns they worked with were of a poor standard, demanding too much attention from the weaver to be adequate for multiple loom systems. Another issue was the fact that not all weavers were equally interested in, or able to exploit, the possibility of raising their wages through operating more looms. All other factors being equal, the multiple loom system would lead to an increased diversity in wages. In one factory, the weavers in 1915 earned from DKr.12 to DKr.31 per week – with an average of DKr.20.61 There was a fear within the Federation that this would undermine workplace solidarity. The tool that the Textile Workers’ Federation used to cope with these challenges was the enforcement of local and general agreements, obtained through bargaining – sometimes backed by conflict. In 1905, the general agreement was to be renewed. As a precondition for any negotiations, the Employers’ Federation demanded that the Textile Workers’ Federation should accept that three- and four-loom systems could be introduced ‘for all appropriate fabrics’, with a piece-rate deduction of 16.66 per cent or 20 per cent respectively. This provoked the great strike of 1905, mentioned above. The result of the strike was that the textile workers had to accept the deductions from the piece-rate put forward by the employers. Judged from this point alone, the result was a defeat for the Federation. But at the same time, the new agreement stipulated specific preconditions – concerning types of fabrics, yarns and looms – that had to be met in order to implement a multiple loom system. Furthermore, it was agreed that the weavers’ pay should always be raised 10 per cent as a minimum when changing to a multiple loom system.62 The strike of 1905 gave a moral boost to the Textile Workers’ Federation. It was not directed against multiple loom systems as such, but aimed at securing the right of co-determination with regard to the conditions under which the system would be implemented. In this respect, it was a success. The principle of stipulating the conditions for what were actually changes in the work process, in the general agreement, would turn out to be a tool of long lasting importance. In the 1911 agreement, the paragraph concerning multiple loom systems had swollen to more than two pages, determining not only which types of fabrics could be woven on which types of looms with which types of yarns, but also the types of shuttles, cops and so on to be used, as well as how many shots per minute the looms had to be capable of.63 61 Textilarbejderen (1915), no. 9. 62 Jensen, Dansk Tekstilarbejderforbund, pp. 196–200. 63 Ibid., pp. 211–212. 161

The Ashgate Companion to the History of Textile Workers It should be remembered that one of the basic principles of the September Compromise of 1899 was the employers’ right to manage and distribute work in the workplace. What the example of the multiple loom system indicates is that labour was in fact able to fight for co-determination on issues, which, if the September Compromise were to be followed by the letter, would have been the exclusive domain of the employers.

Institutionalizing class struggle As questions regarding working conditions became part of the general agreements, they were also lifted out of the workplace. The struggle between labour and capital at the factory level was subsumed under an institutionalized struggle. The daily experiences of the workers could no longer always be transferred into immediate action, but would often have to go through several levels of ‘the system’ in order to be transferred into some sort of action – if any. The reactions to the Bedaux system are an illustration of this. As mentioned above, the Bedaux system of rationalization was introduced in a few factories from 1933. When the system was introduced in a factory in the town of Silkeborg in 1934, the workers felt so intimidated by the foreign experts who were assigned to do time studies that they eventually refused to cooperate.64 The chairman of the Textile Workers’ Federation went to Silkeborg, where he urged the workers to follow the orders of the factory management, since refusal to cooperate would be a violation of the general agreement.65 The next day one of the workers was fired. This sparked a strike, which soon included all 324 workers at the factory. The strike was ruled illegal by the Permanent Court of Arbitration and the workers incurred a fine. The Textile Workers’ Federation instructed its members to go back to work, an order which they rejected by a 97 per cent majority of the votes in a local referendum. Following the referendum, all members on strike were expelled from the Federation. The Federation leadership branded the strike committee in Silkeborg as communists, and the strike was renounced in the national Social Democratic press. However, the strike had widespread support in the community – 4,000 persons attended a local support rally. Other local branches of the Textile Workers’ Federation also supported the strike. But the strikers eventually ended up having not only the Permanent Court of Arbitration, the Ministry of Justice, the police and the Conservative press, but also the CTU, the Textile Workers’ Federation and the Social Democratic press as their adversaries. After nearly two months, the strike ended in complete defeat. The fierce attacks from the Federation against the workers in Silkeborg for being communists reflected the fact that the strike was questioning the general strategy of 64 Keld D. Larsen, Silkeborg. En udpræget industri– og arbejderby (Silkeborg, 1994) p. 57; Jørgen Peter Christensen, ‘Rationalisering og arbejderne. Dansk industri i mellemkrigstiden’, in Mikkelsen, Produktion & arbejdskraft i Danmark, p. 141. 65 Dansk Tekstilarbejderforbund: Forhandlingsprotokol (1935), pp. 29, 74. 162

Denmark the trade union movement: that the strike weapon had to be submitted to the system of bargaining and general agreement. The system of industrial relations, which had so carefully been built up, had to be defended – even against the decisions and interests of a particular group of members. At the Federation’s congress the following year, the chairman made his position clear: ‘What they did over there was an insane child’s play. They thought it was possible to cross the principles of the Employers’ Confederation, by declaring a strike in a single factory.’66 However a congress delegate from Silkeborg presented another interpretation of what was at stake: ‘The point is, that it (the Bedaux system, LKC) creates dissatisfaction, unrest and unemployment, and ruins the workplace.’67 For the workers in Silkeborg the conflict was not just about a new system of piecerate payment, but about such fundamental issues as lack of respect and dignity in the workplace. There is little doubt that the majority of members agreed in principle with the chairman’s defence of the existing industrial relationship. But that did not necessarily preclude them from feeling that the workers in Silkeborg had been let down. Delegates from the two other factories where the Bedaux system had been implemented complained that they had tried in vain for several months to involve the Federation. The following year, negotiations over the system finally took place between the Textile Workers’ Federation and the Textile Employers’ Federation, but they soon ended without result. As early as the beginning of the century, the discussions about the multiple loom system revealed a strong element of determinism in the Federation’s ideas on technological change. However, the strike of 1905 also reflected the knowledge that the technology issue was basically subject to the same antagonistic interests as all other aspects of working conditions. But the 1920s and 1930s saw a change in Social Democratic ideology. Generally, the party downplayed class antagonisms, in order to reposition itself as a ‘people’s party’. In matters of production and wages, some social democrats were inspired by the ideas of F.W. Taylor, Henry Ford and others, who argued that rationalization and technological progress could create such abundant wealth for everyone that class struggle would be superfluous.68 It is likely that this affected the way in which Federation leadership regarded a phenomenon such as the Bedaux system. More specifically, the Bedaux system was mainly considered from a technical point of view, as a system of payment, rather than an instrument in the struggle for power in the workplace. Furthermore, it is clear that what occupied the Federation leadership most was to try to understand the technical details and implications of 66 Ibid., p. 30. 67 Ibid., p. 59. 68 Michael F. Wagner, ‘Industrialismens verdensbillede – et portræt af den moderne amerikanisme og dens ny teknologi’, Den jyske historiker 35–36 (1986): 41–42; Christensen, ‘Rationalisering og arbejderne’, pp. 130–136 passim; Knud Knudsen, Arbejdskonflikternes historie i Danmark. Arbejdskampe og arbejderbevægelse 1870–1940 (Copenhagen, 1999), p. 244. 163

The Ashgate Companion to the History of Textile Workers the system – which they apparently never did.69 The result of these considerations was an inability to take action – illustrated by the general secretary, who declared at the 1935 congress: ‘if we act very carefully, it might just be silenced to death’.70

Gender and work It is a popular belief that as industrialization and mechanization progressed, skilled labour – mainly in the form of male artisans – was generally replaced by unskilled labour – mainly women and children. However, as demonstrated above, women had been active in textile production long before industrialization. They were to a large extent involved in pre-industrial textile production, primarily within domestic spinning or weaving, which also included production for the market. The technology, qualifications and customers of the female domestic producer were probably more or less the same as those of the male, rural artisan. For this reason, it has been claimed that at least in rural areas the distinction between artisans and non-artisans was in reality based on sex rather than skill.71 Furthermore, industrialization did not necessarily exclude men from textile production, as shown in Table 6.3. From the national censuses we can see that some specific jobs seem to have been clearly linked to a specific sex. In the cotton-spinning plants, the positions of overseer, mechanic and so on were almost exclusively held by men, while carders and spinners were predominantly female. In cotton weaving both sexes could occupy the central function of weaver. However, setting up the loom was mainly a male job, while burling was almost exclusively a female job. In contrast to the cotton industry, spinning in the cloth mills was generally carried out by men. Since the source material for these observations are the national censuses, the figures are aggregated numbers. This means that even though jobs statistically appear to be occupied by both sexes in the industry as a whole, at workplace level they might well be reserved for one sex only. It is important to note that this could be a different sex at different workplaces.72

69 The Textile Workers’ Union commissioned the independent Institute of Technology to make a report on the system (Letter from Teknologisk Institut, 18/12 1932. Dansk Textilarbejderforbunds arkiv, Arbejderbevægelsens Bibliotek og Arkiv (hereafter DTA/ ABA) folio no. 403). On the basis of this report, the union leadership concluded that implementation of the Bedaux system would always lead to higher wages. This was later denied by both the engineer who had written the report and the Bedaux company (Letter from Chr. Randstrup, Teknologisk Institut, 25/2 1936, and copy of letter from K. Gartner, Bedaux & Co. to the management of Holger Petersen’s Fabrik, 1/9 1936. Both in DTA/ABA, folio no. 403). 70 Dansk Tekstilarbejderforbund, Forhandlingsprotokol (1935), p. 123. 71 Marianne Rostgård, Kønsarbejdsdeling og teknologiudvikling i dansk tekstilindustri (1991), p. 48. 72 This is confirmed by a recent study of the cloth mills: Tønsberg, Den danske klædeindustri, pp. 120–121. 164

Denmark Table 6.3 Women as percentage of total workforce in different textile branches and jobs 1906

1935

Textile industry (knitwear excluded)

56

61

Cotton-spinning plants

82

81

Cotton-weaving plants

60

60

63

55

42

50

Spinners

30

47

Weavers

44

33

Weavers Cloth mills

Source: Statistisk Tabelværk, V.A.7, table IX (1906) and V.A.21, table VB (1935).

There is no single cause for the distribution of job functions in relation to sex within the textile industry. As women carried out some of the most mechanized jobs (such as cotton spinning) as well as some of the least mechanized (such as burling), no simple relationship between mechanization and ‘feminization’ existed. What happened during industrialization was not simply that women substituted men, but that traditional concepts of male and female labour were partly carried on in new forms, partly uprooted and rearranged.73 As industrialization progressed, and the technology and organization of work changed, there was an ongoing process of social construction as to what were considered ‘male’ and ‘female’ jobs. The changing social organization of labour, and the establishment of modern industrial relations based on collective bargaining, formed a crucial framework and instrument for this process. The industrialization of textile production led to an immensely increased division of labour. Weaving alone became divided into a number of separate processes, such as pattern-drawing, warp-cutting, sizing, beaming, threading and so on, apart from the actual tending of the loom. In a cloth mill, the complete process of production was divided into at least 15–20 major job functions. The motivation for this was primarily technical: it was a prerequisite for mechanization and a rational exploitation of the technology. However, the division of labour also 73 The most comprehensive study of this, regarding the Danish textile industry, was carried out by Marianne Rostgård in her (unpublished) Ph.D. thesis (1991), as well as in several articles, such as ‘Konstruktion af kønsarbejdsdeling i dansk tekstilindustri’, in Hans Buhl and Henry Nielsen (eds) Made in Denmark? Nye Studier i dansk teknologihistorie (Århus, 1994) and ‘Fagbevægelsen og de kvindelige industriarbejdere’, Arbejderhistorie 4 (1997): 36–53. 165

The Ashgate Companion to the History of Textile Workers had an important side effect: it made it possible to divide the job functions into ‘male’ and ‘female’ with a high degree of detail.74 The gender division of labour was a result of decisions principally reached within the domain of the employers, but gradually also made subject to collective bargaining. Tradition played a role in the decision-making of the employers, especially in the early phase of industrialization. There was also a traditional awareness of the gender attached to certain jobs among workers. Male artisans showed a hostile attitude towards taking jobs in the early cotton industry. Since other artisans were tending mechanized looms in the cloth mills, this animosity cannot only be explained as a reaction to mechanization as such. It was also directed against the gender identity attached to a specific job.75 Nevertheless, even if tradition had its part to play, it could only be as far as it fitted into the general rationality of management practice. In a market economy, two things generally govern the recruitment of labour: its availability and its price in relation to productivity. The price of labour, or, as seen from a worker perspective, the level of wages, was the centrepiece in the struggle over gender and work. In the period when the basic principles of industrial relations in the textile industry were formed, the husband was generally considered the family breadwinner, among both workers and employers. Consequently, a woman could legitimately be paid a lower wage than a man could.76 However, the lower wages of women could also be a disadvantage, seen from a male worker’s point of view, since they gave the employers an incentive to substitute female for male labour. It was this aspect that was most frequently stressed when gender and wages were discussed in the first half of the twentieth century. The discussion seems to have popped up regularly, sometimes provoked by attempts made by the employers to increase wage differences. This was the case from 1916 to 1919, when the Federation leadership accepted employer demands that led to temporary gender-based differences in the piece-rate scales. This was met by angry reactions from some of the Federation’s female members, and the 1919 Federation congress demanded that the leadership should ‘constantly work for the fulfilment of the following principle: equal payment for equal jobs’.77 This principle was formally enforced in jobs paid according to a piece-rate, such as weaving. Nevertheless, even in these jobs women generally earned 30 per cent less than men. The main reason for this was that men simply operated the types of looms for which the best piece-rate wages were paid. In 1937, figures show that 27 per cent of all male weavers operated automatic looms – for which the highest wages were paid – in contrast to only 4 per cent of all female weavers.78 74 Rostgård, ‘Konstruktion af kønsarbejdsdeling’, p. 233. 75 For an example, see NIHA, no. 1535. 76 Rostgård, ‘Fagbevægelsen og de kvindelige industriarbejdere’ provides examples of employers who argued for a lower wage for women, see p. 47. 77 Jensen, Dansk Tekstilarbejderforbund, p. 286. 78 Textilfabrikantforeningen: Gennemsnitsfortjenesten i September 1937. Archives of Dansk Tekstilarbejderforbund, folio no. 217. 166

Denmark Women, men and the union Female textile workers became union members at a slower pace than their male colleagues did. In 1897, approximately 25 per cent of male workers were union members, but only 10 per cent of female workers.79 This difference persisted until around the First World War – in the year 1914 the rates were 51 per cent and 32 per cent respectively. After the First World War the difference was ironed out, and the rate of organization reached a relatively stable level around 80 per cent for both men and women. These figures do not necessarily mean that female textile workers were more indifferent to their working conditions. There were a number of conflicts in the early period of the Federation in which women played a role, most noticeably a rather spectacular strike at Rubens Dampvæveri in Copenhagen in 1886, which involved 225 female workers. However, the figures reflect that men and women, because of their roles in society as such, to some extent had different experiences as industrial workers, and to a certain extent developed different work ethics and strategies. Most notably, women were responsible for family reproduction and childcare. The labour movement was a long-term project. But children could not be fed with promises of shorter working hours and daycare centres sometime in the future. At least until the trade union movement was strong enough to enforce agreements, elect shop stewards and so on, it was not necessarily the obvious choice of all women. Short-term solutions such as preparing food in the workplace, sneaking away during breaks to take care of children and other such ‘undisciplined’ actions could in fact be considered a more rational choice. Trade unions were based on a certain work ethic, which is expressed in one of the most popular slogans of the Danish trade unions: ‘Do your duty – demand your rights!’ The ideal worker was one who attended to work in a disciplined and responsible manner, thus making him a respectable citizen with a moral right to decent treatment and to participate in the decision-making process. Women, such as those depicted above, displayed a work ethic which was inconsistent with this ideal. Many male workers probably did so too. Nevertheless, it is noticeable that male union activists of this period often explained such discrepancies in terms of gender: women act spontaneously and short-sightedly because of their sex, thus creating problems for the union. The above-mentioned strike at Rubens Dampvæveri did in fact create a lot of trouble. It ended in defeat, and internal grievances resulted in a split within the union. In his memoirs, J.J. Møller, chairman of the Textile 79 This and the following numbers are calculated on the basis of Møller, Dansk Tekstilarbejderforbunds historie, Appendix, and Jensen, Dansk Tekstilarbejderforbund, p. 482 (membership) and Statistisk Tabelværk V.A.1, table 2; V.A.7, table VIII; V.A.12, table III; V.A.18, table III+IV; V.A.21, table IIIA+B (total number of workers). Because of the nature of the source material, these numbers include members in the knitwear industry. For a further discussion of the method of calculation, see Christensen, Det moderne arbejde, p. 332. 167

The Ashgate Companion to the History of Textile Workers Workers’ Federation until 1918, describes the strike at Rubens Dampvæveri in 1886 as initiated by a ‘hot-blooded’ woman.80 When a spontaneous strike by female spinners in Vejle created problems for the Federation in 1906, Møller used the exact same phrase, describing the female workers as ‘hot-blooded young women, who can be swiftly brought into ecstasy’. He took the opportunity to reiterate the need to maintain discipline among the rank and file.81 In the long term, it became evident that the labour movement was, in fact, able to not only secure better wages and working conditions for both male and female workers, but also improve the everyday life of working-class families through housing, daycare facilities, better schools and so on. Between 1914 and 1925, women rallied in the Textile Workers’ Federation. Nevertheless, in the decisive years at the end of the nineteenth and beginning of the twentieth century, when wages and piece-rate scales were formalized, agreements made and certain categories of jobs marked out as either male or female, women’s interests were clearly underrepresented in the Federation. Consequently, the best-paid jobs were defined as male – and vice-versa.82

Post-war decline The decline in employment after 1950 was of course also reflected in a decline in membership of the Textile Workers’ Federation. At the end of the 1960s, the membership was only half of what it had been in 1951. Furthermore, during the 1960s there was a great interchange of members – the number of members moving in and out of the Federation between 1961 and 1965 was greater than the average total number of members, meaning that statistically the whole membership had been renewed in this period.83 The reason for this was probably that work in the textile industry had come to be regarded as low status – the kind of work one only sticks with until something better crops up, and in times of generally high employment such as the 1960s something better had a tendency to crop up regularly. This was the backdrop against which the Textile Workers’ Federation had to act in the post-war period. In 1950 there was a conflict of principle in a small provincial factory over the right to organize.84 However, apart from this incident, the two decades after the war seem to have been a rather calm period for textile workers. This was also the case for the rest of industry: the level of conflict was low, compared to the interwar period. Still, a couple of major conflicts did take place in connection with the renewal of general agreements, but even here, the Textile Workers’ Federation seems to have played a rather minor role. In the 1970s, however, the general level of conflict in Danish industry increased. In 1973, the 80 Møller, Dansk Tekstilarbejderforbunds historie, p. 115. 81 Dansk Textilarbejder-Forbunds Fagblad (1906), no. 10. 82 Rostgård, ‘Konstruktion af kønsarbejdsdeling’, p. 233. 83 Christensen and Jensen, Textil og skotøj i 100 år, p. 48. 84 Ibid., p. 44. 168

Denmark Danish Employers’ Confederation refused a proposal for a new general agreement, and the textile workers were locked-out as part of the contest that followed. This was the first time since 1936 that the whole textile industry had been in conflict with the employers. Apart from this, the Textile Workers’ Federation was not involved in any significant conflicts, even in this turbulent period. This probably reflected the rather defensive role the Federation was forced into by the structural changes taking place within the textile industry. A major issue in the internal debates of the Textile Workers’ Federation from the mid-1960s and onwards was the need to unite with other, similar unions, to form a more effective organization. This goal was achieved in 1977, when textile workers united with workers in the clothing industry to form the Danish Clothing and Textile Workers’ Federation. In 1981 workers of the footwear industry joined them. Finally, in 1998, the textile workers became part of the much larger General Workers’ Union and the Clothing and Textile Workers’ Federation was dissolved after 113 years of existence.

169

This page has been left blank intentionally

7 Egyptian textile workers: from craft artisans facing European competition to proletarians contending with the state Joel Beinin

In the last 20 years voluminous scholarship on Egypt and other Arab provinces of the Ottoman Empire has undermined the traditional Orientalist thesis that the seventeenth and eighteenth centuries were a period of economic, political and cultural stagnation and decline. Nonetheless, this thesis persists as the default explanation for why there was no industrial revolution, no development of modern science and technology, and other absences compared to the presumed normative case of Europe (which usually means England). For example, David Landes’s global economic history relies on crude culturalist arguments to explain the failure of the Arab Middle East to follow the developmental trajectory of Europe. He argues, anticipating the title of a highly touted book by Bernard Lewis, that ‘History had gone awry.’ As textile manufacturing was the leading sector in England’s industrial revolution, it seems logical to investigate that sector in Egypt to inquire if differences in technique or the social organization of production before the 1740s to the 1790s The classic statement of this thesis is H.A.R. Gibb and Harold Bowen, Islamic Society and the West: A Study of the Impact of Western Civilization on Moslem Culture in the Near East (Oxford, 1950). See also Bernard Lewis, The Emergence of Modern Turkey, 3rd edn (New York, 2002).  David S. Landes, The Wealth and Poverty of Nations: Why Some are so Rich and Some so Poor (New York: W.W. Norton, 1998), especially ch. 24 entitled ‘History Gone Wrong?’ The quote is from p. 394. The title of Bernard Lewis’s book What Went Wrong: Western Impact and Middle Eastern Response (Oxford, 2002) suggests a similar argument. 

Map 7.1 Egypt

Egyptian textile workers – when the mechanical devices that ultimately transformed the English textile industry were invented – might explain the divergent developmental trajectories of the two countries. This is a false logic based on the flawed assumption that economic and social development is a modular process that can be transferred from one society to another. Moreover, it ignores the global dynamics that prompted England’s industrial transformation and provided a good part of the capital to sustain it. The British imperial project in India brought cheap, high quality Indian cottons into competition with English linens and woollens. This encouraged the shift from human labour to machine power in England, while the abundance of cheap coal was among the important factors that made this possible. Wealth obtained from ruling India provided a substantial part of the capital to finance this shift, while overseas colonies constituted a captive market for machine-manufactured cottons. Egypt’s textile industry did not and could not replicate this trajectory. It should be understood in its own context, not as a series of failures and absences compared to Europe. Egypt was historically a producer and exporter of cotton, linen, wool and silk. Indian cottons did not pose a significant threat to its textile industry. In Egypt, like India but unlike England, food and labour power were cheap. Hence, there was no pressure to mechanize textile production. Long-distance merchants, the main possessors of liquid capital in the seventeenth and eighteenth centuries, did not invest in organizing textile production, in all likelihood because greater profits could be made with less risk by trading in luxury goods. By the time machinewoven European cottons began to compete with Egyptian textiles, the established advantage of England and France was already an obstacle to Egypt emulating their developmental trajectory. Until about 1820 Egyptian textile production was characterized by small workshops, guild organization of craft workers and a highly specialized division of labour. Cottons and linens were exported – primarily to other provinces of the Ottoman Empire and neighbouring Muslim regions. In 1816–1818 Mehmed ‘Ali, the Ottoman governor of Egypt, effected a partial shift in the techniques and social organization of production by establishing state-owned textile workshops. The second half of the nineteenth and early twentieth century was a transitional period when handicraft spinners and weavers struggled to adjust to the influx of European yarn and cloth. A more substantial transformation of the forces and social relations of textile production began with the first successful mechanized textile mill in 1912 and the first Muslim Egyptian-owned mechanized mill in 1927. Concurrent with the growth of the mechanized textile industry, from the late 1930s until the mid-1950s textile workers organized in trade unions which became a significant force in the nationalist movement demanding evacuation of the British occupation  

This argument is based on Robert B. Marks, The Origins of the Modern World: A Global and Ecological Narrative (Lanham, Md., 2002), pp. 95–111, which presents a brief summary of recent scholarship placing England’s industrial revolution in its global context. André Raymond, Artisans et commerçants au Caire au XVIIIe siècle (Damascus, 1973– 1974), p. 225. 173

The Ashgate Companion to the History of Textile Workers army from the Nile Valley. The authoritarian-populist regime of Gamal Abdel Nasser nationalized most of the textile industry in the early 1960s. A combination of material and social gains and repression rendered textile workers politically quiescent. They have participated prominently in recurrent waves of protest against neo-liberal economic policies since the mid-1970s. But textile workers are no longer an organized political force or a symbol of industrial modernity, as they were from the 1930s to the mid-1950s.

Textile production and trade in the seventeenth and eighteenth centuries In the seventeenth and eighteenth centuries textile production constituted the leading artisanal activity in Egypt, in respect of both the number of people employed and the value of their product. In the 1670s the Ottoman traveller Evliya Celebi (1611–1682) observed that about one-fifth of Cairo’s artisans (12,102 out of 59,214) worked in the textile crafts organized in 18 of the city’s 136 guilds. André Raymond’s magisterial Artisans et commerçants au Caire au xviiie siècle is the source of most of what we know about the social history of eighteenth-century Egypt. Over one-third (97 out of 283) of the wills deposited in the Cairo shari’a court Raymond studied were those of textile artisans. On this basis he estimated that they comprised a quarter to a third of all the craft workers of Cairo. Extrapolating from the estimates of the French scholars who accompanied Napoleon’s 1798 expedition to Egypt and assuming that the relative position of textiles remained roughly the same as in the previous 125 years, there would have been some 5,000 textile artisans in Cairo at the end of the eighteenth century. In addition to spinners and weavers of cotton, silk, linen and wool, the textile sector included carders, dyers, lace-makers, embroiderers and tailors. Spinning and other preliminary activities were often undertaken by urban or rural women and children. Women often worked in the craft of their husband. But they usually did so at home and were not guild members, so there are few records of their activity. Muslims, Copts and Jews were all engaged in silk weaving in Cairo, and a number of Copts were also tailors. Cairo was the principal textile production and distribution centre specializing in the weaving of silk, cotton and linen. Fabrics manufactured in upper Egypt were distributed through its port of Bulaq. Dyeing was a semi-industrial specialty of Cairo. Workshops with an average of 20 workers were concentrated in three main areas of the city according to Evliya Çelebi. Weaving was dispersed throughout

 Ibid., pp. 204, 229.  Ibid., pp. 457–458, 460. 174

Egyptian textile workers Cairo in workshops of only a few artisans. After coffee and spices, textiles were the most important product sold in Cairo’s markets. Several provincial towns were major textile producers. Mahalla al-Kubra, Mansura, Rosetta, Damietta, Minuf, Fayyum, Bani Suwayf and Asyut undertook cotton and linen weaving, while Mahalla al-Kubra and Cairo were the leading centres of silk weaving. Raw silk from Mount Lebanon was imported through the port of Damietta, which was also a silk-weaving centre. Bani Suwayf had 600 textile artisans, and Qena 250, probably working in wool. The most important provincial textile centre was Mahalla al-Kubra, located in the central Nile Delta. At the end of the eighteenth century it was Egypt’s third largest city after Cairo and Damietta with a population of 17,500 and over 2,000 textile artisans.10 They wove locally supplied cotton and linen and imported silk.11 There was generally no separation between the processes of production and distribution of textiles. Typically a male artisan, using little working capital, owned or rented his own shop and tools, bought raw materials, and produced and sold commodities on demand using his own labour, that of family members and a small number of journeymen or apprentices. Evliya Çelebi observed an average of 3.5 workers per artisanal workshop in Cairo in the 1670s. This changed little through the early twentieth century. In the textile sector, only the workshops of rug weavers and dyers exceeded this average. In Cairo and its urban extensions, as in several other major cities of the Ottoman Empire, production and commerce was organized in guilds. Workers in each process and product of the textile industry were organized into separate guilds.12 Each guild was headed by a master or shaykh who was chosen by a combination of election by the membership and appointment by the Ottoman governor. The guilds were not centrally controlled, and there was great variety in their internal regulations and practices. By the end of the eighteenth century a certificate (gedik) was required to engage in most urban occupations in Cairo.13 But the guilds never exercised absolute control over the quality of commodities or techniques of production. The guild structure was sufficiently flexible to accommodate new crafts and production processes.14  Ibid., pp. 317–319.  Ibid., pp. 229–230.  Roger Owen, Cotton and the Egyptian Economy, 1820–1914: A Study in Trade and Development (Oxford, 1969), p. 13. 10 Sawsan El-Messiri, ‘Class and Community in an Egyptian Textile Town’ (Ph.D. diss., Hull University, 1980), p. 58; Owen, Cotton and the Egyptian Economy, p. 12. 11 El-Messiri, ‘Class and Community’, pp. 57–58 relies on the observations of DuboisAymé, a scholar who accompanied the French expedition of 1798, to claim that Mahalla was the only locale for silk weaving in Egypt, and that its products were exported to other Ottoman regions. This is inconsistent with Raymond’s very solid information on Cairo’s silk weavers. 12 Raymond, Artisans et commerçants au Caire au XVIIIe siècle, pp. 213–215, 222. 13 Ibid., pp. 549–550. 14 Ibid., pp. 225, 584–585. 175

The Ashgate Companion to the History of Textile Workers Nor did guilds block expansion of production in the face of competition from European manufactures. Throughout the eighteenth century Egypt’s textile production was of sufficient volume and quality to be exported to Istanbul and other Ottoman provinces, North Africa, the interior of Africa and Europe. Locally manufactured cotton and linen cloth made up about one-fifth of Egypt’s total exports.15 Linen produced in Asyut, Manfalut, Abu Tig and Cairo and especially cotton woven in Cairo, Mahalla al-Kubra and Rosetta were the only textiles exported to Europe in the eighteenth century. France imported over 90 per cent of these textiles and then re-exported a significant quantity to Holland and Spain.16 Textile exports to Europe peaked in 1730–1739, when 30.5 per cent of all exports to France consisted of textiles. European consuls explained the subsequent downturn as partly due to the declining quality of Egyptian textiles and partly due to the unreasonable demands of the Marseilles Chamber of Commerce. Exports rose somewhat in the last quarter of the eighteenth century; a French scholar accompanying the Napoleonic expedition estimated that 9.6 per cent of the value of all of Egypt’s exports to Europe was composed of textiles – two-thirds cotton and one-third linen.17 While trade with Europe was still a relatively minor factor in the economy of mid-eighteenth-century Egypt, textiles constituted more than half of all imports from Europe in 1776–1781.18 Hence, Egyptian textile artisans were among the first to be affected by competition from European manufactured goods. This competition seems to have obliged them to maintain relatively low wages and prices. The absolute value of textile workers’ legacies studied by André Raymond and Pascale Ghazaleh declined (in constant paras) from 50,387 in 1679–1700 to 34,165 in 1776–1798 to 20,864 in 1799–1849. Silk weavers comprise the largest single group of artisans whose legacies Raymond examined. The value of their estates declined from an average of 67,151 constant paras in 1679–1700 (N = 25) to 30,304 paras in 1776–98 (N = 21). In the last two decades of the seventeenth century the incomes of textile artisans were average for artisans of the time (though the elite silk weavers left legacies that were more than one-third greater than the average). Despite competition from Europe, their estates rose above the general average for artisans in the last quarter of the eighteenth century; and they declined in the first half of the nineteenth century, probably due to state intervention and competition from imported British machine-manufactured cottons.19 Despite constituting the sector most exposed to competition from European imports, textile artisans continued to make a reasonable, though not a lucrative living, and handicraft production continued to dominate the local market well into the nineteenth century.

15 Ibid., pp. 185–198. 16 Ibid., p. 181. 17 Ibid., pp. 180–182. 18 Ibid., pp. 193, 197. 19 Ibid., pp. 229–230; Pascale Ghazaleh, Masters of the Trade: Crafts and Craftspeople in Cairo, 1750–1850, Cairo Papers in Social Science 22 (Cairo, 1999), pp. 84–85, 134. 176

Egyptian textile workers

Textile workers and the state in the nineteenth century In 1800 the French occupation force in Cairo appointed a Frenchman as Director of Guilds (mudir al-hiraf). A list he compiled in January 1801 enumerates 278 guilds in Cairo, Old Cairo, Bulaq and Giza. Eleven of the guilds were textile-related, including weavers of Cairo and Bulaq, weavers of Giza, four different guilds of dyers differentiated by colour and type of fabric, tailors, makers of silk braids and spinners of wool. This list is incomplete, as it does not include guilds mentioned by the principal Egyptian chronicler of the period, al-Jabarti, and parts of the list itself are illegible.20 This effort to impose tighter state regulation on the guilds foreshadowed much more comprehensive intervention in the economy by Mehmed ‘Ali – an ethnic Albanian officer in the Ottoman army who participated in the Anglo-Ottoman campaign to expel the French from Egypt. In 1805, after establishing himself as governor of Egypt with support from Cairo’s guilds, he proceeded to centralize all power in his hands. He established a government monopoly over all manufacturing and foreign trade, instituted a forced industrialization programme and made his family the hereditary rulers of Egypt until 1952. The centrepiece of Mehmed ‘Ali’s statist industrial programme was a textile workshop in the Khurunfish quarter of Cairo opened sometime between 1816 and 1818.21 The enterprise first produced silks, but was soon converted to cotton spinning and weaving. That decision was probably connected to the development of Egyptian long-staple cotton, which is well suited to mechanical looms because of the strength of its fibre. Machines, skilled workers and advisors were brought from Europe to Khurunfish. Existing workshops were closed, and the artisans compelled to work there. A second textile workshop was established at Bulaq. In the 1820s steam engines imported from England were installed at these two Cairo workshops and a flax mill in the Delta town of Mansura. After briefly experimenting with using black Sudanese slaves, the state forcibly recruited workers for its workshops with the assistance of the headmen of the quarters of Cairo. There were three categories of workers: daily paid, monthly paid (mostly not artisans or manual workers but bureaucrats) and piece-rate workers. Salaries varied widely. Europeans received more than Egyptians. Skilled weavers working on piece-rate could earn 4 piasters (160 paras) a day – less than they earned previously on their own account. Peasants employed in workshops outside Cairo earned 1–2 piasters a day.22 In part due to the low wages offered, workers often ran

20 André Raymond, ‘Une liste de corporations au Caire en 1801’, Arabica 4, (2) (May 1957): 150–63. 21 Pascale Ghazaleh, ‘Manufacturing Myths: Al-Khurunfish, a Case Study’, in Nelly Hanna (ed.), Money, Land and Trade: An Economic History of the Muslim Mediterranean (London, 2002), pp. 124–138; Roger Owen, The Middle East and the World Economy, 1800– 1914 (London, 1981), pp. 69–70. 22 Judith E. Tucker, Women in Nineteenth Century Egypt (Cambridge, 1985), pp. 75–76. 177

The Ashgate Companion to the History of Textile Workers away or otherwise resisted labour in the state enterprises. The Cairo authorities often asked guild masters to help retrieve workers who fled. In 1823 between 1,188 and 2,000 workers were employed at the Khurunfish workshop. By the early 1830s they operated 370 carding machines, 100 mulejennies and 300 looms, not all of which were in working order. Work discipline was maintained by three overlapping control systems. Guild members were supervised by their masters. Government officials of the Administration of Spinning and Weaving oversaw all aspects of production. Ultimately, soldiers were available to enforce discipline. In 1840 workers from factories in Alexandria, Cairo and Mahalla al-Kubra were pressed into the army, an expression of the militarization of labour control in the state enterprises.23 Even after the limited introduction of machinery, just as in the early stages of the industrial revolution in Europe, there was no radical transformation of the techniques or social organization of production at Khurunfish and the other staterun textile workshops. Guild members had previously worked for wages and rented tools and premises. At Khurunfish the ‘putting-out system’ was employed, just as in England, where this remained the prevalent method of manufacture as late as the 1820s. Hence, Pascale Ghazaleh concludes, ‘Khurunfish … may be seen as a group of guilds working on the same premises. But these guilds were supervised by government officials and produced goods which contributed to the creation of one final product – textiles.’24 The major innovation was that the state owned the premises and the means of production and organized the labour process. The Khurunfish workshop was the start of an extensive state-owned textile manufacturing programme. By 1830 there were some 30 textile workshops in Cairo and throughout lower and upper Egypt including a tarbush (fez) factory, spinning and weaving mills for cotton, jute, linen and silk, and a textile bleaching and printing works. Provincial workshops spun yarn using raw materials and equipment supplied from Cairo. They would then return the yarn to Cairo for weaving or export. The two textile workshops in Mahalla al-Kubra, each with 120 mule-jennies, 60 carding machines and 200 looms, unlike the other provincial workshops, undertook all the stages of producing finished cottons.25 This was enabled by the abundance of skills in this historic textile centre. The tarbush factory established in 1825 produced 24,000 wool hats a month of a quality equal to the established Tunisian competition.26 Flax and wool spinning and weaving workshops were dispersed throughout the country.27

23 Ghazaleh, ‘Manufacturing Myths’, p. 130; Tucker, Women in Nineteenth Century Egypt, pp. 73, 76. 24 Ghazaleh, ‘Manufacturing Myths’, p. 134. 25 Moustafa Fahmy, La Révolution de l’industrie en Égypte et ses conséquences sociales au 19e siècle (1800–1850) (Leiden, 1954), pp. 25–26. 26 Owen, Cotton and the Egyptian Economy, p. 71; Tucker, Women in Nineteenth Century Egypt, p. 73 says 60 hats a day. 27 Tucker, Women in Nineteenth Century Egypt, p. 73. 178

Egyptian textile workers The textile workshops altogether employed 1,381 mule-jennies (each with 180–220 spindles), 1,124 carding machines and 1,759 looms.28 The mule-jennies were operated by animal power, except in three workshops where steam engines imported from England were employed. Most of the machinery for these enterprises was built in Egypt with imported tools under the supervision of the Frenchman who developed long-staple cotton, F. Jumel. Some 12–15,000 workers were employed in these textile mills (the Egyptian nationalist school of history claims a much larger figure).29 An indeterminate number of the workers were women. The tarbush factory was established with the assistance of Tunisian women, who taught Egyptian women how to spin and net the wool into hats. By the 1830s most of the work was done by Egyptian women and girls. The dominance of women in the tarbush factory may be due to this being a new industry for Egypt. Their presence did not violate the established gender division of labour. A similar example of employing females in a new situation is the dispatch of eight women and four men to Syria to establish a cloak factory to supply the Egyptian occupation army.30 Women and girls typically worked in the state workshops cleaning and spinning cotton and flax – tasks they formerly performed at home – for lower wages than men. Most of the mule-jennies and power-looms were run by men. The carpenters, blacksmiths and turners employed to fabricate and repair machines were all men. Judith Tucker succinctly summarizes the gender division of labour in Mehmed ‘Ali’s textile workshops: ‘Female and child labor was more heavily used in animal-powered factories, on smaller machines, and in less mechanized and auxiliary tasks.’31 A combination of the effects of the international recession of 1836–1837, the state’s limited administrative capacity, lack of a local source of fuel and British opposition put an end to Mehmed ‘Ali’s monopoly system and industrial programme. In 1834 spinners of yarn, weavers of cloth and makers of mats were permitted to sell their products on the open market after paying a tax.32 The entire monopoly system was dismantled, and free trade was imposed when the 1838 Anglo-Ottoman Trade Convention was applied to Egypt following the ouster of Mehmed ‘Ali’s army from Syria in 1840. The Treaty of London drastically reduced the size of the Egyptian army, the main consumer of textiles and other products of the state workshops, from 130,000 to 18,000. Nearly all the textile workshops established in the 1820s, which had absorbed one-sixth to one-fifth of Egypt’s cotton harvest at the high point of their operation in the 1830s, lay idle by the 1840s. While Egypt abandoned mechanized cotton spinning and weaving, cultivation of long-staple cotton expanded dramatically. During the American Civil 28 Fahmy, La Révolution de l’industrie en Égypte et ses conséquences sociales au 19e Siècle, p. 24. 29 Owen, The Middle East and the World Economy, pp. 69–70. 30 Tucker, Women in Nineteenth Century Egypt, pp. 87–88. 31 Ibid., p. 88. 32 Ibid., p. 73. 179

The Ashgate Companion to the History of Textile Workers War (1861–1865) cotton became the principal crop of most landowners in Egypt. Ginned and pressed cotton was exported mainly to England.

Machine-manufactured imports and local craft production In the 1840s textile production declined throughout the Ottoman Empire. Cheap European machine-manufactured cottons flooded the region, assisted by the low import duties imposed by the Anglo-Ottoman Trade Convention and similar agreements. Textile producers subsequently recovered by finding market niches that European goods did not satisfy and by accepting lower wages and prices. At mid-century textiles may well still have been the largest single commodity produced in Egypt. Due to the uneven effects of competition from imports, some textile crafts endured with little change. Local dyeing, using natural indigo imported from India and other materials, survived because much of the imported English cotton cloth was undyed. In 1868 there were 1,703 members of the dyers guild in Cairo – over one-third of all textile craft workers – working in 441 workshops. Large quantities of flax and linen, commodities not imported to Egypt, were brought to Cairo from the countryside and woven into finished cloth.33 Silk weaving continued in Mansura and Damietta, probably because the weavers worked for exceptionally low wages to undercut imported cottons.34 But silk weaving declined sharply in Mahalla al-Kubra, which became a centre for collecting, ginning and pressing cotton for export.35 There were, however, significant changes in the social structure of textile production. European imports began to dominate the market for luxury goods. This may have resulted in the decline and even disappearance of certain crafts and guilds even earlier than the mid–nineteenth century.36 The historic primacy of textile craft workers in major cities declined. Around 1868–1873 there were about 4,400–4,500 weavers and dyers in Cairo constituting only 14–18 per cent of its artisan population.37 Three ‘modern sector’ textile establishments employed 1,617 workers.38 Widespread consumption of cotton cloth and yarn imported from

33 John Chalcraft, The Striking Cabbies of Cairo and Other Stories: Crafts and Guilds in Egypt, 1863–1914 (Albany, 2004), p. 57. 34 Ibid., p. 56. 35 El-Messiri, ‘Class and Community’, pp. 65–66. 36 Raymond, Artisans et commerçants au Caire au XVIIIe siècle, pp. 508–512. 37 According to the unpublished 1868 census of the governorate of Cairo discovered by Chalcraft, the city’s 4,406 weavers and dyers comprised less than one-sixth (14 per cent) of its artisans, Striking Cabbies of Cairo and Other Stories, p. 55. The 1873 Statistique d’Égypte suggests a slightly less dramatic decline in the proportion of textile artisans: 4,529 male textile workers (18 per cent) out of 25,149 male artisans in Cairo. 38 Owen, The Middle East and the World Economy, pp. 148–149. 180

Egyptian textile workers England in the 1860s and 1870s by popular strata of the population diminished demand for hand-woven cottons.39 Textile production was maintained by expanded use of the putting-out system and feminization and ruralization of production. Home production by women in the nineteenth century took three different forms. Merchants or hand-loom weavers might put out raw cotton or linen to peasant women who cleaned and spun it. Some peasants bought raw cotton or linen on their own account and sold the spun product to weavers. Other peasant families spun and wove cotton cloth for their own use. Peasant families also cleaned, spun and wove wool, which was the primary fabric for peasant wear.40 Imported European fabrics affected the market for specialty items produced and worn by women much less than they did the spinning and weaving of standard cotton cloth. Women continued to work at embroidery and dressmaking. Girls were apprenticed in their teens, and then became seamstresses working on their own account out of their homes or in a dressmaker’s shop. Legal opinions (fatwas) studied by Judith Tucker indicate that the possibility of a woman earning a lucrative income as a seamstress was recognized.41 The 1873 Statistique d’Égypte reveals that while textile crafts declined in the principal cities, they continued to flourish in the provinces. It enumerated 6,073 men employed in the textile sector in Cairo and Alexandria and 22,187 in the rest of Egypt.42 In the provinces textile workers comprised nearly one-third of all artisans – roughly their historic proportion in Cairo. An even greater proportion of the weavers were located in the provinces. The 1872 census listed only 685 weavers in Cairo and Alexandria and 16,997 in the rest of the country.43 While spinning and weaving were becoming feminized and ruralized, the garment trades were expanding and concentrated in the cities. According to the unpublished 1868 census of the Cairo governorate, 16 per cent of all artisans (4,874) were tailors, seamstresses, shirt-makers and shoemakers – a higher proportion than spinners, weavers and dyers. The 1872 census lists 1,648 tailors in Cairo and Alexandria, but only 840 in the rest of Egypt.44 The British occupation of 1882 accelerated Egypt’s integration into the world capitalist market and, consequently, the adaptation of the textile and garment trades along previously established lines. Employment in the garment trades grew faster than any other sector in the late nineteenth and early twentieth centuries. Tailors, seamstresses, shirt-makers and shoemakers comprised 10 per cent of all Egyptian craft workers in 1897 and 30 per cent by 1917. The number of dressmakers

39 Chalcraft, Striking Cabbies of Cairo and Other Stories, p. 55. 40 Tucker, Women in Nineteenth Century Egypt, pp. 84–85. 41 Ibid., pp. 86–87. 42 Owen, The Middle East and the World Economy, p. 149. 43 Chalcraft, Striking Cabbies of Cairo and Other Stories, p. 56. 44 Ibid., pp. 57–58. 181

The Ashgate Companion to the History of Textile Workers and other makers of women’s clothing grew even more rapidly than the garment trades as a whole.45 The growth of the garment trades was due to the low cost of local labour, the low price of imported cloth, and minimal competition from imported ready-made clothes, which were consumed only by a small Europeanized elite. Egyptian tailors earned above average incomes for craft and service workers by copying European designs. Their low wages relative to European standards enabled them to undersell imported garments by a wide margin. For example, overcoats imported from Austria, which won a certain share of the local market in the early twentieth century, sold for E£1–3. Male tailors in Cairo bought imported English cloth, buttons and lining for 30 piasters and sold the finished product for 50 piasters – half the cost of the cheapest Austrian model.46 Women earned even less than men, and several items worn by women were locally produced specialties. Hence, there was even more room for growth in the women’s needle trades. Shawls were woven by many women and girls working in their homes in and around Asyut. Merchants imported cotton yarn and sold it to the women, who needed little training, as they generally had learned needlework at home. In all likelihood the same merchants bought the finished shawls from the women and sold them in Cairo, where they were popular among both indigenous women and tourists. Because the income of the Asyuti women supplemented their family income, they could be paid below subsistence wages. The dispersion of the workers in several villages as well as the town facilitated their super-exploitation.47 In the early twentieth century Asyuti merchants also organized production of tulle veils. The merchants distributed fabric and machine-made thread imported from France and Germany to women who worked at home in villages around Asyut. The raw materials required to produce one veil cost 80 piasters. A woman received 1–2 piasters for a finished veil, which probably entailed a day’s labour – far from a living wage.48 The 1907 census enumerated 83,378 people (about 75,000 males) working in the textile industry, including weavers of straw, linen and hemp as well as dyers and makers of rope and lace. There were 45,456 hand-loom weavers (42,355 males). The largest concentrations of weavers were in Mahalla al-Kubra (3,183), Cairo (2,869), Qalyub (2,405), Minuf (2,394), Shibin al-Kom (1,628), Sanuris (1,530), Damietta (1,431) and Akhmim (1,266). Sixty-six districts throughout Egypt had more than 100 weavers. Mahalla, Cairo and Damietta were centres of silk weaving. Akhmim, ‘the Manchester of Egypt’, was a centre of cotton weaving. Wool was woven in many rural towns. In Abu Qurqas, which was close to the herds of sheep, goats and camels concentrated in upper Egypt, at least 750 spinners supplied wool thread for the town’s 400 looms.49 Sidney Wells, director-general of the Department of 45 Ibid., p. 111. 46 Ibid., pp.113–114. 47 Ibid., pp. 111–112. 48 Ibid., pp. 112–113. 49 F. Moore Gordon, ‘Rapports sur les industries du tissage à Beni Adi, Beni Souef et Abou 182

Egyptian textile workers Agricultural, Industrial and Commercial Education, surveyed the textile industry in 1910 and concluded, ‘One can affirm that the textile industry, far from declining as is often claimed, is in a state of growth and, to the contrary, is first among the local industries.’50 Mahalla al-Kubra offers a good example validating Wells’s judgement and illustrating the flexibility of Egypt’s textile workforce. In the early twentieth century Mahalla had recovered from its mid-nineteenth-century decline and was the leading centre of silk weaving with 2,455 handlooms, far more than any other locale. But by 1917 silk weaving in Mahalla had collapsed. There were only 158 weavers, probably due to the unavailability of raw silk from Mount Lebanon during the First World War. Many silk weavers apparently shifted to rayon or artificial silk, as the 1917 census lists 2,528 weavers working in ‘yarns of non-animal origin, or unspecified yarns’. Despite the competition from European imports, the number of weavers working in cotton increased from 43 in 1907 to 646 in 1917. Consequently the number of workers in textiles and related industries (about 3,900) remained stable in Mahalla between 1907 and 1917. They comprised over 30 per cent of the economically active population of 10,458, a proportion consistent with the historic position of textiles in Egypt’s urban crafts.51 Feminized and ruralized textile production meant continuing downward pressure on wages. Wages in the provinces were lower than in Cairo and women’s wages were typically far lower than men’s. The daily wage of a male weaver around 1910 ranged between 4 and 10 piasters.52 In Qalyubiyya province wool weavers earned 6–8 piasters and flax weavers 3–4 piasters.53 Silk and cotton weavers in Mahalla earned 3–5 piasters a day.54 The annual income of a weaver working six days a week and earning 5 piasters a day in the provinces would amount to slightly more than three-quarters of the estimated minimum of 2,000 piasters a year required to sustain a rural family.55 Such low wages could be sustained because rural weaving was not a full-time occupation but largely a part-time activity concentrated around saints’ birthdays and other holidays. Competition from the provinces contributed to the miserable working conditions and wages in Cairo. Weavers worked in ‘quarters where most of the buildings are ramshackle, at the back of courtyards, in sheds without roofs … in long, narrow, and badly lit spaces’. During part of the year spinners and weavers worked an extra half shift until midnight or 1 a.m. to earn an extra 75 centimes or a franc.56 Kerkas’, L’Égypte Contemporaine 2 (5) (January 1911): 71–73. 50 Sidney H. Wells, ‘Note préliminaire sur l’industrie du tissage en Égypte’, L’Égypte Contemporaine 1 (4) (November 1910): 579. 51 El-Messiri, ‘Class and Community’, pp. 69–74. 52 Wells, ‘Note préliminaire sur l’industrie du tissage en Égypte’, p. 581. 53 Sidney H. Wells, ‘L’industrie du tissage en Égypte’, L’Égypte Contemporaine 2 (5) (January 1911): 57. 54 F. Moore Gordon, ‘Notes sur l’industrie textile à Mehalla-Kebir’, L’Égypte Contemporaine, 1 (2) (March 1910): 336. 55 Chalcraft, Striking Cabbies of Cairo and Other Stories, p. 119. 56 Germain Martin, Les Bazars du Caire et les petits métiers arabes (Cairo and Paris, 1910), 183

The Ashgate Companion to the History of Textile Workers The prevalence of hand-loom weaving should not lead to the conclusion that Egyptians were unwilling or unable to introduce machinery. Without the ability to enact protective tariffs, the only way that Egyptian textiles could compete with European imports was to rely on super-exploited labour. This was a rational, even if not a humane, business decision. Machinery could be used if it was manufactured locally, hence cheaply, rather than imported. An industrial school was established in Abu Tig in 1903. By the end of the First World War the weavers there had learned to construct wooden mechanical looms for silk weaving modelled on those imported from Manchester.57 More advanced looms were also used in Mahalla, Akhmim and Bani Suwayf.58

Mechanization, trade unions and the nationalist movement During the first decade of the twentieth century Egyptian workers began to organize themselves in trade unions. Almost from the start, the trade union movement was allied to the nationalist struggle against the British occupation of Egypt and European domination of its economy. However, textile workers were marginal to the early trade union movement. No more than 193 weavers ever belonged to the Manual Trades Workers Union established by the Nationalist Party in 1909 – an amalgam of a federation of guilds and a trade union which was the first institutional effort to link trade unionism to the nationalist movement. There were no unions of textile workers before the First World War. The 1919 nationalist uprising, in which the Wafd emerged as the representative of populist Egyptian nationalism, was accompanied by a wave of strikes and the formation of several dozen new trade unions. Among them was the union of Egypt’s first mechanized textile mill, the Filature Nationale d’Égypte. The Filature Nationale was descended from one of the two mechanized spinning and weaving mills established in 1899: Egyptian Cotton Mills Ltd in Cairo and Anglo-Egyptian Spinning and Weaving Co. in Alexandria. The investors assumed that they would not have to pay any local tax on their products and were surprised when Lord Cromer, the British viceroy, imposed an 8 per cent countervailing duty equivalent to the tariff paid on British imports. Deprived of tariff protection, the firms could not survive. Egyptian Cotton Mills went bankrupt in 1907. Anglo-Egyptian Spinning and Weaving persuaded the government to suspend the countervailing duty for five years. In 1912 it was reorganized by a group of German investors and renamed the Filature Nationale d’Égypte. By the First World War the Filature Nationale employed 1,200 workers operating 20,000 spindles and 400 power-looms. They produced cotton thread and cloth valued at E£50,000 a year. Annual imports of these same products from pp. 52, 83. 57 Chalcraft, Striking Cabbies of Cairo and Other Stories, p. 123. 58 Wells, ‘Note préliminaire sur l’industrie du tissage en Égypte’, p. 581. 184

Egyptian textile workers 1900 to 1913 ranged from E£2.7 million to E£40 million. The mill used 25–30,000 kantars of cotton (1 kantar = 157 kg); Egypt’s total annual production amounted to 6–7.5 million kantars. By the Second World War the workforce had grown to 8–10,000.59 In the early 1920s, while the cost of living remained elevated compared to preFirst World War levels, many employers sought to take back concessions they had granted during the nationalist and trade union upsurge of 1919–1923. The Filature Nationale union was in a weak position because its political patron, the Nationalist Party, was no longer an important political force; it was overwhelmed by the Wafd in the 1923 elections, the first in nominally independent Egypt. After a slowdown in November 1923 and a sit-in strike in February 1924, the Filature Nationale workers were forced to accept a 10 per cent wage cut. The sit-in at the Filature Nationale was part of a city-wide workers’ mobilization in Alexandria led by the first Communist Party of Egypt. The newly installed Wafd government crushed both the Alexandria workers’ movement and the Communist Party. Bank Misr was founded in 1920 during the nationalist uprising, proclaiming itself ‘an Egyptian bank for Egyptians only’. The bank sought to provide capital to establish Egyptian-owned, large-scale, industrial enterprises. Its flagship industrial enterprise was the Misr Spinning and Weaving Company, established in Mahalla al-Kubra in 1927. Misr Spinning and Weaving undertook the entire cotton spinning, weaving and dyeing process. Soon after beginning full-scale operations, it employed some 15,000 workers. By the end of the Second World War the mill employed over 25,000 workers and was the largest industrial enterprise in the Middle East. In 1938 Bank Misr established two more mechanized textile mills as joint enterprises with the British Bradford Dyers firm: Misr Fine Spinning and Weaving and Beida Dyers. The two mills at Kafr al-Dawwar, near Alexandria, employed some 11–12,000 workers.60 In 1937 a Syrian Muslim émigré family established the Sbahi Spinning and Weaving Company. It employed 5,500 workers in three mills in Alexandria and the northern Cairo suburb of Shubra al-Khayma. Workers in these and other mechanized textile mills formed the centre of an emergent Egyptian urban proletariat. In 1945, according to the Census of Industrial Production, 117,272 of the 316,144 employees in manufacturing enterprises were textile workers and 9,425 of the 22,220 manufacturing enterprises were spinning and weaving mills. Thirteen of the 45 enterprises employing over 500 workers were textile mills. Over 50,000 of the textile workers (about 42 per cent) were employed at the seven large mills of the Filature Nationale, Misr and Sbahi. Thousands more were employed at medium-sized mills in the Cairo suburbs of Shubra al-Khayma, Zaytun, Matariyya and Imbaba.

59 Owen, The Middle East and the World Economy, pp. 237–238; Joel Beinin and Zachary Lockman, Workers on the Nile: Nationalism, Communism, Islam, and the Egyptian Working Class, 1882–1954 (Princeton, 1987), p. 273. 60 Beinin and Lockman, Workers on the Nile, p. 273. 185

The Ashgate Companion to the History of Textile Workers Many workers in these mills were from peasant backgrounds. Fikri al-Khuli, one such peasant boy, went to work at Misr Spinning and Weaving in 1928 when he was 11 and remained until 1942. His autobiographical novel, al-Rihla (The Journey), the only first-hand account of life from the perspective of the mill floor by an Egyptian textile worker, vividly relates the harsh labour control measures and supervisors’ frequent abusive treatment of workers. Al-Khuli and his workmates understood their work experience in radically different terms than the Misr management and its political allies.61 Peasant workers saw themselves as working temporarily to save a sum of money before returning to their villages. Turnover was very high. In the 1930s the dominant basis of identity in and around the mill was the antagonism between the local residents of Mahalla (mahallawiyya) and the peasants recruited to work at the Misr mill (company men, or shirkawiyya).62 The shirkawiyya identified themselves by their villages of origin, as non-Mahalla residents and as Muslims. Even when they came to believe they had common interests and to act on them, they often regarded the Mahalla residents as their enemies as much as the Misr management. Most of the males fiercely opposed the entry of females to their workplace. Some even physically attacked them.63 These circumstances made it difficult to organize the workers. Nonetheless, in 1936 Prince ‘Abbas Halim, who fancied himself ‘Prince of the Workers’, visited Mahalla intending to form a union under his leadership. Borne on the shoulders of workers, he entered a marquee where he addressed an assembled crowd in a stammer. Many workers were distressed to find that they could not understand him. ‘What’s he saying? Is he speaking Polish?’ they asked each other. ‘Is he speaking to us? He’s speaking to the mechanics.’ One said, ‘He’s speaking a foreign language.’ Another said, ‘Of course he’s ignoring us, man. Why should he talk to us? He’ll speak to people who understand him.’64 ‘Abbas Halim had spent much of his childhood in Europe and never perfected his Arabic. On this formal occasion he may have thought it appropriate to speak in his flawed standard Arabic rather than colloquial Egyptian. The better educated skilled mechanics probably understood him far better than uneducated or minimally educated unskilled workers, who would have had some difficulty understanding even faultless standard Arabic. This first attempt to organize a union at Misr Spinning and Weaving failed, not only because the workers literally did not understand what ‘Abbas Halim proposed, but because they felt little connection to Cairo-centred political and trade union activity.

61 Fikri al-Khuli, al-Rihla (3 vols, Cairo: Dar al-Ghad, 1987–1892). For an extended analysis of this text, see Joel Beinin, Workers and Peasants in the Modern Middle East (Cambridge, 2001), ch. 4. 62 Khuli, al-Rihla, vol. 1, pp. 55ff. 63 Ibid., vol. 3, pp. 51–64. 64 Ibid., vol. 3, p. 31. 186

Egyptian textile workers On 18 July 1938, 1,500 weavers at Misr Spinning and Weaving, with no help from any outsiders, struck demanding an eight-hour day and a higher piece-rate.65 The company shut down for 45 days and reorganized the weaving mill into smaller workshops to diminish the danger of future collective action. About a hundred workers were arrested for their role in the strike and paraded through town as an example; 55 were convicted for participating in the strike. The judge who presided over the case chastised the workers in paternalistic language, expressing the disdain of most of Egypt’s political classes towards the emergent working class and incredulity that ignorant peasants were capable of knowing their own interests.66 The judge, like the Misr management and its political supporters, believed that the Mahalla mill was the emblem of Egyptian modernity, industrial progress and economic development. Ignorant peasants should therefore accept the roles the company assigned to them in this project. Deviations would be dealt with harshly, just as the Wafd had crushed the 1924 strike movement in Alexandria. However, the modernity of Misr Spinning and Weaving was derivative. Contrary to the hopes and expectations of many Egyptians, it was incapable of generating the industrial takeoff that mechanization of the textile industry produced in England. The machinery at Misr Spinning and Weaving was purchased second-hand from European firms who were moving on to technologically more advanced equipment. Most workers were poorly educated, if at all, so the quality of their product was low. Crude methods of labour discipline and a high rate of absenteeism meant that the number of workers per loom at Misr’s enterprises was far greater than in Japanese, European and American textile mills. Hence, Egyptian textiles were uncompetitive on the international market. The local market could only be secured by enacting a protective tariff after Egypt regained tariff autonomy in 1930. Maintaining a larger than necessary labour force and controlling it through harsh discipline was economically ‘rational’ in the short term because of the low cost of labour. But low wages also meant that the local market for Egypt’s premium long-staple cotton was limited. Misr therefore used this long-staple cotton to produce cheap cloth for the popular classes, squandering Egypt’s comparative advantage.

Textile workers’ unions and the nationalist movement During the 1930s ‘Abbas Halim and the Wafd were first allies and then rivals in claiming to represent Egyptian workers. In 1942, during one of the brief periods when the British and the King allowed the Wafd to rule, trade unions were legalized. Consequently, the Wafd retained popularity among some trade unionists. But by the late 1930s others had despaired of the political figures who sought to be their patrons and to reject the corporatist conceptions embraced by all the political 65 Ibid., vol. 3, pp. 165ff. 66 André Eman, L’industrie du coton en Égypte: Étude d’économie politique (Cairo, 1943), pp. 183–184. 187

The Ashgate Companion to the History of Textile Workers forces except the Marxists. These elements advocated a trade union movement ‘independent of political parties’. Although this demand began outside the textile industry, by the middle of the Second World War the textile workers of Shubra al-Khayma had become its strongest proponents. The mechanized textile industry mushroomed during the Second World War, when Egypt became a rear base and supply centre for the Allied war effort. After the war, employment dropped sharply. Mill owners squeezed their remaining workers to make a profit in a radically diminished market. As most textile workers were from peasant backgrounds, the notion that they could be hired and dismissed or made to work harder than the previous norm because of the ‘logic’ of the capitalist market violated their conceptions of a moral economy. ‘Arbitrary dismissals’ and low wages in relation to the increasing cost of living were recurrent issues for textile workers after the Second World War. During and after the war textile workers, especially those of Shubra al-Khayma and other northern suburbs of Cairo, drew the attention of the reinvigorated Egyptian communist movement. Consequently, until the collapse of the monarchy in 1952 and for some time beyond, these textile workers formed the core of a radical labour movement linking the economic demands of workers with the nationalist demand for ‘total evacuation’ of British troops from the Nile Valley. The government attempted to repress this movement by dissolving the General Union of Mechanized Textile Workers of Shubra al-Khayma and Cairo in April 1945. In December security forces imposed a virtual armed occupation on Shubra al-Khayma. This provoked a nearly general strike from 1–9 January 1946 during which union leaders Taha Sa‘d ‘Uthman and Mahmud al-‘Askari and another senior worker-organizer, Yusuf al-Mudarrik – all of them connected to the New Dawn Marxist group – were arrested. The strike was actively supported by the several Marxist factions and the Wafdist Vanguard, the youthful left wing of the principal nationalist party. In February a major clash between student demonstrators demanding abrogation of the 1936 Anglo-Egyptian Treaty and evacuation of British troops from Egypt and Sudan led to the formation of the National Committee of Workers and Students (NCWS). Its leadership included two textile workers from Shubra al-Khayma, Mahmud al-Dumrani and Sayyid Khudayr. The NCWS called a successful general strike on 21 February, when workers from Shubra al-Khayma joined tens of thousands of others demonstrating in Cairo. In May, impelled by the nationalist mobilization, trade union leaders attempted to overcome the differences between the Marxist factions and form a national trade union federation: the Congress of Egyptian Trade Unions (CETU). Simultaneously, workers in at least nineteen textile mills in Shubra al-Khayma struck in response to continuing plant closures and layoffs – a nearly total shutdown of the regional textile industry. The CETU called a national general strike for 25 June to support its call for complete evacuation of British forces from the Nile Valley and a long list of trade union issues, including solidarity with the Shubra al-Khayma strikers. The general strike failed. This signalled the recently appointed Prime Minister, Isma‘il Sidqi, that he was free to attack the leftist trade union-nationalist alliance. On 11 July trade union and nationalist leaders were arrested, the CETU was banned, 188

Egyptian textile workers the left and labour press was closed down, several Marxist associations were outlawed and a new anti-communist law was submitted to parliament. The date 11 July 1946 marked a decisive defeat for the popular movement of 1945–1946. It ended a conjuncture when both the government and the left wing of the nationalist and trade union movements took seriously the possibility of a revolution.67 The Marxists’ success in Shubra al-Khayma led them to clash with the Society of Muslim Brothers. In 1944 some of the Brothers’ adherents were elected to the executive committee of the Shubra al-Khayma textile workers’ union. In the spring and summer of 1946 the Muslim Brothers organized their own textile workers’ union to replace the one dissolved by the government. In December 1947 they established a small textile mill in Shubra al-Khayma employing less than 60 workers. It was intended to relieve the problem of unemployment and demonstrate the proper way to run an enterprise on what the Muslim Brothers considered Islamic principles. Although the Brothers never gained as great a following in Shubra al-Khayma as the left commanded in 1945–1946, their activity obstructed the re-establishment of the left until after the 1948 Arab–Israeli War.68 Consequently, the Delta and Alexandria were the main centres of workers’ collective action during a second upsurge of converging radical labour and nationalist mobilization, from September 1947 to April 1948. Muhammad Hamza and other members of the newly formed Democratic Movement for National Liberation (DMNL) – which united the former Iskra and Egyptian Movement for National Liberation Marxist groups – began to work in Mahalla al-Kubra in the summer of 1947. The announcement of new disciplinary rules sparked a violent and intense strike of the 26,000 Misr Spinning and Weaving workers from 2 September to 4 October 1947. But its underlying causes were resentment over years of humiliating treatment by the management and anxiety about layoffs. There was a company-sponsored union at Misr Spinning and Weaving. As the strike developed its principal demand came to be the formation of an independent, representative union.69 The other major labour struggle of autumn 1947 was a 17-day strike at the Filature Nationale from 26 September to 13 October prompted by wage reductions related to the institution of an eight-hour day.70 In response to these strikes, in January 1948 a government commission issued a report recommending standard conditions for the textile industry. The standards, although not obligatory, legitimized the concepts of a minimum wage, an eight-hour day and a six-day week. A rash of strikes at the Filature Nationale and Sbahi from February to April 1948 accompanied by solidarity strikes at other mills suggests that textile workers thought the government’s proposals were inadequate.71

67 Beinin and Lockman, Workers on the Nile, pp. 338–349. 68 Ibid., pp. 366–375. 69 Ibid., pp. 353–355. 70 Ibid., pp. 356, 359. 71 Ibid. 189

The Ashgate Companion to the History of Textile Workers The textile workers of Cairo’s northern suburbs once again became a centre for Marxist activity during a third upsurge of the radical labour and nationalist movement from mid-1951 until January 1952. The Greater Cairo Mechanized Textile Workers’ Union based in Zaytun and Matariyya was led by DMNL member Muhammad ‘Ali ‘Amir. Another DMNL member, Muhammad Shatta, led the Founding Committee for a Union of Mechanized Textile Workers in Shubra al-Khayma and its suburbs, which sought to re-establish a textile workers’ union in Shubra al-Khayma throughout 1951. The police repeatedly thwarted its efforts to convene an assembly to formally reconstitute the union. Nonetheless, a Congress of Egyptian Textile Workers managed to meet in Cairo on 5 January 1952 under the leadership of Muhammad ‘Ali ‘Amir. That gathering was intended to prepare for a founding congress of a General Federation of Egyptian Trade Unions which was to convene on 27 January 1952. The Cairo fire of 26 January ended the popular mobilization and provided a pretext for banning the trade union congress. The Wafd government was dismissed, martial law was declared and many trade unionists were jailed. The stage was set for the military coup that ended the monarchy on 23 July 1952.

The Free Officers and the trade union movement The Free Officers led by Gamal Abdel Nasser seized power with a vaguely articulated programme of nationalism, ‘anti-feudalism’ and political reform. The political sympathies of the officers ranged from the Muslim Brothers to the DMNL. The military regime was widely supported by workers and peasants, especially after it enacted a land reform in September 1952. Workers at the Misr mills in Kafr al-Dawwar believed that the new regime empowered them to renew their campaign for representative trade unions. The 3,000 workers at Beida Dyers won this demand after a brief strike on 9 August 1952. At Misr Fine Spinning and Weaving there was a company-controlled union. A union reform movement led by Muhammad Mutawalli al-Sha`rawi and Ahmad al-Yabani, who were close to the DMNL, had been active for some time. It did not initiate the sit-down strike of some 500 workers on 12 August. The next day troops arriving from Alexandria fired on large demonstrations of workers as they shouted their support for the new regime and demanded the removal of abusive company officials and the formation of a freely elected union. Two soldiers, one policeman and four workers were killed in an exchange of shots. The military authorities arrested 545 workers and charged 29 of them with various offences. A hastily convened military tribunal sentenced Mustafa Khamis and Muhammad Baqri to death, and they were executed on 7 September. The government’s principal consideration in imposing this harsh penalty was its suspicion that they were members of Marxist organizations, although Baqri almost surely was not. Ten other workers including Muhammad Shihab, a Muslim Brother, received jail sentences.72 72 Ibid., pp. 421–426. 190

Egyptian textile workers To balance its repression of the Kafr al-Dawwar strikers, the military regime enacted three labour laws in December 1952. Law 318 established a mandatory arbitration and conciliation procedure for grievances, which effectively banned strikes. Law 317 provided longer annual vacations, free transportation to factories in remote areas and free medical care in enterprises employing more then 500 workers. The most important provision of the law was increased severance compensation, which made it expensive to lay workers off during business downturns. Decree 165 of April 1953 forbade dismissing workers for union activity. However, this was offset by quadrupling the length of time a worker could be hired on probation (with lower wages and benefits) – from three periods of one month to two periods of six months. Many employers evaded the high cost of laying off workers by dismissing new hires before a year was up.73 By increasing severance compensation the Free Officers partially alleviated the single most important grievance of the labour movement since the end of the Second World War – one that had been particularly acute among textile workers due to the sharp ups and downs of the industry. This was among the reasons that most trade unionists supported Gamal Abdel Nasser during his confrontation with Muhammad Naguib and others who demanded the restoration of parliamentary democracy in the crisis of March 1954. However, the textile unions, in part due to the presence of Marxists, generally did not participate in the strikes of 27–29 March 1954 that confirmed Abdel Nasser in power. Many trade union leaders believed that it was in the interest of their constituency to become clients of the new regime, which seemed much more amenable to their interests than the monarchy. In contrast, the textile unions in which Marxists were prominent continued to uphold the banner of trade union independence. However, the political limits of that option became obvious when security forces violently suppressed a strike and sit in at the Shurbagi textile mill in Imbaba in August– September 1953.74 The regime was adamant that order and discipline would be maintained and that social and political initiatives could not come ‘from below’. Consequently, it fought relentlessly against communists and other leftists in the labour movement, especially in the textile unions which were their stronghold. However, this had to be done with some delicacy because in the 1950s Egypt’s 150,000 mechanized textile workers comprised about 43 per cent of its industrial workforce.75 Such a large and strategically located constituency could not be alienated by a regime claiming to represent the popular strata. From the mid-1950s to 1965 the regime adopted progressively more radical anti-capitalist measures while institutionalizing its corporatist conception of state and society. Trade unions effectively became part of the state apparatus. The most concentrated resistance to the state’s absorption of the labour movement was among the textile trade union leaders. Muhammad Mutawalli al-Sha‘rawi from 73 Marsha Pripstein Posusney, Labor and the State in Egypt: Workers, Unions, and Economic Restructuring, 1952–1996 (New York, 1997), pp. 47–48. 74 Beinin and Lockman, Workers on the Nile, p. 434. 75 Robert Tignor, Egyptian Textiles and British Capital, 1930–1956 (Cairo, 1989), p. 57. 191

The Ashgate Companion to the History of Textile Workers Kafr al-Dawwar and Sayyid Fa’id and Ahmad Fahim from Shubra al-Khayma led efforts to establish a federation of textile unions in 1953–1954. Fahim had formed a breakaway textile foremen’s union in 1942, when the Marxists were ascendant among the Shubra al-Khayma textile workers. Although he was not a constant ally, Fahim was a pragmatist and willing to cooperate with Marxists like al-Sha‘rawi and Fa’id. Therefore, the regime unsuccessfully tried to prevent Fahim from becoming president of the Textile Workers’ Federation. Under his leadership the Federation remained one of the few institutions where workers could independently discuss issues and take initiatives. Its semi-annual conference in 1956 was the first forum to demand doubling the minimum wage.76 Fearful that what happened in the textile federation might occur on a larger scale, the regime repeatedly blocked efforts to establish a national trade union confederation. It was finally formed in January 1957 on the basis of a corporatist bargain – the government would permit the formation of a trade union confederation but exercise veto power over its leadership and supervise its activities. Ahmad Fahim became vice-president of the Egyptian Trade Union Federation (ETUF).77 Nationalization of Egypt’s large-scale financial, commercial and industrial enterprises in 1960–1962 brought most of the textile industry under direct government control. Only small and medium-sized firms employing less than 200 workers remained in private hands. For example, by 1973 the 3,600 privately owned mechanical looms in Mahalla al-Kubra were dispersed in 600 small shops, most with fewer than 30 looms each.78 Textile workers, like all those employed in the state sector, benefited from decreased hours of work, higher wages, social insurance, consumer cooperatives and medical clinics.79 In 1962 Ahmad Fahim became president of the ETUF, creating the impression that the labour movement was now led by a textile worker connected to the left tradition in the labour movement. But Fahim’s leadership did not enhance the capacity of textile workers or others for autonomous action.

Conclusion: towards a neo-liberal order From the mid-1950s to the present, despite the nationalization of most of the textile industry in the early 1960s, the encouragement of private enterprise in the mid-1970s and the partial denationalization of a relatively small number of enterprises in the 1990s, the state has tightly controlled trade unions. Elections to union executive committees have often been fixed. In 1987 the High Constitutional 76 Posusney, Labor and the State in Egypt, pp. 58, 69. 77 Ibid. Elsewhere I have rendered the name as a direct translation from the official Arabic name: the General Federation of Egyptian Trade Unions. 78 El-Messiri, ‘Class and Community’, p. 128. 79 Mahmoud Abdel-Fadil, The Political Economy of Nasserism: A Study in Employment and Income Distribution Policies in Urban Egypt, 1952–1972 (Cambridge, 1980). 192

Egyptian textile workers Court ruled that the right to strike was constitutionally protected, but striking workers continued to risk severe repression. Legislation adopted in 2003 formally legalized strikes but under severely restricted circumstances. Labour news has rarely been reported on the radio and television and in the government press, though in the mid-1980s the legal leftist weekly al-Ahali consistently covered labour issues. Egypt’s defeat in the 1967 Arab–Israeli War undermined the regime’s legitimacy and exposed it to significant political and social criticism for the first time since 1954. Since then there have been sporadic outbursts of collective action and strikes in response to declining real wages in 1971–1972 and 1975–1976, reductions in consumer subsidies and the imposition of Washington consensus neo-liberal policies in 1977, price increases and increases in workers’ contributions to health insurance and pension plans in 1984–1989, and the privatization of the public sector in the 1990s and early 2000s. Textile workers in the large public sector mills – the Misr mills in Mahalla al-Kubra, Kafr al-Dawwar and Helwan, and the ESCO mills in Shubra al-Khayma and Qalyub – were prominent in all these waves of protest and strike activity along with other public sector workers. Private sector workers participated much less actively in collective actions of any sort. In the mid-1980s several ‘alternative’ newspapers and organizations emerged to give workers a voice outside the framework of the state-dominated trade union federations, which had long ceased to represent them. Veteran textile union activists Taha Sa‘d ‘Uthman and Muhammad Mutawalli al-Sha‘rawi were among the editors of Sawt al-‘Amil (Workers’ Voice), which began publication in 1985. Sawt al-‘Amil sharply criticized the historic absorption and repression of the labour movement by the Nasserist state. It tried to connect the workers’ movement of the mid-1980s with that of the 1940s and 1950s by publishing biographies of veteran trade union activists and other historical material. One of its major projects was a series of articles presenting a legal case for the innocence of the Kafr al-Dawwar textile workers who struck on 12–13 August 1952. Textile workers at Mahalla al-Kubra formed the core of an area-wide organization, al-Fajr (Dawn), which had several hundred members and issued a newspaper with the same name in the mid-1980s. Another regional newspaper of the period was based among the textile workers of Shubra al-Khayma (‘Ummal Shubra al-Khayma). The fiercest confrontations between workers and the state during the 1984–1989 strike wave were two sit-in strikes at the Iron and Steel Co. in Helwan in July and August 1989.80 Its leaders were among the founders of the Centre for Trade Union and Workers’ Services, which subsequently established a branch in Shubra al-Khayma. After the Iron and Steel strikes, the regime of Husni Mubarak became less tolerant of labour dissidence, an aspect of its generally more repressive character in the 1990s. This revival of labour militancy occurred in the context of a structural crisis of the textile industry concentrated in the public sector. The proportion of textile workers in the industrial workforce began declining in 1960; the absolute number of 80 Omar El Shafei, Workers, Trade Unions, and the State in Egypt: 1984–1989, Cairo Papers in Social Science 18 (Cairo, 1995), pp. 22–29, 33–35. 193

The Ashgate Companion to the History of Textile Workers textile workers began declining in 1976.81 The index of real wages of textile workers declined by more than one-third: from 100 in 1986 to 61 in 1994, an even more precipitous decline than that of the industrial workforce as a whole.82 Due to lack of capital investment, productivity was actually lower in 1999 than in 1985; and it is much lower than in neighbouring countries like Tunisia and Turkey. A quarter of the machinery is outdated and needs to be renovated or replaced.83 Average wages of Egyptian textile workers are among the lowest in the world: 85 per cent of wages in Pakistan and 60 per cent of wages in India. A weaver in a well-run private sector enterprise makes about E£1,000 a month (including base pay of about E£500–600 plus incentives); a spinner makes about E£800 a month (including base pay of about E£400 plus incentives). It is, of course, not accidental that the lower paid spinners are mostly women. Private sector textile workers work 12-hour shifts and rarely receive health insurance or other social benefits (whose total cash value is about E£400 a month) despite the fact that they are legally entitled to receive these benefits. Wages are about half this amount in public sector enterprises, where daily shifts are 8 hours and workers receive time-and-a-quarter for overtime.84 Nonetheless, the industry is globally uncompetitive. Exports began to decline in 1990, exacerbated by a global textile recession in 1991. The value of total production reached the lowest point since government statistics began to be collected in 1996/97.85 The private sector has been encouraged by the state since Anwar al-Sadat’s declaration of an ‘open door’ economic policy in 1974. There was relatively little private investment in any industrial sector in the 1970s. Private sector investment in textiles declined precipitously from 1983 to 1987. It grew following the 1991 Gulf War, but stalled again in mid-2000.86 Egypt’s support for the United States in the 1991 Gulf War resulted in cancellation of nearly half of its foreign debt. This created the conditions allowing Egypt to sign an Economic Reform and Structural Adjustment Program agreement with the IMF and the World Bank. Law 203 of June 1991 was the first tangible step towards privatizing the public sector, which the international financial institutions had been urging for over a decade. After resisting privatization since 1974, the leadership of the Egyptian Trade Union Federation endorsed this legislation.

81 Tamer Abdel-Kader, ‘State, Capital and Workers’ Protests in Egypt’ (M.A. thesis, American University in Cairo, 1998), p. 79. 82 Ibid., p. 84. 83 American Chamber of Commerce in Egypt, The Textile and Clothing Industry in Egypt (Cairo: The Chamber, August 2004), p. 70. 84 Interview with the owner of a private sector textile enterprise who wishes to remain anonymous, Cairo, 19 March 2005. 85 Abdel-Kader, ‘State, Capital and Workers’ Protests’, p. 81; Egypt, Central Agency for Public Mobilization and Statistics, On-Line Census of Industrial Production, 2000/01 ‘`Adad al-mansha’at wa-`adad al-mushtaghalin hasba fi’at al-sin wa’l-naw`’. 86 Abdel-Kader, ‘State, Capital and Workers’ Protests’, p. 81. 194

Egyptian textile workers Several state-owned textile firms were sold to Egyptian and foreign investors during the mid-1990s spurt of private investment. They generated 12 per cent of all proceeds from the sale of public sector enterprises. From 1992 to 2000 the market share of the private sector in cotton spinning grew from 8 per cent to 58 per cent.87 In response to the crisis in spinning and weaving, production of ready-made garments became a more salient factor in the textile industry. This replicated the practice established in the late nineteenth and early twentieth centuries, when competition from imported European cloth led to a decline in Egyptian spinning and weaving, but a sharp rise in garment production. In 2000/01 there were 209,272 spinning, weaving, and dyeing workers and 163,241 garment workers (compared to 457,191 spinning, weaving, and dyeing workers in 1976). Private investors are well represented in garment production. The value of private sector garment production increased nearly 30 per cent from 1996/97 to 2000/01, while the value of public sector production fell by approximately the same amount. In other words, neo-liberal privatization in the garment industry did not result in a substantial increase in the value of total production, nor did it provide enough employment opportunities to replace the lost jobs in spinning and weaving. Moreover, the garment workforce is highly feminized, which has resulted in lower wages, just as was the case 100 years ago. Of all public sector garment workers, 51 per cent are women; 40 per cent of all shirt-makers are women, all in the private sector; and 52 per cent of all other private sector garment makers are women.88 Precise information about working conditions in privately owned textile and garment enterprises is difficult to come by. While they are legally required to provide the same social benefits and health insurance as public sector firms, most do not. They evade the law by bribing inspectors and producing for the local market, which requires lower standards of documentation than the export market. Firms that pay social insurance benefits must pay a severance package equal to two months wages for every year worked if they dismiss a worker. Firms that successfully avoid paying social insurance benefits are able to fire workers with no restrictions, a strong incentive for illegal behaviour.89 Before enactment of the 2003 Unified Labour Law it was widely reported that workers signed undated letters of resignation when hired so that they could be dismissed without receiving severance pay if they ‘made trouble’. Firms often hired workers on three-month temporary contracts and fired them before they became permanent. The 2003 legislation legalized indefinitely renewable temporary contracts.90 This means that many workers remain in limbo for years living in 87 Jolynn Khamky, ‘Liberalization to Divestment: Egypt, 1960–2000’ (senior honors thesis, Department of History, Stanford University, 2003), pp. 50, 65. 88 Central Agency for Public Mobilization and Statistics, On-Line Census of Industrial Production, 2000/01 ‘‘Adad al-mansha’at wa-‘adad al-mushtaghalin hasba fi’at al-sinn wa’l-naw’’. 89 Interview, Cairo, 19 March 2005. 90 Marsha Pripstein Posusney, ‘Globalization and Labor Protection in Oil-poor Arab 195

The Ashgate Companion to the History of Textile Workers fear of exercising their rights and with no job security. Several of the new private enterprises are located in the new cities surrounding Cairo. Workers are therefore cut off from the social networks required to support collective action; few private sector firms are unionized. By 1999, 137 of the 314 public sector firms declared eligible in 1991 had been privatized. Although the 1991 privatization legislation forbade mass layoffs after privatization of a firm, managers of public sector firms under consideration for privatization often attempted to make them more attractive by reducing the workforce before the sale. Anxieties about unemployment and other possible consequences of privatization prompted a renewal of strikes and collective action in the mid-1990s, with major strikes at Misr Fine Spinning and Weaving in Kafr al-Dawwar in November 1994 and Misr Helwan Spinning and Weaving in August 1998. At the Misr Helwan mill, the entire workforce of 8,700 was given a three-week vacation. Then management announced that only 2,800 workers should return to work. Rumours spread that the enterprise would be leased to a private investor.91 Some 400 workers of ESCO’s Qalyub Spinning Mill sat in at their enterprise for over four months from February to June 2005 protesting the sale of their firm to a private investor because neither the government nor the new owner would commit to maintaining the level of their wages, benefits and pensions.92 Despite the moves towards privatization, the textile industry remains the core of the public sector. The single largest state-owned enterprise is the Misr Spinning and Weaving Company in Mahalla. It employed over 20,000 workers in 2000 and generated revenues of E£700.2 million. By contrast, one of the most successful new private firms, Alexandria Spinning and Weaving, employed 3,200 workers and earned revenues of E£160.68 million. The recently privatized Arab Polvara Spinning and Weaving earned E£213.52 million.93 The pattern of collective action since 1971 indicates a significant disparity between the work experience, trade union and political consciousness, and militancy of public and private sector workers. Even among the generally more militant public sector textile workers, gender, locality, the differential effects of foreign competition, the particular commodity manufactured, the fabric worked and the political character of the local trade union leadership all produced important differences. These disparities prompt this concluding reflection. Surveying the history of Egyptian textile workers over the course of more than 300 years, as I have done, may encourage the illusion that the subject is unified and coherent. This is certainly Countries: Racing to the Bottom?’, paper presented at the conference on ‘The Jordanian Economy in a Changing Environment’ (Amman: University of Jordan, 22–23 July 2003). 91 Agnieszka Paczynska, ‘Globalization and Pressure to Conform: Contesting Labor Law Reform in Egypt’, paper presented at the DC Area Workshop on Contentious Politics (College Park: University of Maryland, 23 October 2002). 92 Joel Beinin, ‘Popular Social Movements and the Future of Egyptian Politics’, Middle East Report Online, 10 March 2005, http://www.merip.org/mero/mero031005.html 93 Khamky, ‘Liberalization to Divestment’, p. 67. 196

Egyptian textile workers not the case. Gender, locality, the differential effects of foreign competition, the particular commodity manufactured, the fabric worked and the absence or presence of trade unions and their political character all produced differences that make it problematic to speak of a set of common experiences and understandings that define ‘the’ Egyptian textile worker. It is possible to construct a coherent narrative of change over time, as this chapter has done, and it is useful to note that neither competition from European imports nor the failed industrial project of Mehmed ‘Ali created a radical discontinuity in textile craft production. But it is also important to recognize the ruptures and reformations of textile workers as a component of the Egyptian working class represented by the nationalist uprising of 1919 and its aftermath, the movement for independent trade unions from the late 1930s to 1952, the nationalizations and authoritarian populism of the Nasserist era and the still uncertain efforts to re-privatize the economy since 1991.

197

This page has been left blank intentionally

8 The German wool and cotton industry from the sixteenth to the twentieth century Dietrich Ebeling, Marcel Boldorf, Stefan Gorißen, Michael Mende, Anke Sczesny and Michaela Schmölz-Häberlein

Before the coming of factory industry, cotton and wool processing had a distinctly regional nature and it was during the expansion and specialization of the international market that the manufacturing landscape formed. Locations in the north-west of Germany profited notably from their proximity to Atlantic trading centres and their development came under the influence of the progressive industry in the Netherlands, north-western France and England. The textile industry did not conform to governmental units before or after the territorial realignment that was a result of the Napoleonic expansion. Nevertheless, governmental interventions influenced its location. The rural cottage industry (the phase of proto-industrialization) was partly the result of the restrictions of the old trade cities and their trade guilds. Larger territorial states pursued mercantile politics by granting privileges to single businesses, by the formation of government manufactories, by commissioning and by prohibitive tariffs in the eighteenth century. The integration of the trade region on the left bank of the Rhine into Napoleonic France and the use of the continental system led to short-term relocations, but did not fundamentally change the image of the Gewerbelandschaften.  Hans Pohl (ed.), Gewerbe- und Industrielandschaften vom Spätmittelalter bis ins 20. Jahrhundert (Stuttgart, 1986), pp. 112–202. The annotations are limited to the German literature. National and international research debates are not considered.  Overview in Peter Kriedte, Spätfeudalismus und Handelskapital; Grundlinien der europäischen Wirtschaftsgeschichte vom 16. bis zum Ausgang des 18. Jahrhunderts (Göttingen, 1980).  For Prussia, see Carl Hinrichs, Die Wollindustrie in Preußen unter Friedrich Wilhelm I. (Berlin, 1933).

Map 8.1 German textile areas

The German wool and cotton industry This was also the case for the process of industrialization in the nineteenth century that took place due to free market policies and thus under international competition. Production was concentrated in the urban centres, which had already developed in the eighteenth century, and which went on to become modern industrial cities. However, until the second half of the nineteenth century rural areas retained their prominence in hand-weaving. Cotton-producing centres in north-western Germany were located in parts of the Rhineland (Wuppertal, Gladbach/Rheydt). In the nineteenth century, under Dutch influence (Twente), cotton-producing locations in north-western Westphalia (Gronau, Bocholt, Rheine) developed. The Rhenish-Westphalian cotton industry of the late eighteenth and nineteenth century was partly affiliated to the older proto-industrial linen industry. In the east, the Saxon regions were dominant (Upper Lusatia, south-western Saxony and Vogtland). In southern Germany, especially in the city of Augsburg and its surrounding region in Franconia, the decline of the former fustian industry was balanced by the production of pure cotton fabrics and by the introduction of calico printing at the end of the seventeenth century. Here, as well as in other smaller locations of the former cotton trade, mechanical cotton-spinning mills were introduced (Augsburg, Kaufbeuren, Kempten, Hof, Bayreuth and Kulmbach) in the nineteenth century. During the First World War, textile production collapsed due to the lack of raw cotton and the regulations imposed on business, but then recovered and overcame the hyperinflation of 1923 and the worldwide economic crisis (1929/1933) relatively well. After the Second World War, the textile industry in Augsburg recovered initially, but went on to be concentrated within a few affiliated business groups. With increasing international competition from the 1960s onwards, a shrinkage process set in, reducing the Augsburg textile industry to insignificance.

 For the nineteenth-century textile industry, see Horst Blumberg, Die deutsche Textilindustrie in der industriellen Revolution (Berlin, 1965).  Gerhard Adelmann, Die Baumwollgewerbe Nordwestdeutschlands und der westlichen Nachbarländer beim Übergang von der vorindustriellen zur frühindustriellen Zeit 1750–1815; Verflechtung und regionale Differenzierung (Stuttgart, 2001).  Albin König, Die Sächsische Baumwollindustrie am Ende des vorigen Jahrhunderts und während der Kontinentalsperre (Leipzig, 1899); Karin Zachmann, ‘Die Kraft traditioneller Strukturen. Sächsische Textilregionen im Industrialisierungsprozeß’, in U. John and J. Matzerath (eds), Landesgeschichte als Herausforderung und Programm (Stuttgart, 1997), pp. 509–535.  Wolfgang Zorn, Handels- und Industriegeschichte Bayerisch-Schwabens 1648–1870 (Augsburg, 1961); Claus-Peter Clasen, Textilherstellung in Augsburg in der Frühen Neuzeit, vol. I: Weberei (Augsburg, 1995); Rolf Kießling, ‘Entwicklungstendenzen im ostschwäbischen Textilrevier während der Frühen Neuzeit’, in Joachim Jahn and Wolfgang Hartung (eds), Gewerbe und Handel vor der Industrialisierung; regionale und überregionale Verflechtungen im 17. und 18. Jahrhundert (Sigmaringendorf, 1991), pp. 27–48. 201

The Ashgate Companion to the History of Textile Workers Compared with these textile centres, the north-west does not appear as prominent. The centres of production seem to have been quite isolated and scattered. Their development was not autochthonous, but depended on external stimulation: from the thirteenth to sixteenth centuries mainly from Flanders, in the late seventeenth century from Huguenot hosiers immigrating from southern France to Hamelin, in the eighteenth century from carded and worsted makers and also master dyers from Thuringia and Saxony, and in the early nineteenth century from Dutch linen merchants establishing a cotton industry in the, then Hanoverian, county of Bentheim. As a result of the mechanization of cotton spinning in the first half of the nineteenth century, the availability of economical manual labour became less crucial. The increasing application of steam power left the spinning factories in the second half of the nineteenth century without the need for running water. In 1898, spinning factories were operating in 187 locations in Germany. However, a fundamental change in the industrial scenery did not occur. Of the 16 locations in which spinning mills with more than 100,000 spindles were installed, four were located in Saxony, three in Upper Franconia, three in Alsace, three in north-western Westphalia, two on the Lower Rhine and one in Augsburg. Though there was never a considerable cotton industry in the sense of both spinning and weaving in Bremen or its close neighbours, until the late 1960s this town was second only to Liverpool as the largest trading centre of cotton (particularly US cotton) in Europe. In the middle of the 1960s the South of the Federal Republic was dominant. Bavaria and Baden-Württemberg together had a share of 57 per cent in spinning and 58 per cent in weaving. Alongside this, Westphalia’s production was insignificant with shares of 25 per cent and 28 per cent respectively.

 Irma Butke, Zur Entwicklung der Textilindustrie in der Grafschaft Bentheim (Nordhorn, 1939); Ottfried Dascher, Das Textilgewerbe in Hessen-Kassel vom 16. bis 19. Jahrhundert (Marburg, 1968); Clemens Wischermann, ‘Vom Heimgewerbe zur Fabrik; Industrialisierung und Aufstieg der Nordhorner Textilindustrie im 19. und 20. Jahrhundert’, in Clemens von Looz-Corswaren (ed.), Nordhorn. Beiträge zur 600jährigen Stadtgeschichte (Nordhorn, 1979), pp. 190–228.  Horst J. Salzmann, ‘Die Wachstumsraten der Baumwollindustrie in der Bundesrepublik Deutschland seit 1950’ (Ph.D., Freiburg/Switzerland, 1971), p. 48. 202

The German wool and cotton industry Table 8.1 Regional division of spindles in main German cotton-producing states Prussia

Saxony

Bavaria

Württemberg

Baden

AlsaceLorraine

Total

1815

15

79

1

2

3

–

100

1825

15

76

1

4

4

–

100

1835

17

66

5

6

7

–

100

1845

20

55

7

3

16

–

100

1855

25

52

7

3

13

–

100

1865

22

31

23

10

16

–

100

1875

17

15

20

6

9

33

100

1885

18

20

19

7

8

28

100

1895

23

22

20

8

6

21

100

1905

30

22

18

8

5

17

100

1913

33

20

19

7

5

16

100

Source: H. Kiesewetter, Industrielle Revolution in Deutschland: 1815–1914 (Frankfurt am Main, 1989), p. 180.

Table 8.2 Shares of German states in cotton weaving and spinning, 1846/1861 Looms 1846

Spinning mills

1861

1846

Spindles

1861

1846

1861

Prussia

46

51

84

63

72

59

Saxony

28

24

7

19

13

27

Thuringia

8

14

2

5

3

4

Bavaria

6

4

3

2

4

2

Württemberg

5

3

2

3

6

4

Hanover

2

2

–

2

–

1

Others

4

3

2

5

3

3

100

100

100

100

100

100

Total

Source: Kiesewetter, Industrielle Revolution, p. 180. 203

The Ashgate Companion to the History of Textile Workers Wool processing was ubiquitous. In particular the production of woollen stockings can be found in many places. The region of Aachen (Aachen, Burtscheid, Monschau and smaller places) was a centre of fine woollen cloth production for the world market in the seventeenth and eighteenth centuries.10 During the proto-industrial phase, this region divided its labour with the production regions around the cloth centres of Verviers and Eupen in present-day Belgium.11 A less important centre for the production of fine cloth was Lennep on the right-hand bank of the Rhine.12 The Berliner Lagerhaus, founded in 1713, was a typical manufacture foundation of the mercantile era in Prussia, and never achieved the same importance as the Rhenish cloth industry.13 The worsted makers were concentrated in south-western Germany and Thuringia. Around 1690, there were approximately 900 and around 1770 there were approximately 1,300 clothiers in Württemberg. In 1755, in the area of the Calwer Moderation, the most important centre of wool processing in Württemberg, approximately 900 masters and 251 assistants were employed. The peak was reached in 1771, with approximately 1,000 weavers. This ‘Moderation’ entailed that between 1650 and 1797 worsted spinners and weavers in the region produced exclusively for the cloth-trade company in the small town of Calw. The first early industrial cloth factory in Calw was founded in 1806.14 Along with the factory industry, noticeable regional emphasis developed in the nineteenth century. Following the Prussian provinces of the Rhineland and Brandenburg, the realm of Saxony became the third most important location; in worsted yarn spinning and weaving it was the leader. The main centres of cloth weaving were the Saxon Vogtland and Saxon Lusatia. Worsted yarn spinning was also concentrated in the Saxon Vogtland as well as in Glauchau and Meerane. Further centres of cloth production were located in Württemberg and Bavaria, in Kurhessen, and in the Grand Duchies of Hesse and Hanover.15 One important wool-processing region in the north of Germany in the eighteenth century was Upper Eichsfeld, then politically part of the archbishopric of Mainz. This territory of just 1,500 km2 and little more than 100,000 inhabitants when it was ceded to Prussia in 1802, was a hub of German wool manufacture during the whole 10 Dietrich Ebeling and Martin Schmidt, ‘Zünftige Handwerkswirtschaft und protoindustrieller Arbeitsmarkt; die Aachener Tuchregion (1750 bis 1815)’, in Dietrich Ebeling and Wolfgang Mager (eds), Protoindustrie in der Region; europäische Gewerbelandschaften vom 16. bis zum 19. Jahrhundert (Bielefeld, 1997), pp. 321–346. 11 Pierre Lebrun, L’industrie de la laine a Verviers pendant le XVIIIe et le début du XIXe siècle (Liège, 1948). 12 Richard Robert Isenburg, ‘Untersuchungen über die Entwicklung der bergischen Wollindustrie’ (Ph.D., Heidelberg, 1906). 13 Hinrichs, Wollindustrie. 14 Walter Troeltsch, Die Calwer Zeughandelskompagnie und ihre Arbeiter; Studien zur Gewerbe – und Sozialgeschichte Altwürttembergs (Jena, 1897); Reiner Flik, Die Textilindustrie in Calw und Heidenheim 1750–1870; eine regional vergleichende Untersuchung zur Geschichte der Frühindustrialisierung und der Industriepolitik in Württemberg (Stuttgart, 1990). 15 Blumberg, Textilindustrie, p. 71. 204

The German wool and cotton industry of the eighteenth century. It is estimated that about 20 per cent of the population was directly involved in the woollen and worsted industries. Combing, spinning and weaving of light wools and worsteds were concentrated in several industrial villages, of which Gross Bartloff with about 1,000 inhabitants had become the most important. Though Eichsfeld exported finished fabrics, considerable parts of its production crossed the border in a semi-finished state. The destinations of these products were both nearby regions and England or France. For example, woollen and worsted yarn went to England or Picardy, however far more went to the early ‘industrial towns’ of the various territories surrounding Eichsfeld: to the Imperial Free City of Muhlhausen, to Langensalza in Saxony, to Eisenach, to Kassel and other places in Hesse, or to Göttingen and Osterode in Hanover. Mühlhausen was also the main purchaser of unfinished fabrics to be dyed, finished and then exported by Mühlhausen firms, mainly to the international fairs of Frankfurt am Main or Leipzig. In the course of the nineteenth century the Upper Eichsfeld woollen industry lost its importance, but continued to be a considerable source of workers, for combing in particular.16 In the nineteenth century, the Rhineland, Brandenburg and particularly Saxony were able to assert and expand their position as leaders in carded yarn production. In the decades before the 1870s, this production shifted in favour of Thuringia, Brandenburg and the Rhineland. Saxony and the south-western German states suffered losses. The annexation of Alsace-Lorraine again brought an extensive increase. Table 8.3 Looms weaving wool in the German tariff union and Hanover (%) 1846

1875

Brandenburg

12.5

20.4

Rhine provinces (including Arnsberg)

23.9

28.4

Thuringia

9.4

20.0

Saxony

24.3

12.3

Silesia

7.1

6.5

Southern Germany

11.7

5.6

Others

11.1

6.8

Source: Blumberg, Textilindustrie, p. 292.

16 Helmut Godehard, ‘Der Anteil der eichsfeldischen Wollkämmer an der Entwicklung der Machinenkammgarnspinnerei’, Eichsfelder Heimathefte 10 (1970): 210–228. 205

The Ashgate Companion to the History of Textile Workers The winners of this shift were Brandenburg, the Rhineland and Thuringia. Losses were recorded in Saxony and Southern Germany. The gain was a result of the expansion of worsted yarn processing in particular. The Rhineland, the region on the right bank of the Rhine, also benefited from this development. In contrast, the region around Aachen, which concentrated on carded yarn processing, became relatively less important. The decline of private spinning mills and the purchase of worsted yarn from Saxony and Alsace, for example, contributed to the fact that the share of the wool processing industry decreased when measured by the number of workers compared to other branches of industry. Nevertheless, worsted yarn processing remained concentrated in the ancient centres of Saxony and in the Rhineland, but it also expanded in Brandenburg (Lusatia). In 1875, these three areas claimed 72 per cent of the total output volume. Table 8.4

Percentage distribution of fine spindles in wool spinning Carded yarn

Worsted yarn

1846

1875

1846

1875

23.2

25.9

3.6

6.6

8.4

4.3

3.8

8.6

Rhine provinces

23.5

23.5

4.5

16.3

Saxony

13.4

31.0

39.4

31.2

5.1

4.1

35.1

14.0

10.3

2.4

10.1

17.3

Brandenburg Silesia

Thuringia Southern Germany

Source: Blumberg, Textilindustrie, p. 298.

Around the middle of the nineteenth century, the wool industry in Württemberg experienced a new boom, and due to the application of water power and modern machinery comparatively few workers were needed. Around 1850, the turnover of wool weaving and worsted yarn spinning in Württemberg added up to 7.8 million florins and by 1861, 74 wool-spinning mills employed 4,351 workers. In northern Germany Bremen became an important trading centre not only for cotton (Bremer Baumwollbörse), but also for wool from Australia and La Plata between 1860 and 1870. In 1884 Bremen wool merchants established a woolcombing mill (Bremer Wollkämmerei or BWK) as well. Another company was established in the same year at the nearby Oldenburgian town of Delmenhorst, the North German Wool-Combing and Worsted-Spinning Mill (Norddeutsche Wollkämmerei und Kammgarnspinnerei, NWK or shortened to Nordwolle). At the time of its breakdown in 1931 this was one of the leaders on the global market and 206

The German wool and cotton industry comprised every activity from sheep breeding in Argentina to worsted spinning and weaving mainly in Thuringia. Together with washing and combing mills in Doehren, south of Hanover, and Leipzig, both these Bremen-based enterprises ruled the entire industry until the early 1970s. Though the industry itself has almost vanished, in 2000 the BWK remained Europe’s leading supplier of slivers, meeting the demands of a market which stretched from northern Italy to Iran.17

Production The use of cotton for the production of garment cloth and bed and table linen since late medieval times, as well as its combination with flax yarn for the fabrication of siamoise, bombasines and fustians was complemented by the production of calicos in the eighteenth century. Until the end of the 1860s, only low-budget articles were mass-produced in the cotton industry while finer yarn was purchased from England. Only later did an improvement in the quality of the yarns and cloths occur, even though strong, coarse fabric was predominant during the entire nineteenth century.18 Cotton weaving on the left bank of the Lower Rhine encountered keen English competition in the first years after the collapse of the Napoleonic economic system, but only in the production of siamoise and bed linen that was adopted from the Wupper valley. However, this sector was lost to Saxony and Silesia in the 1830s. A number of businesses therefore took up the production of printed cotton skirt and trouser fabrics; however, they suffered on their part from fierce Swiss and Belgian competition from the 1840s. During the crisis resulting from the American Civil War (1861–1865), the production of trouser fabrics was abandoned entirely and replaced by semi-woollen trouser fabrics with a cotton warp and a woollen weft. After the war, the production of semi-woollen fabrics continued to be the second most important textile product of the Lower Rhine after Baumwollbiber (a warp of firmly twisted cotton yarn with a weft of loosely spun yarn of cheap Indian cotton). After the crisis of the 1860s and the boom phase in the 1870s, the production of trouser fabrics was restarted – with higher quality results than in the 1840s. This article was especially suited for production on mechanical looms.19 In wool processing, heavy and expensive carded yarn fabrics were displaced by lighter fabrics from the seventeenth century onwards. On the one hand, there were high quality fine draperies made from merino wool, and on the other hand, there were lighter and cheaper new draperies made from carded and worsted yarns.

17 Gerhard Kaldewei (ed.), Ansichten der Nordwolle 1884–1996; von der NWK 1884 zum Fabrikmuseum Nordwolle Delmenhorst 1996 (Oldenburg, 1996). 18 Blumberg, Textilindustrie, p. 291. 19 Willy Fränken, Die Entwicklung des Gewerbes in den Städten Mönchengladbach und Rheydt im 19. Jahrhundert (Köln, 1969), pp. 53ff. 207

The Ashgate Companion to the History of Textile Workers The shift in production to the finer drapery occurred as a result of the adoption and imitation of other kinds of finer fabrics that had already gained popularity in the markets. In the eighteenth century merino wool obtained through Amsterdam from Portugal and Spain replaced the English wool that had been traded through Brugge and Dordrecht since the seventeenth century (which was increasingly processed in England itself and from 1672 was banned from exportation). Towards the end of the eighteenth century wool from Saxony, Bohemia and Poland was increasingly processed, and, after cross-breeding with merino sheep, high quality wool was produced. Around 1803, the number of merino sheep in the electorate of Saxony alone amounted to around two million. In addition to patterned fabrics for specialist market segments, high quality single-colour fabrics in different gauges and glazes were produced for the Near or Middle East, carmine- or scarlet-dyed thick fabrics were produced for the Russians on a large scale, and other different fabrics for the French, Italian and Spanish markets. As a way of adapting to the modified market structures of this period, simple fabrics were also produced, as well as cashmere (made of fine worsted yarn for the chain and woollen yarn for the weft, softly milled and roughened for summer clothing). Worsted yarn was also interwoven in fabrics and in combination with cotton and silk was used in great quantities in stocking embroidery. In the 1860s, the transfer from expensive carded yarn fabrics to the cheaper and fashionable worsted yarn fabrics began. At the same time, the use of cotton and artificial wool for mixed fabrics in the woollen yarn industry as well as in worsted yarn industry began.20

Output volume and labour market In the 1780s, the textile industry in Prussia comprised 72.6 per cent of non-agrarian production. Wool processing with an output value of 37.7 per cent came just behind the linen industry (42.5 per cent). In 1846/47, the wool-weaving industry, measured by the number of looms, took third place behind the linen and cotton industries, yet it was the leader in exported fabrics. In the second half of the 1840s, the cotton industry employed 6.4 per cent of the workforce of the entire industry sector (including handicrafts). In 1861, the number of workers in wool processing amounted to 157,990 and in cotton processing to 193,000. In 1875, 15 per cent of the non-agrarian active 20 Karl Dechêne, ‘Die Entwicklung der Aachener Tuchindustrie in der preussischen Zeit bis zum deutschen Zollverein im Jahre 1834’ (Ph.D., Tübingen, 1922); Josef Strauch, ‘Die Aachener Tuchindustrie während der französischen Herrschaft (1794–1814)’ (Ph.D., Münster, 1927); Richard Wichterich, ‘Die Entwicklung der Aachener Tuchindustrie von 1815–1914’ (Ph.D., Köln, 1922); Hermann Weinberg, Die wirtschaftliche Entwicklung der Stadt Aachen von der Einführung der Gewerbefreiheit (1798) bis zur Gegenwart (Aachen, 1931); Ernst Barkhausen, Die Tuchindustrie in Montjoie ihr Aufstieg und Niedergang (Aachen, 1925). 208

The German wool and cotton industry population and accordingly 2.2 per cent of the entire population were employed in the German textile industry. In businesses with more than five employees, wool processing (136,000 workers) claimed second place behind cotton processing (148,000 workers) and above flax and linen processing (40,000 workers).21 Cotton spinning experienced an enormous boom as a result of the Napoleonic continental system. After the system ended, the industry could hardly keep up with British competition and it recovered only after the 1820s. During the Great Depression of the 1870s, growth slowed down again, only to accelerate in the 1880s.22 Table 8.5 Spindles, production, workers and added value in cotton spinning Spindles

Yarn produced Employees Net added value (tons) (millions of Marks)

1815

373,000

1,963

7,600

4.1

1825

390,000

2,499

11,900

5.1

1835

580,000

3,786

15,400

5.3

1845

830,000

13,093

18,500

9.8

1855

1,400,000

24,918

20,500

16.9

1865

2,350,000

37,128

31,900

22.7

1875

4,237,609

93,613

51,100

38.1

1885

5,000,000

127,950

53,700

38.0

1895

6,877,126

240,872

66,600

91.4

1905

8,832,016

350,939

78,200

161.1

1913

11,951,583

447,264

96,800

214.8

Source: Kiesewetter, Industrielle Revolution, p. 183.

The number of spindles increased only slowly after 1815 (1815–1825 around 4.6 per cent). Growth rates of over 40 per cent were not achieved until later decades. In 1911 Germany took third place behind Great Britain and the USA in cotton production. 21 Günter Kirchhain, Das Wachstum der deutschen Baumwollindustrie im 19. Jahrhundert; eine historische Modellstudie zur empirischen Wachstumsforschung (Münster, 1973), p. 246; Blumberg, Textilindustrie, pp. 131, 301, 302. 22 Kirchhain, Wachstum, p. 202. 209

The Ashgate Companion to the History of Textile Workers The increases in productivity were incomparably higher due especially to technological developments. In 1815, 5.1 kg yarn were produced on one spindle, in 1845, 14.6 kg (up 168.6 per cent) and in 1905, 39.7 kg (up 171.9 per cent). The growth of labour productivity was considerably higher than this. In 1815, one worker produced 258 kg of yarn, in 1845, 708 kg (up 174.4 per cent) and in 1905, 4,488 kg of yarn (up 533.9 per cent). While in a century the absolute number of workers in the cotton-spinning industry increased from 7,600 (1815) to 96,800 (1913), in the same period the number of spindles per worker increased from 49 to 124. Table 8.6 Looms, production, workers and added value in cotton weaving Looms

Fabric produced (tons)

Employees

Net added value (millions of Marks)

1815

40,000

3,600

62,200

16.2

1825

55,000

6,450

82,500

16.7

1835

87,000

14,237

130,000

33.6

1845

150,000

35,225

208,000

66.4

1855

n.a.

46,711

218,400

55.7

1865

194,250

41,294

n.a.

104.6

1875

226,000

101,415

227,000

167.4

1885

236,000

134,451

228,500

124.9

1895

251,250

242,878

229,100

167.2

1905

300,560

340,559

261,500

314.2

1913

343,200

426,837

284,000

380.2

Source: Kiesewetter, Industrielle Revolution, p. 183.

Although the cotton-weaving industry lagged behind the spinning industry in technological development – in 1861 the share of power-looms was only 12 per cent – it nevertheless showed a considerable increase in output as well as in productivity. In 1815, 90 kg of cloth were produced on a loom, whereas the output in 1846 was 253 kg (up 181.1 per cent) and in 1905 1,133 kg (up 347.8 per cent). Three periods of growth can be discerned: 1815–1840 when growth was relatively strong, and 1840–1860 and 1872–1913 when growth was relatively weak.23 The weaving 23 Kirchhain, Wachstum, p. 224. 210

The German wool and cotton industry industry developed particularly dynamically in the region of Gladbach/Rheydt where it increased by more than 6,000 looms and 10,000 workers. In particular the number of handlooms used in cottage industry continued to grow even in the late 1840s: between 1846 and 1849 from 9,942 to 12,062.24 From the 1830s, and accelerating in the 1840s, the cotton industry experienced a rapid boom in western Münsterland as it worked in close cooperation with the neighbouring Dutch region of Twente. The number of cotton looms increased from approximately 300 in 1831 to 12,600 in 1849 and 16,900 in 1858, and the number of machine spindles increased from less than 9,000 in 1843 to 15,200 in 1849, 30,800 in 1858, 500,000 in 1895 and 900,000 in 1903/05. Simultaneously, the number of machine looms increased from 628 in 1858 to 12,118 in 1892 and 20.611 in 1901. The basis of this boom was the drastic mechanization of the 1850s, initially, however, exclusively in the area of spinning: in 1849 there were just two mechanical looms in use in the administrative district, yet their number leaped to 1,317 in 1861. The number of cotton spindles in the administrative district also saw a parallel increase from 24,220 in 1849 to 59,900 in 1861.25 The growth in labour productivity which accompanied the mechanization resulted in employment stagnation in the spinning sector in Saxony. Around 1800 approximately 14,000 to 18,000 manual spinners were recorded in the Vogtland and in the Erz Mountains. For 1861 the number of workers in the Saxony cottonspinning industry can be estimated at approximately 14,000.26 After the Second World War, the number of spindles in the West German cotton industry rose from 5,807,291 (1950) to 6,684,951 (1952) and then fell until the end of the 1960s. The share of the entire textile industry in the industrial workforce amounted to 11.1 per cent in 1950 and had sunk to 6.2 per cent by 1962. The share of the cotton industry alone in the number of industrial employees in 1950 was 3.7 per cent and had dropped to 1.4 per cent by 1968. The number of cottonspinning mills fell in the same period from 160 to 139 and the number of cottonweaving mills fell from 513 to 315. The number of spindles was reduced by a good third, the number of looms by more than half. The number of workers sank from 24 Fränken, Entwicklung, p. 56. 25 Reinhard Schüren, Staat und ländliche Industrialisierung; sozialer Wandel in zwei Dörfern einer deutsch-niederländischen Textilgewerberegion 1830–1914 (Dortmund, 1985); Clemens Wischermann, ‘Die Industrialisierung des Baumwollgewerbes im Münsterland’, in Karl Ditt and Sidney Pollard (eds), Von der Heimarbeit in die Fabrik; Industrialisierung und Arbeiterschaft in Leinen- und Baumwollregionen Westeuropas während des 18. Jahrhunderts (Paderborn, 1992), pp. 192–222; Karl Ditt, ‘Die Industrialisierung in Baumwollund Leinenregionen Europas; eine Einführung’, in ibid., pp. 1–42, at p. 7 ; Gerhard Adelmann, ‘Zur regionalen Differenzierung der Baumwoll- und Seidenverarbeitung und der textilen Spezialfertigungen Deutschlands 1846–1907’, in Hans Pohl (ed.), Gewerbe- und Industrielandschaften vom Spätmittelalter bis ins 20. Jahrhundert (Wiesbaden, 1986), pp. 284–345. 26 Hubert Kiesewetter, Industrialisierung und Landwirtschaft. Sachsens Stellung im regionalen Industrialisierungsprozeß Deutschlands im 19. Jahrhundert (Köln/Wien, 1988), pp. 441, 455. 211

The Ashgate Companion to the History of Textile Workers 70,940 to 40,504 in the spinning industry and from 100,289 to 69,560 in the weaving industry. Nevertheless, the quantitative output increased due to technological ameliorations by 28.7 per cent in the spinning industry and by 17.2 per cent in the weaving industry.27 In the wool-processing industry, worsted yarn production and processing increased noticeably faster than carded yarn spinning and weaving, especially from the middle of the nineteenth century. In 1860/68, the index of production (compared to 1834/38) lay at 969 for worsted yarn weaving and at 261 for carded yarn weaving. The growth index of the cotton-weaving industry amounted to 295.28 An exact determination of the output volume of the Aachen cloth industry was not possible for the entire region until the nineteenth century. The development in Monschau, which was representative for the region, during the last quarter of the eighteenth century makes it clear that a constant upward trend occurred into the 1780s, after which decline set in. Table 8.7 Output volume of fine drapery industry in Monschau, 1775–1794 Output (Kümpfe)

Output (Kümpfe)

1775

2,588

1785

3,935

1776

2,389

1786

4,223

1777

3,006

1787

4,325

1778

3,480

1788

3,970

1779

3,821

1789

3,283

1780

3,770

1790

3,783

1781

3,967

1791

3,740

1782

4,151

1792

3,371

1783

3,710

1793

2,876

1784

3,676

1794

2,597

Source: Barkhausen, Tuchindustrie, pp. 55f.

As a result of the market expansion during the French period, the output volume, the production value and the employment numbers initially increased rapidly. Between 1810 and 1814, production fell by 22 per cent due to cyclical 27 Salzmann, ‘Wachstumsraten’, pp. 15, 51, 58, 97, 116. 28 Blumberg, Textilindustrie, p. 39. 212

The German wool and cotton industry influences. In the administrative district of Aachen, a total of 9,572 workers were employed in nine locations in the cloth industry in 1811. After the end of the French occupation and the restructuring of the markets, the output volume fell again to approximately one-third. The domestic market could not replace the lost markets of Napoleonic times. It was not until the 1830s that the economic cycle recovered through the opening of new markets and to a certain extent through the expansion of the domestic market with the formation of the tariff union. Statistics, collected from 1834, show a moderate growth in production (80,000 to 88,820 units) until 1847. The 30 per cent growth in value added (12.0 to 15.6 million Marks) was much steeper. Until 1870, the output volume grew to 200,000 units with a production value of 36 million Marks. In 1861, 293 male and 318 female workers were working in 39 spinning mills (number of fine spindles 70,475). There were 224 masters and 2,316 assistants operating 2,272 machine looms and 4,184 male and 1,897 female workers operating 1,019 manual looms. The high demand for cotton on the North American market after the end of the Civil War (1865) led to a shortage of workers in the Aachen cloth industry. The Chamber of Commerce regretted an increase in wages due to the introduction of compulsory education. In 1876/77, a total of 8,375 people worked in the textile industry in Aachen and Burtscheid. This number had increased to 15,466 by the beginning of the First World War.29 During the interwar depression, the number of workers fell to 6,000 (1931), then rose briefly to 9,300 before the start of the Second World War. After the war, the cloth industry recovered for a while. In 1953 the industry employed 8,000 persons. The number of businesses as well as the number of workers then fell, to 64 and 5,525 respectively by 1960. At this time, another 96 businesses existed, employing 8,616 workers who processed other textile resources and produced different products.30 Due to the crisis in camlet production, the output of the Upper Swabian cottonprocessing industry sank drastically in the seventeenth century. In Augsburg alone, a decline in production from 500,000 units (at the beginning of the seventeenth century) to 50,000 units (between 1635 and 1646) could be recorded. The number of weavers fell from approximately 3,000 (in 1601) to 550 (in 1677). From the middle of the eighteenth century, camlet production became completely negligible. It was only through the production of cotton fabrics and the introduction of calico printing that Augsburg and the other Upper Swabian cities could record an economic improvement from the end of the seventeenth century. Other centres of camlet production (Ulm and Memmingen) shifted their activity to the production of linen.

29 Wichterich, ‘Entwicklung’, p. 181; Alphons Thun, Die Industrie am Niederrhein und ihre Arbeiter, 1. Teil: Die linksrheinische Textilindustrie (Leipzig, 1879), p. 19; Marieluise Schultheis-Friebe, Die französische Wirtschaftspolitik im Roer-Departement 1792–1814 (Bonn, 1969), pp. 209–210; Josef Dahmen, Das Aachener Tuchgewerbe bis zum Ende des 19. Jahrhunderts, 2nd edn (Berlin, 1930), pp. 132–133. 30 Clemens Bruckner, Zur Wirtschaftsgeschichte des Regierungsbezirks Aachen (Köln, 1967), p. 327. 213

The Ashgate Companion to the History of Textile Workers In the north of Germany the three biggest mills in Nordhorn employed between 9,000 and 11,000 people in total, between the late 1920s and early 1970s. Of these workers about 35 per cent were engaged in spinning and preparatory work, about 50 per cent in weaving and about 15 per cent in dyeing, printing and other finishing operations. Compared with the number of workers within the cotton industry throughout Germany, however, these proportions were rather small and formed a mere 2.4 per cent.

Markets The semi-woollen fabrics of the Rhenish cotton industry were primarily exported to South America from the 1870s onwards. After the formation of the Reich in 1871, the competition from the cotton industry in Alsace-Lorraine was especially noticeable at the time of the depression. Nevertheless, the industry profited from a strong and rapidly growing domestic demand. At the same time, more and more products from the textile industry in the Netherlands and Münsterland were exported, especially to England, Belgium and Switzerland as well as to Central and South America. The import duty on yarn, imposed in 1879, kept English competition out.31 Table 8.8 Export shares of single sectors in the textile industry 1828

1837

1850

1864

1869

Linen

47

28

29

9

10

Silk

20

15

21

23

32

9

19

19

13

14

Wool

24

38

31

55

44

Total

100

100

100

100

100

Cotton

Source: Blumberg, Textilindustrie, p. 289.

The Saxon cotton industry also had an international focus. Around 1800, the Saxony cotton region exported muslin worth around one million thalers to the Levant. After the foundation of the tariff union in 1834, Chemnitz was one of the leading centres for the production of printed tablecloths (together with Elberfeld), and much of this merchandise was sold on overseas markets.

31 Fränken, Entwicklung, p. 77. 214

The German wool and cotton industry Due to the demand from the army and neutral states at the beginning of the First World War, the cotton industry in Chemnitz experienced a temporary boom, but unemployment never fell below the pre-war level of around 10 per cent. From 1920 the Chemnitz cotton industry participated to a disproportionate extent in the inflation that was part of the trade cycle between the wars.32 Due to the backlog in demand, an import surplus developed after the Second World War, but this had already transformed into an export surplus by the 1950s. From the end of the 1960s the West German cotton industry came under pressure from cheap foreign competition. Despite a fourfold increase of the export volume compared with 1950 and a twofold increase compared with 1958, the ratio of exports to imports sank to 54 per cent from 1968.33 In the wool-processing industry, Prussia – much like England and France – sealed off its domestic market before the nineteenth century, also opposing the producers of its western provinces, in favour of domestic producers (Berliner Lagerhaus). Unlike the south-western German wool-processing industry, whose simple products mainly went to the domestic market, and to a lesser extent to Switzerland and Italy, the Aachen fine-cloth industry relied on remote markets. In addition to the different markets in the Reich that were served by the Frankfurt and Leipzig Trade Fair, a large part of the production went to the southern Netherlands, to Switzerland, to France and the Baltic. In the Middle East, so-called seraglio cloths were delivered via Vienna and during the French period also via Marseille. This market had opened up in the second half of the eighteenth century because of the downturn in the production of this sort of cloth in southern France. The entire spectrum of cloth production reached Russia, a market which promised a high profit margin, but long delays in payment. This market alone provided employment for 5,000 workers.34 The domestic market remained subordinate for almost the entire nineteenth century. Until the foundation of the tariff union in 1834, the political fragmentation and tariff barriers were obstructions to trade. However, the crucial cause was the lack of consumption power despite an increasing population. The average export quota of approximately 20 per cent for the entire German wool industry in the 1830s was far exceeded by individual regions. In the region of Aachen, exports amounted to 70 per cent. The traditional markets in Europe and the Middle East were increasingly replaced by the North American market. The latter, however, was subject to extensive fluctuations. In 1895, the share of German woollen cloths in North American imports amounted to 13.2 per cent. With that figure, Germany was in second place ahead of France (4.4 per cent) and Belgium (3.4 per cent), but far behind England (78.1 per cent). The markets in South America and the Caribbean 32 Karlheinz Schaller, „Radikalisierung aus Verzweifelung“; Geschichte der Chemnitzer Arbeiterschaft vom Ersten Weltkrieg bis zur Inflation (Bielefeld, 2003), pp. 53, 130–131. 33 Salzmann, ‘Wachstumsraten’, p. 86. 34 Strauch, ‘Aachener Tuchindustrie’, pp. 58–59; Anton Seidl, ‘Die Aachener Wollindustrie im Rahmen der Rheinischen bis zur Gewerbefreiheit 1798’ (Ph.D., Köln, 1923), p. 21; Wichterich, ‘Entwicklung’, p. 58. 215

The Ashgate Companion to the History of Textile Workers played a relatively minor role in the German wool industry as both markets continued to be mainly the domain of the English. The market of the Middle East, which was reanimated after the Crimean War (1853–1856) and was served mainly by Silesian producers, was still important. Trade with Russia and from there to Asia decreased due to high Russian import duties; when these were lowered in 1854, trade grew. Export to East Asia experienced a growth in the 1860s, especially exportation to China and Japan. In Western Europe, France and Belgium, and also Spain, Portugal and the Scandinavian countries played no greater role as export markets due to their high import duties. Italy remained important as a market area. Competition from Austria could be ousted as a result of the independence movement there. In sum, the sales of the 1840s were not reached again. Export to the Netherlands and its colonial regions remained stable. Even though the capacity of the domestic market had increased since the 1860s, due to the general industrial and demographic development and the rise in the price of cotton as a result of the American Civil War, the transfer from foreign markets to domestic markets only took place in the 1890s. This was partly promoted by the prohibitive customs policy of the United States and its developing industry.35

Technological development In 1783, Johann Gottfried Brügelmann, a businessman from Elberfeld, undertook an initial attempt at the mechanization and centralization of cotton spinning with the installation of a mechanical cotton-spinning mill using hydraulic energy in Ratingen, according to the English model.36 The Napoleonic era is considered to be the phase when breakthroughs towards the mechanization of the cotton-spinning industry in Saxony occurred. The number of mule-spindles increased from 13,000 in the year 1806 to 256,000 in the year 1813 and 283,700 in 1816. In the mid-1840s, 116 machine cotton-spinning mills were recorded and the number of spindles increased to 475,000. By 1830 at the latest, the technological backlog had been overcome. The technology was now sufficiently developed to produce internationally competitive fine yarn. The number of spindles in this branch of production increased rapidly and was accompanied by an increase in the number of businesses. Within the German tariff union, Saxony attained a

35 Blumberg, Textilindustrie, pp. 211–275. 36 Eckhard Bolenz, Johann Gottfried Brügelmann; ein rheinischer Unternehmer zu Beginn der Industrialisierung und seine bürgerliche Lebenswelt (Köln, 1993); Burkhard Dietz, ‘Aufbruch zu neuen Ufern; Johann Gottfried Brügelmanns Entscheidung für den Standort Cromford und die Vorbildfunktion einer erfolgreichen Firmengründung’, in Gerda Breuer (ed.), „Die öde Gegend wurde zum Lustgarten umgeschaffen…”; Zur Industriearchitektur der Textilfabrik Cromford 1783–1977 (Köln, 1991), pp. 16–33. 216

The German wool and cotton industry leading position. By 1861 Saxony’s share had fallen to just under a third, whereas Bavaria, Prussia, Württemberg and Hanover had caught up considerably.37 Around 1840, the mechanization of cotton spinning was completed in Bavaria (Nuremberg-Fürth, Hof and the Bavarian Vogtland, Kaufbeuren, Kempten and the Palatinate). By the middle of the 1860s, self-actor spindles were dominant in the spinning industry. Nevertheless, Germany clearly lagged behind England and other states in the introduction of modern spinning methods. While the self-actor had already been launched in the 1830s in England, it only became widely accepted in Germany in the 1850s. Germany also lagged approximately 20 years behind with the conversion to the highly productive ring spinning method (launched in the USA in the 1860s and in England in the 1870s). The spread of this new spinning method did not proceed simultaneously in the different regions of Germany. The southern regions and Alsace-Lorraine overcame the head start of Prussia and especially Saxony. This was due to their low wage level. Table 8.9 Percentages of self-actor spindles (1), and water and ring spindles (2) Bavaria

Württemberg

Alsace-Lorraine

Prussia

Saxony

1

2

1

2

1

2

1

2

1

2

1875

89.4

6.6

71.9

13.4

94.5

0.2

56.3

41.3

48.0

20.8

1913

46.2

53.8

33.2

66.8

46.4

52.5

31.9

63.3

80.0

19.3

Source: Kirchhain, Wachstum, pp. 45–46.

Until well into the second half of the nineteenth century, production at home and in small manual-weaving mills was dominant. In 1861, their share in the total number of establishments was still 86.6 per cent. Admittedly, in the 940 businesses registered as factories there were nearly double the number of machine looms operating. The share of employees in the small businesses amounted to 82 per cent. In Saxony especially, the cotton-weaving industry insisted on manually operated production methods. In machine-weaving mills, Saxony was placed only in sixth position among the German states. In the chamber district of Gladbach the number of mechanical looms in 1874 already amounted to 6,284, while simultaneously the number of handlooms had sunk in the early 1870s from 5,249 to 1,491. This radical trend towards mechanization was primarily due to high raw material prices. The enhancement of manual looms allowed only a moderate rise in working speed compared with the improvement provided by the power-loom. In 1829 the number

37 König, Sächsische Baumwollindustrie, pp. 248–353; Kiesewetter, Industrialisierung, p. 449. 217

The Ashgate Companion to the History of Textile Workers of strokes on handlooms lay at 25 and at 75 on power-looms, while by 1913 these figures were 40 and 200 respectively.38 Another thrust of innovation set in after the First World War. The cotton-spinning mill and weaving mill in Esslingen in Württemberg took up machine weaving in 1924. After the Second World War, a series of technological ameliorations (including high-performance carding machines and heavy-draft flyers) were introduced and led to lower demands for labour. In spinning machines alone, the growth in speed halved the demand for labour.39 The mechanization of labour-intensive spinning also began in the woolprocessing industry. With the application of English technology, the Aachen cloth industry assumed a leading role. This mechanization occurred on a large scale at the time of the French Occupation (1798–1814). The technology, developed in the Rhineland, also expanded to the Bergish wool industry. In 1817, 21 shearing machines were already in use, and by 1829, 57 shearing machines and 16 roughing machines were in operation. In 1819 there were 10 independent machine-spinning mills. The changeover to mechanical looms continued into the second half of the 1850s and was spurred by the transfer towards the production of simple fabrics. However, in 1861, 23 mechanical looms and 145 handlooms were in operation, with only a slightly higher performance from the power-looms. The final switch-over to mechanical looms took place at the end of the nineteenth century.40 However, only the further development of the pre-spinning machines made the production of high quality and fine yarn possible. With the introduction of the mule-jennies in the worsted yarn spinning industry of the 1830s – and later in the carded yarn industry – the number of spindles could be increased considerably. In the region that was part of the tariff union in 1846, 429,469 fine spindles were operational in carded yarn production and 125,955 fine spindles in the worsted yarn production. The self-actor was also employed in the mid-1850s, first in worsted yarn spinning, then later in carded yarn spinning at the end of the 1850s. The mechanization of the spinning mills effected a much greater increase in productivity than in other sectors of production. In the 1840s a worker working on a mechanical spinning machine already produced 500 times what he could produce through manual spinning. Due to the initially insignificant increase in productivity compared to the manual-weaving mills, the mechanization of the weaving mills by the power-loom took place substantially more slowly, and this was also the case in wool processing. A series of technological improvements eventually led to the renunciation of the handloom in the 1870s, however, even in the mid-1870s, only one-third of woollen fabrics were produced in mechanized weaving mills. 38 Kirchhain, Wachstum, p. 237. 39 Salzmann, ‘Wachstumsraten’, p. 61. 40 Blumberg, Textilindustrie, p. 75; Isenburg, ‘Untersuchungen’, pp. 26–29; Friedrich Zunkel, ‘Gewerbe- und Industrielandschaften von der Frühindustrialisierung bis 1914: Wolle’, in Hans Pohl (ed.), Gewerbe und Industrielandschaften vom Spätmittelalter bis ins 20. Jahrhundert (Stuttgart, 1986), pp. 254–283. 218

The German wool and cotton industry The mechanization of the finishing process (mainly in shearing mills) began parallel to the mechanization of the spinning mills. However, this was initially concentrated in individual regions. The application of steam power for the starting of the machines set in unequally in time and space and stood in a competitive and complementary relationship with the starting of the machines by water power. It did not supply three-quarters of the required energy until the 1870s.41

Organization of production The domestic industry took root in the countryside. Except where uncooperative institutions opposed the putting-out system, this system developed in cotton weaving throughout Saxony. With a few exceptions, the same applied until the 1830s in the Rhineland. The only centralized production facility of the eighteenth century was the cotton-spinning mill in Ratingen, founded after the English model in 1783 by Johann Gottfried Brügelmann, a businessman from Elberfeld. The installation of further centralized cotton-spinning mills and weaving mills in the first half of the nineteenth century took place only hesitantly. One cause for this was the lack of technological competence and the lack of modern mechanical engineering coupled with extremely strong English competition and, secondly, the noticeable lack of adequately qualified machine spinners. While the spinning industry in the town region of Gladbach/Rheydt continued to develop slowly, the number of operating spindles kept increasing slowly from 14,587 in 1836 to 16,950 in 1849. In the rural district of Gladbach, 29,470 spindles were operating in 1849 with 1,218 workers working on them. At the beginning of the 1850s, manual home spinning reached its climax, then rapidly fell due to the mechanical spinning mills that began breaking through on a large scale from 1854. The centralization of weaving in factories occurred significantly later than spinning. Around 1860, the weaving industry in western Münsterland and on the Lower Rhine was still almost exclusively organized as a cottage industry. However, the mechanization of cotton weaving was making rapid progress in the boom years following the formation of the Reich and from the 1880s onwards. In the course of the following decade, just over 1,000 manual looms that had been recorded in 1880 disappeared. Apart from a few successful individual foundations, the transfer of the cotton industry to a factory industry did not succeed in the region of Mark/Berg or in the Siegerland in the mid-nineteenth century. As a reason for the marginalization of this branch of business, which was still of great importance at the turn of the nineteenth century, contemporaries mentioned the relatively high wages and the relatively good opportunities to earn money in these regions where the iron and steel processing industries were strongly developed. For example in the Siegerland, all five cotton-spinning mills operating in 1843 were closed down by 1855. In the 41 Blumberg, Textilindustrie, pp. 89–91, 112, 132. 219

The Ashgate Companion to the History of Textile Workers middle of the nineteenth century, the Rhenish-Westphalian cotton industry was concentrated in the two production centres on the left bank of the Lower Rhine and in western Münsterland, two noted low-wage regions.42 As in other sectors of textile production, the expansion of the production and the intensification of competition in wool processing during the seventeenth and eighteenth centuries lead to ruralization and accordingly to the regionalization of production. The high standards in quality of production and the utilization of expensive raw materials, however, set limits to this process, especially in the production of fine cloth. Due to the lack of government-protected domestic markets the pressure for modernization was already high in the proto-industrial phase. The reduction of wage costs was possible due to the recruitment of the rural workforce in spinning and, to an extent, in weaving. At the end of the eighteenth century a great part of the spinning and weaving mills as well as the associated operating processes were already settled in the urban centres of cloth production, yet remained outside the central workshops in which the finishing work (shearing, dyeing) took place (operating as a decentralized manufacture system). The shift to the factory system was gradual. A complete integration of all production sectors was the exception. The deterioration of the sales situation after the French period accelerated the concentration process at the expense of the small, financially weak businesses. In the 1830s, the process of centralization in the Aachen cloth industry was largely complete. The mechanization of spinning mills did not result in the immediate formation of large firms everywhere. The first jennies were suitable for the maintenance of domestic production. The beginnings of industrial machine-spinning mills go back to the first decade of the nineteenth century. In 1846 approximately 93 per cent of worsted yarns and at least 80 per cent of carded yarns were already produced in spinning factories. In the worsted yarn spinning industry, the share of factory businesses increased to 63 per cent by 1858, and the share of spindles even increased to 98 per cent. In comparison, in the weaving industry the share of factory production in worsted yarn and semi-woollen processing lay below 10 per cent; in the carded yarn industry it was approximately twice as high. In 1875 only one-third of the workers in the weaving mills worked in businesses with more than five employees, whereas in the spinning mills this figure was already 82 per cent.43 By the 1870s, the concentration process was particularly far advanced in Brandenburg (84 per cent), Silesia (84 per cent) and the Rhineland (79 per cent). Among the leading regions Saxony was lagging behind (50 per cent). The carded

42 Fränken, Entwicklung, pp. 50, 81. 43 Blumberg, Textilindustrie, pp. 112–113, 122–123; Karin Zachmann, ‘Die Durchsetzung des kapitalistischen Fabriksystems in der deutschen Textilindustrie des 19. Jahrhunderts aus der Sicht der Verdrängung vorindustrieller Produktionsformen’, Jahrbuch für Wirtschaftsgeschichte 3 (1991): 31–52, at p. 47. 220

The German wool and cotton industry yarn industry in Alsace-Lorraine, which was annexed in 1871, was almost completely organized into large firms.44 Wool weaving in cities in the south-west of the Reich as well as in country towns in the Duchy of Württemberg, the Baden Margraviate and smaller territories kept the character of an urban craft guild. In 1622 dyers and traders in Calw in Württemberg merged to become a fabric trade company, which possessed a quasi-monopoly over the dye works and the distribution of the production of the so-called Calwer Moderationsbezirk. The state exercised extensive influence on businesses through the allocation of cooperative privileges to existing institutions (guilds and trading companies). As a return service for these privileges, the corporations assisted the state in its military, fiscal and regulatory duties. Attempts to further the textile industry by the well-directed foundation of manufactories led to no definite success, even though 31 manufactories were constructed between 1740 and 1833 in Kurbayern, of which 10 were for the silk industry and nine each for wool and cotton manufacture. A tendency towards a Kaufsystem rather than towards the putting-out system can be seen in the region of Eastern Swabia. In Augsburg putting-out was prohibited by the organization of the weavers. When putting-out started to rear its head in the countryside, it was prohibited there too. Only towards the end of the eighteenth century did putting-out increase in importance. In a similar way to the cities, the countryside was organized into guilds, with the difference that the guilds were spatially partitioned. Therefore, there were guild sectors that coincided with the district of sovereignty of a territory and comprised only one craft. Furthermore, there were accumulative guilds that incorporated several crafts equipped with only a few masters within one territory. Finally there were comprehensive territorial guilds that subsumed the craftsmen of several sovereignties. A few figures can demonstrate that the crafts in the countryside were an absolute necessity for textile production in the early modern age. In the territory of Ulm, the number of country weavers was approximately 330–600 in the year 1513 and this increased to over 1,000 towards the end of the seventeenth century. In the territory of Memmingen, approximately 300–600 country weavers worked in the late seventeenth century and in Augsburg in 1600 only half of the production level of 500,000 camlet units was manufactured by the weavers of Augsburg, suggesting that the other half was probably produced in the vicinity of the city.

Workers Textile workers differed from other groups in the industrial workforce because of their roots in the pre-industrial period and the fact that they needed to have some skills; some of them even needed a high level of qualification and specialization. 44 Blumberg, Textilindustrie, pp. 296–297. 221

The Ashgate Companion to the History of Textile Workers Qualified workers in the factories were primarily recruited from the crafts. In the weaving industry, the factory system eliminated the relative independence of the weaving masters and created the institute of the foreman, on whose technical and organizational knowledge those manufacturers who were operating as businessmen without technical expertise were dependent to a great extent. Unlike the handweaver, the skills needed by the machine weaver were marginal. In the 1860s, a training period of four to six weeks was enough to operate a machine loom.45 According to the demands on the cottage industry, the urban trade, the locally concentrated home businesses and the wage labour in larger workshop units (manufactories), the spectrum of workers in the Rhenish cloth industry was extraordinarily versatile before the factory era. In the rural cottage industry the integration with agriculture was already partly dissolved; only rarely did the family form a home-production unit. Only to a small extent was the workforce recruited from the local population in the new textile production locations that were rapidly growing during the eighteenth century. The quick transfer to fine cloth production was only possible due to the recruitment of skilled workers (shearers and dyers) from outside. Some came from the French and Belgian cloth industries and some were recruited from the Aachen craft guilds, but the bulk of weavers originated from the surrounding countryside In Augsburg, where more than 3,000 weavers lived at the time of a production peak shortly before the Thirty Years’ War, the weaver craft could not recruit sufficient numbers locally, and had to recruit from the area around Swabia, but also from as far as Silesia and Pomerania. They dominated the weaving workforce with up to 68 per cent coming from further afield in the eighteenth century. Unlike the urbanborn weavers, the immigrants who had recently settled in the town could usually fall back upon a small plot of land. In the weaving village of Langenneufnach, located to the west of Augsburg, at least 113 of the 263 weavers in the seventeenth and eighteenth centuries possessed agriculturally useful ground, a situation that tended to apply to most of the localities in the country.46 The factory workers of the Aachen cloth industry also came partly from the surrounding villages and partly from Belgian and Dutch border locations and they had previously lived in the countryside as home-weavers. In the 1870s many winders, spinners and weavers commuted daily from the surrounding villages to the factories in Aachen and accepted two hours of daily travel or slept in the factory during the week.47 The first generation of early industrial workers in Württemberg was recruited from nearby, for example from Heidenheim an der Brenz. In Esslingen in the 1880s, 30.4 per cent of the skilled, but only 5.8 per cent of the unskilled workers 45 Jürgen Kocka, Weder Stand noch Klasse; Unterschichten um 1800, Bonn 1990; Jürgen Kocka, Arbeitsverhältnisse und Arbeiterexistenzen; Grundlagen der Klassenbildung im 19. Jahrhundert (Bonn, 1990). 46 Clasen, Textilherstellung, pp. 15–89. 47 Blumberg, Textilindustrie, p. 311; Wichterich, ‘Entwicklung’, p. 167; Thun, Industrie, pp. 40, 63. 222

The German wool and cotton industry were commuters. It was not until the following decades that the share of unskilled commuters increased to 52.2 per cent (1909/14), while the share in skilled workers decreased to 18.3 per cent. In the early cotton-spinning mills, the majority of the workers with basic occupations (stretching, carding, spooling, reeling, pre-spinning, twisting and fine spinning) were either employed in purely manual labour or worked on the still primitive machines. In the spinning mills in Chemnitz, for instance, only 2 per cent of the workforce were supervisory staff. Another 42 per cent were employed in preparatory work, 49 per cent in fine spinning and 7 per cent in reeling. Of these workers 31 per cent were men, 36 per cent women and 33 per cent children. Because of the unhealthy and badly paid factory work, many workers made use of the convenient economic cycle during the Continental System to work as unskilled workers in calico printing or to turn back to home weaving and manual spinning. Few of the men in the spinning factories had the opportunity to qualify as mechanical engineers.48 Despite the lack of area-wide data, it is assumed that girls and women in the textile industry made up the mass of the workforce in manual spinning. At the beginning of the nineteenth century in the region of Aachen, 82.1 per cent were female, of which a quarter were younger than 20 years old. Female workers also had an over-proportional share in diverse auxiliary functions. In the woolprocessing industry, besides wool washing, this particularly concerned mending work on finished cloth. In contrast, manual weaving remained reserved exclusively for men. The same applies to the qualified functions in finishing.49 In cotton-spinning mills, about 5 per cent more female workers were hired than male workers throughout the nineteenth century.50 The numbers for Prussia from the 1840s and 1850s prove that the share of adult women in mechanical woolspinning mills increased consistently, particularly in worsted yarn production (1858: 66 per cent). Weaving continued to be a male domain in the factory as did centralized manual weaving until the 1870s. In Prussian cloth and woollen factories, men’s share in the workforce in the 1840s and 1850s amounted to 25–30 per cent. It was not until the widespread introduction of machine looms in the 1870s that the share of female workers increased.51 In those parts of the proto-industrial regions that were strongly influenced by guild structures, a policy prevailed which excluded women: even though the Reich’s valediction of 1771 explicitly admitted women to the crafts, Saxon guilds restricted entry to the wives and daughters of guild members. Nevertheless, 146 employed women were counted in the weaving industry in Chemnitz in 1847, and they were probably employed as relatives of the 1,274 weaving masters with jobs 48 Karlheinz Schaller, Einmal kommt die Zeit; Geschichte der Chemnitzer Arbeiterschaft vom Ende des 18. Jahrhunderts bis zum Ersten Weltkrieg (Bielefeld, 2001), pp. 50–57. 49 Dietrich Ebeling et al., ‘Die Aachener Textilregion zwischen Ancien Régime und Fabrikzeitalter’ (forthcoming). 50 Kirchhain, Wachstum, p. 164. 51 Blumberg, Textilindustrie, pp. 345–348. 223

The Ashgate Companion to the History of Textile Workers in all areas of weaving. Complaints against the work of women on weaving and knitting looms were addressed to the weaving guilds in the district authority of Zwickau until 1850.52 In Württemberg, the participation of women in the production of cloth was widely restricted to spinning and unskilled work until the 1830s. After the introduction of the flying shuttle, women were also allowed to operate looms. Once the trade regulations of 1828 had declared the production of cotton yarn and cloth as work not bound to guilds, women were allowed entry to the weaving industry. Manufactories, and later factories, freed women and girls from their close links to the family household. Until the end of the nineteenth century, hardly any married women worked in the textile industry in Württemberg; a mere 5 per cent of female workers were married. In the year 1900, new trade regulations came into effect that strongly regulated female labour. Women were no longer allowed to work at night between 8:30 p.m. and 5:50 a.m. as well as no later than 5:30 p.m. on Saturdays and on the days before holidays. The daily work schedule was not allowed to add up to more than 11 hours and women who gave birth were not allowed to work for four weeks afterwards. The spread of female factory workers led to the establishment of nursery schools in a series of firms. The use of children from the age of 6 was one of the most important advantages of the rural cottage industry. The decline in child labour in mechanical woolspinning mills was less due to legal regulations than to technological progress in pre-spinning. Child labour was extremely high in several regions of the Rhineland, in Württemberg and in Bavaria. Until the middle of the 1870s, child labour represented 3–4 per cent of the workforce; in the Saxon worsted yarn industry it was even as high as 7 per cent. In the wool-weaving industry, child labour remained widespread until the demise of cottage industry. Children were particularly useful for yarn spooling.53 In machine weaving, children were deployed for knitting together broken threads. The percentage of children employed in woollen cloth factories remained particularly high until the 1860s. During the second half of the 1840s, it was around 15 per cent and corresponded to that of the cotton, linen and silk industries. The laws to protect juvenile workers of 1839 and 1863, which prohibited work until their thirteenth birthday and restricted the work schedule for school-age children, met with tough resistance from the businessmen who had to pay higher wages due to the additional employment of adults. In 1860 the Aachen Chamber of Commerce demanded a change in the law that would allow ‘that children, as soon as they had

52 Karin Zachmann, ‘Die Kraft traditioneller Strukturen; sächsische Textilregionen im Industrialisierungsprozeß’, in U. John, and J. Matzerath (eds), Landesgeschichte als Herausforderung und Programm (Stuttgart, 1997), pp. 509–535, at p. 518; Kiesewetter, Industrialisierung, p. 476. 53 Blumberg, Textilindustrie, pp. 353–357. 224

The German wool and cotton industry completed their school attendance, should be admitted to 10 hours of work in the factories per day’.54 In the dispersed cloth industry in the region of Aachen in 1812, children, in addition to their use as unskilled labour, were deployed in the spinning mill from as young as 3 years of age. Of the children employed in the spinning mills up to the age of 16, almost two-thirds were girls. However, girls and women were not employed at all in yarn spooling, where two-thirds of the work was done by boys who were employed in weaving from the age of 12. Altogether 13 per cent of the entire workforce in the cloth industry was younger than 16 years old. In spite of compulsory schooling, which was introduced in 1863, in 1864 there were 200 workers aged between 12 and 14 years old and 1,060 workers between the ages of 14 and 16. This made up a share of 13.7 per cent of the entire workforce.55 Child labour was self-evident during early industrialization in Baden and Württemberg, as children needed to earn a portion of the family income. Cotton factories had the greatest percentage of child and female labour of all industrial businesses. Incidentally, the textile companies of the nineteenth century predominantly employed younger workers. Performance on the job decreased notably among workers over the age of 40 and there were hardly any older workers who were able or willing to accomplish the piecework. In Saxony too, supervisory school authorities, teachers and priests fought child labour and promoted factory schools from the late 1830s on. It was not until almost a quarter of a century after they appeared in Prussia that the legal restrictions on child labour reached Saxony.56 Nominal wage development in the German wool industry proceeded very unevenly between the early 1840s and the early 1870s. In the 1830s, maximum weekly wages in the wool-spinning mills in Württemberg averaged 7.30 Marks. They increased to 11.50 Marks by the 1860s. Similar wages were paid in cottonspinning mills. In different regions comparable wage levels and a similar wage development could be observed. Wages were clearly higher in large cities (Aachen, Berlin, Augsburg, Leipzig). Between 1860 and 1872 the difference in the wages of female and male workers oscillated at around 50 per cent for the same occupation. However, since men were usually employed in machine work or undertook supervisory functions, the actual gap was even greater. In times of crisis, businesses lowered their wages by up to a third or even a half. The nominal wages that had increased only moderately in the first phase of industrialization between the 1830s and the 1860s could not keep pace with the cost of living. In the first half of the 1860s, the workers in the wool-spinning mills had to accept a loss in spending capacity of 35 per cent. It was only with the boom of the early 1870s that the real wage level of 1860 was reached again. In other branches of industry real wages rose faster.

54 Wichterich, ‘Entwicklung’, p. 171. 55 Dahmen, Aachener Tuchgewerbe, pp. 133–134. 56 Schaller, Einmal kommt die Zeit, pp. 68–73. 225

The Ashgate Companion to the History of Textile Workers Table 8.10 Wages in cotton spinning, 1815–1913 Annual wage (Marks)

Annual wage (Marks)

1815

165

1875

585

1825

250

1885

540

1835

330

1895

635

1845

355

1905

755

1855

385

1913

860

1865

390

Source: Kirchhain, Wachstum, pp. 158–160.

A comparison with wage development in the weaving industry shows that this can primarily be ascribed to the substitution of male labour by cheaper female labour. The inclusion of weaving into the factory system initially meant a reduction in wages of one-third for the formerly autonomous cloth workers.57 After the Second World War, wages in the West German cotton industry nearly quadrupled between 1950 and 1968, and the increase was clearly higher among the female workers than among the male workers. As the wages dropped approximately 20 per cent below the average for the entire industry, a decline in employment was unstoppable. However, within the European Economic Community (EEC) textile employment in Germany was still high.58 The transfer from home industry to factory industry was connected to a considerable extension in the daily work schedule for the workers. As most workers did not live in the immediate proximity of the factories, those who were not granted accommodation in the factory had a half to one hour daily walk to work and back. In 1850 a general investigation of the work schedule was carried out in the factories in Baden. While 15 hours were worked daily in the weaving mill Merian in Hoellstein, the work schedule in most cotton-spinning mills averaged 13 hours. On average, the daily working time amounted to 10 hours at the beginning of the nineteenth century. One century later, it had been reduced to 8.3 hours daily. After the Second World War, the average number of hours worked per week decreased from 50 in 1950 to 41 in 1970 and underbid the level in Italy, the Netherlands, Belgium (all at 45 hours per week) and Japan (48 hours per week). In the normal net working time, in which holidays were considered, the Federal Republic had the lowest level on an international scale with 1,900 yearly hours in 1967.59 57 Blumberg, Textilindustrie, pp. 317–325. 58 Salzmann, ‘Wachstumsraten’, pp. 120–121. 59 Ibid., pp. 127, 131. 226

The German wool and cotton industry Of the weavers, 88 per cent lived in three districts of Augsburg and paid low rents of 15 Kreuzer in 1625. Most weavers could not afford to have a separate house. In the year 1809, 46 per cent of the 552 weavers lived in groups of two or three per house. Another 20 per cent lived in houses in groups of four or five. Including their family members, up to 47 people could be found living under one roof.60 The urban concentration of production, which increased with factory industrialization, led to a deterioration of living standards. The existing houses were not enough and were used to their maximum by integrating rear buildings, barns and barracks. In 1875, as many as 85 people lived in a house in a typical workers’ housing area. Working conditions had a detrimental effect on health. Only 18 per cent of those examined for military duty in Saxony between 1852 and 1854 were spinners, and 18 per cent were weavers. Domestic commercial weavers suffered especially badly from abdominal and chest diseases. The exposure to dust in the spinning factories together with the high temperatures led to diseases of the respiratory organs in young, mostly female, workers. Additionally, it was discovered in 1907 that many workers suffered from anaemia. In the machine-weaving mills the almost permanently loud noise led to hearing problems. The poor living conditions caused a high death rate. Almost half of the children born to the workers of the Aachen workforce in the 1870s died before their first birthday.61 The textile workers’ trade unions and their actions must be studied in close relation to the organization of the sector. The craft roots of the workers continued to have an effect far into the factory period. For example, cloth-weaving assistants went on strike in Guben to have their fraternity recognized, but were unsuccessful.62 In the period preceding the factory era, conflicts between the putting-out companies and the manufacturing companies were dominated by regionalization and by the definition of the putting-out systems. Efforts of the cloth-makers to defend the local sites were not limited to the guild city of Aachen. In Monschau too, the assignment of orders to rural weavers created constant potential for conflict between the local weavers and the cloth manufacturers. In Aachen the small producers organized in guilds (masters) and their hired workers (assistants) likewise defended their respective labour markets. The cause of the unrest that occurred in 1830 was a deduction in wages in one of Aachen’s greatest cloth factories, which was felt to be arbitrary by the workers. Very quickly a large crowd (approximately 1,000 persons) formed and turned against the introduction of the new machines. However, they did not receive any support from the workers of the factory themselves, which prevented machinebreaking from occurring.63 60 Clasen, Textilherstellung, pp. 15–89. 61 Blumberg, Textilindustrie, pp. 312; Elisabeth Janssen, ‘Zur Wohnsituation Aachener Textilarbeiter um 1830’, Zeitschrift des Aachener Geschichtsvereins 92 (1985): 209–217. 62 Blumberg, Textilindustrie, pp. 370–371. 63 Kurt Düwell, ‘Die Unruhen der Aachener Textilarbeiter 1830’, in Kurt Düwell and Wolfgang Köllmann (eds), Rheinland-Westfalen im Industriezeitalter, vol. 1: Von der Entstehung der Provinzen bis zur Reichsgründung (Wuppertal, 1983), pp. 114–129; 227

The Ashgate Companion to the History of Textile Workers The atmosphere in the revolutionary year of 1848 had a significant effect on the reorganization of the workforce. In Chemnitz, the waged weavers formed a union by implementing agreed wages with the putting-out merchants against the resistance of the guild. The substantially better organized workers in the calico-printing plants had already fought against the reduction of wages and the introduction of new machinery with strikes between 1819 and 1823.64 During the 1848/49 Revolution a labour association developed in Lennep, predominantly formed of textile workers. After having primarily engaged in social issues, it also participated in the political debate of the Revolution period. Following this lead, groups of the labour union of Wuppertal formed in several centres of the cloth industry in 1850. In 1863 in Wuppertal the Allgemeine Deutsche Arbeiterverein was founded by Ferdinand Lassalle, which by 1865 had 1,260 members and therefore was the largest branch organization in Germany. In the Bergish cloth industry in 1850 a 10-day strike occurred, in which approximately 2,000 workers participated. The goals of the strike were the implementation of wage increases, the introduction of a weekly wage instead of an hourly wage, the reduction of working hours, the provision of overtime payment and the improvement of working conditions. The strike failed particularly because a strike fund was lacking (which was felt sharply due to the tactical error of striking at all the large enterprises at once), because of the unfavourable situation relating to market conditions (the strike did not come at an inconvenient moment for the entrepreneurs) and also because of government intervention (through strike prohibition).65 It was not until the 1860s that the individual strikes and actions became more political in nature. In 1865 a strike against the imposition of fines for starting work late was supported by the Social Democrats (Allgemeiner Deutsche Arbeiterverein), which took up a collection in support of the strikers. In the elections to the North German Reichstag in 1867, the politicization of the workforce in the wool industry gained importance. In the region of Mark/Berg the candidate for the Allgemeine Deutscher Arbeiterverein of Lassalle received many votes. August Bebel was elected by the votes of the weavers in Glauchau-Meerane. The Social Democratic Labour Party, founded in 1869 (Bebel/Liebknecht), was in vogue especially in Saxony and Thuringia.66

Heinrich Volkmann, ‘Wirtschaftlicher Strukturwandel und sozialer Konflikt in der Frühindustrialisierung; eine Fallstudie zum Aachener Aufruhr von 1830’, Soziologie und Sozialgeschichte 16 (1972): 550–565. 64 Schaller, Einmal kommt die Zeit, pp. 87–97. 65 Dieter Dowe, ‘Legale Interessenvertretung und Streik; der Arbeitskampf der Tuchfabriken des Kreises Lennep (Bergisches Land) 1850’, in Klaus Tenfelde and Heinrich Volkmann (eds), Streik (München, 1981), pp. 31–51; Wolfgang Köllmann, ‘Industrielle Unterschichten im Bergischen Land und im Ruhrgebiet in der Früh- und Hochindustrialisierung’, in Hans Mommsen and Winfried Schulze (eds), Vom Elend der Handarbeit (Stuttgart, 1981), pp. 319–333. 66 Blumberg, Textilindustrie, pp. 373–374. 228

The German wool and cotton industry In Aachen, the influence of social democracy remained moderate due to the dominant role of the Catholic Church. The influence of the church on the beginnings of the labour organization was accordingly high. In 1873 and 1875, congresses of Christian workers from the Rhineland and Westphalia took place in Aachen, where a series of demands were put forward: a normal daily work schedule of 10 hours, extra payment for overtime, prohibition of Sunday work and work by married women, factory regulations and the participation of the workers in the trade tribunal. The level of organization in the Aachen cloth industry particularly increased following the introduction of the fast-running machine loom and the double-loom system, which were connected with increasing unemployment to the concern of the weavers. In 1887 the Catholic Weavers’ Association was founded, from which the Christian-social Textile Workers’ Association emerged in 1896, and the Central Association of Christian Textile Workers in 1900/1901.67

67 Thun, Industrie, p. 32; Wichterich, ‘Entwicklung’, pp. 176–177; Dahmen, Aachener Tuchgewerbe, pp. 143–146. 229

This page has been left blank intentionally

9 Great Britain: textile workers in the Lancashire cotton and Yorkshire wool industries Alan Fowler

This chapter compares the experiences of textile workers in cotton and wool in Britain from the mid-seventeenth century to the mid-twentieth century. The emphasis is on the contrast between the responses of labour in Lancashire and Yorkshire to industrialization and shows that despite the similarities of technology, workers in cotton and wool adopted different industrial strategies. The most significant of these contrasts was the development of trade unionism by nineteenth-century Lancashire cotton workers and the absence of effective trade unionism in Yorkshire wool until 1914. The chapter examines why, given their similar geographical and technological background, these responses of labour were so different.

Textiles in Lancashire and Yorkshire c.1700 Production of wool textiles before 1700 was a feature of many parts of England but was mainly concentrated in Yorkshire, East Anglia and the West Country. The period 1500–1700 witnessed a number of developments in textile production. Growth in production took place largely outside the towns, certainly outside those towns where guild restrictions existed. Government legislation, the Statute of Artificers 1563, influenced both wages and apprenticeship with local magistrates given a role in their determination. Legislation was also passed to control the quality of production of wool, partly explained by the fact that wool was Britain’s premier export industry throughout the period. So important was wool to Britain’s prosperity that the seat on which the Lord Chancellor sat in the House of Lords was  My thanks to Terry Wyke and Chris Wrigley for reading this chapter in draft.  D.T. Jenkins and K.G. Ponting, The British Wool Textile Industry 1770–1914 (London, 1982), pp. 1–8.

Map 9.1

Textile districts of Nothern England

Great Britain known as the Woolsack. Nor was the period static in terms of economic change. The development of new cloths, ‘the new draperies’, enabled Britain to compete more effectively on the continent of Europe. The industry was also boosted by the arrival of Protestant refugees from Catholic countries, especially France, who introduced new textile traditions to England. The general effect was to stimulate trade and encourage economic growth. The use of the term ‘wool industry’ is something of a misnomer. Effectively the industry was divided between woollen and worsted cloths, the division being dependent on the type of wool used. During the eighteenth century the wool industry continued to expand in Yorkshire, which developed as the country’s major centre of wool production. Significant as the expansion of woollen textiles was, there was also a rapid development of the worsted industry around Bradford. Prior to the eighteenth century worsted production had been concentrated in East Anglia but as the century progressed Yorkshire began to successfully compete with East Anglia in production of worsted cloths. Yorkshire is located in the north of England to the east of a line of hills, the Pennines, which runs down the centre of England. To the west lies Lancashire which was much less economically developed than Yorkshire up to 1700. The cotton industry, which was established in Lancashire during the eighteenth century, arose out of an existing textile tradition based on wool and linen. Cotton yarn and cloths developed in Lancashire after 1700 for a number of reasons. Demand for cotton goods had been encouraged by imports from India but the powerful wool interest had successfully lobbied for protection. This led Lancashire to develop fustian cloths (a mixture of linen and cotton) as an import substitute. Despite the relative economic backwardness of Lancashire, the county enjoyed a number of natural advantages, climate, proximity to Atlantic ports and later coal, and this combined with favourable economic circumstance explains the industry’s rapid development as demand increased in the second half of the eighteenth century. The development of textiles in both Lancashire and Yorkshire was closely associated with agricultural production. There were many marginal producers in agriculture on both sides of the Pennines who added to their subsistence by engaging in textiles. These agricultural producers increasingly came to rely on textiles for their major source of income. Before the Industrial Revolution all textile production was organized under the ‘domestic systems’ based in the home. There were significant differences in the  For a general account of the industry before the Industrial Revolution, see Herbert Heaton, The Yorkshire Woollen and Worsted Industries: From the Earliest Times up to the Industrial Revolution (Oxford, 1965) (first published 1920) and Alfred P. Wadsworth and Julia De Lacy Mann, The Cotton Trade of Industrial Lancashire 1600–1780 (Manchester, 1965) (first published 1931), pp. 3–10.  R.G. Wilson, ‘The Supremacy of the Yorkshire Cloth Industry in the Eighteenth Century’, in N.B. Harte and K.G. Ponting (eds), Textile History and Economic History (Manchester, 1973). 233

The Ashgate Companion to the History of Textile Workers organization of the cotton and wool domestic systems and, as important, between the two branches of the wool industry in Yorkshire. Economic historians from John Clapham and Herbert Heaton at the beginning of the twentieth century to Eric Sigsworth in the mid-twentieth century and Pat Hudson towards the end of the twentieth century have emphasized this difference. Woollen textiles and worsted differed in that they used different staples of wool: worsted the long staples, woollen textiles the shorter. Worsted was a finer cloth and particularly important in the development of the trade in ladies clothing. However, these technical differences do not explain the differences in economic organization between the two sectors. Woollen textile production under the domestic system in Yorkshire was essentially small scale and ‘in the hands of small independent clothiers’. The family owned the tools and materials that were worked by family labour.10 The clothier bought his own wool and also sold the finished product at the local cloth hall. Both the clothier and his family might also be involved in agrarian production. According to Heaton the success of the small clothier was only possible because of low capital expenditure.11 The model in East Anglia and the West Country was somewhat different with a ‘big’ clothier employing people outside his family, people who were clearly wage labourers. Although there was some replication of this model in the Yorkshire woollen industry, the small clothier was dominant. The model of the ‘big’ clothier employing wage labour was adopted in the newly developing worsted industry in Yorkshire after 1700. Both Heaton12 and Hudson13 share the view that this reflected the later development of worsted production in Yorkshire and the need to mobilize large capital in order to compete with the traditional producers of East Anglia. The clothier in worsted employed substantial numbers of wage workers compared with his counterpart in woollen textiles. The clothier purchased the raw material which was put out to workers in their homes. The clothier retained ownership of the raw material as well as the finished product and he then sold on the goods to the merchant. Worsted was ‘much more capitalistic’ than woollen textiles and the clothier therefore enjoyed higher levels of profit than his counterpart in woollen textiles.14 The probable explanation for the greater capital required in worsted rather than woollen textiles was the need of the former to import wool from outside the region and also the need to compete  J.H. Clapham, The Woollen and Worsted Industries (London, 1907).  Heaton, Yorkshire Woollen and Worsted Industries.  Eric Milton Sigsworth, Black Dyke Mills: A History … with introductory chapter on the development of the worsted industry in the nineteenth century (Liverpool, 1958).  Pat Hudson, The Genesis of Industrial Capital: A Study of the West Riding Wool Textile Industry c. 1750–1850 (Cambridge, 1986).  Heaton, Yorkshire Woollen and Worsted Industries, p. 293. 10 Hudson, Genesis of Industrial Capital, pp. 26–27. 11 Heaton, Yorkshire Woollen and Worsted Industries, pp. 293–301. 12 Ibid., pp. 296–297. 13 Hudson, Genesis of Industrial Capital, pp. 28–29. 14 Heaton, Yorkshire Woollen and Worsted Industries, pp. 248–321; Hudson, Genesis of Industrial Capital, pp. 25–52. 234

Great Britain with East Anglia where there was already a high concentration of capital in the industry.15 The growth of capital in the Yorkshire worsted industry also brought a response from labour as early trade unions developed in the eighteenth century reflecting the growth of wage labour within the domestic system.16 These nascent trade unions were to lead in the early nineteenth century to organized hostility to technological change. Historians of early developments in textiles in Lancashire have emphasized that the domestic system was dominated by small manufacturers during the sixteenth and early seventeenth centuries.17 However, as the Lancashire cotton industry emerged in the eighteenth century, so began the process of transformation from ‘peasant farmers into wage earning textile workers’.18 Wadsworth and Mann in their classic account of this development emphasize the growth of capitalist organization in the industry under the domestic system. Wadsworth and Mann’s account of change in Lancashire bears many similarities to that of Heaton’s explanation of Yorkshire and highlights the development of collective organization of wage labour. The early development of the cotton industry saw the emergence of trade unions among handloom weavers in the mid-eighteenth century drawing upon existing traditions of labour organization, firstly the tradition of apprenticeship and secondly that of the friendly society. These early attempts at organized trade unionism in cotton were essentially among handloom weavers, male workers, but were largely unsuccessful.19 What was to prove significant, though, was that the tradition of combination and collective organization was passed from hand to factory workers in Lancashire but no similar process took place in Yorkshire.20

Textiles during the Industrial Revolution The transformation of the Lancashire cotton industry from the domestic system to the factory system was one of the key moments in world history. The role of labour in this transformation has attracted less attention because of earlier economic historians’ concentration on technology and entrepreneurship. Past accounts of the industrialization of Lancashire emphasize the heroic role of entrepreneurs and inventors, focusing on the role of individuals such as Richard Arkwright. This is not to suggest those individuals were insignificant. The growth of the factory system in 15 Hudson, Genesis of Industrial Capital, p. 29. 16 Heaton, Yorkshire Woollen and Worsted Industries, pp. 312–321. 17 Norman Lowe, The Lancashire Textile Industry in the Sixteenth Century (Manchester, 1972) p. 20. 18 Wadsworth and Mann, Cotton Trade, p. 3. 19 Ibid., pp. 340–383. 20 Heaton, Yorkshire Woollen and Worsted Industries, p. 321. 235

The Ashgate Companion to the History of Textile Workers Lancashire clearly depended on a series of technological breakthroughs relating to carding and spinning. The three major innovations, Hargreaves’ jenny, Arkwright’s water frame and Crompton’s mule, did lay the basis for the factory system, first with water power and later with steam. It was the novelty of the factory system that impressed contemporaries as textile production moved from the home to the factory. Female and child labour, an important part of the domestic system, remained significant in the new factory system, though spinning, previously a female occupation in the home, became a key male occupation in the factory. The factory system transformed the Lancashire cotton industry, enabling it to expand at a rapid pace and to conquer the world market for cotton yarns and cloth helped by Britain’s growing imperial power. The role of labour was central to that transformation.21 The domestic system of production required around four spinners, usually women workers, to keep one weaver, usually a male worker, supplied. It was this bottleneck in production which led to developments in spinning machinery. The advent of the factory system created an entirely new labour force in the spinning sector of the Lancashire cotton industry in which the key workers were the male spinners, though women and children also entered the factories in large numbers. Initially the effect of these developments in spinning was to increase the demand for handloom weavers. Domestic workers in weaving only began to be replaced by power-loom weavers in mills as the pace of technological development quickened in the 1820s and 1830s, though the fall in their living standards preceded this, and was not complete until the 1840s.22 The factory system imposed new disciplines on the labour force. Steam-driven machinery removed workers’ control over their pace of work, was dangerous to operate and subjected workers to both a mechanical and a human discipline as hours worked and breaks from work became determined by the overlooker or the employer as well as the clock. Workers became known as ‘hands’. Factory hands were entirely dependent on their wage for subsistence in contrast to workers in the domestic system, some of whom might retain links with agriculture. However, there were continuities between domestic and factory system. Workers were paid by the piece; that is according to the amount of yarn or cloth produced. The price paid for each piece was established according to local wage lists, a system advantageous to employers as well as labour.23 Apprenticeship was not a significant feature of the industry but new skills such as mule spinning were created. Mule spinning was a technically skilled operation, 21 Mary B. Rose, ‘The Rise of the Cotton Industry in Lancashire to 1830’, in Mary B. Rose (ed.), The Lancashire Cotton Industry: A History Since 1700 (Preston, 1996), pp. 1–28. 22 D.A. Farnie, The English Cotton Industry and the World Market, 1815–1896 (Oxford, 1979) pp. 277–284; Geoffrey Timmins, The Last Shift (Manchester, 1993); Duncan Bythell, The Handloom Weavers: A Study in the English Cotton Industry during the Industrial Revolution (Cambridge, 1969) pp. 40–57, 96–119. 23 Michael Winstanley, ‘The Factory Workforce’, in Rose (ed.), The Lancashire Cotton Industry, pp. 121–153, at p. 136. 236

Great Britain though lacking a formal apprenticeship system. Mule spinners were part of a team assisted by ‘big’ piecers and ‘little’ piecers. The nineteenth-century term ‘team’ is perhaps open to question as the mule spinner employed the piecers on a system of subcontract and was also responsible for recruiting his assistants. The mule spinner did not have complete control over the process as employers could dismiss a spinner failing to attract suitable labour. Piecers provided the pool of labour from which spinners were recruited and were promoted according to ability and opportunity, the latter depending on the expansion of the industry. The system posed a risk to mule spinners, as piecers as a group, though essential to the production process, were a potential alternative source of labour which could be used to substitute for the mule spinner in any dispute with employers. Mule spinners were well-paid workers relative to other factory workers but nevertheless posed a constant threat to employers. Early in the nineteenth century the mule spinners developed trade unions and became the focal point of a series of industrial disputes. Partly as a response to the strength of the mule spinners, employers in the industry sought further technical change in order to undermine the spinners’ privileged position within the mill. The self-actor spinning machine was developed in the 1820s by Richard Roberts as a response to this problem with the specific aim of substituting female for male operatives as women were seen as more tractable as well as providing cheaper labour. This was never achieved. The self-actor replaced the mule between the 1830s and 1870s but male labour remained firmly in control of the spinning process and though initially workers using selfactors were known as minders rather than spinners, they were able to retain their elite status. To contemporaries they were known as ‘barefoot aristocrats’ descriptive of the fact that they worked without footwear but also of both their status and their earning power in the industry. The spinners became the first and most highly organized of the different occupations of workers in the cotton industry.24 Beginning in the 1820s there were a series of attempts to form a federal spinning trade union covering the whole of the cotton industry. Initially the aim was to raise wages to the level of those in Manchester, the highest paid area. Set against a background of legal hostility until the repeal of the combination laws in the 1820s, it was not until the 1870s that a permanent countywide organization was successfully established. The federal union flourished and gained a mass membership among Lancashire cotton spinners. These well-paid workers were able to pay membership fees which enabled permanent paid trade union officials to be appointed, friendly society benefits to be paid and a substantial strike fund to be accumulated.25 24 John Mason, ‘Cotton Spinning in the Industrial Revolution’ (pp. 1–13); ‘Mule Spinners Societies and Early Federations’ (pp. 14–35); ‘Spinners and Minders’ (pp. 36–58), in Alan Fowler and Terry Wyke (eds), The Barefoot Aristocrats: A History of the Amalgamated Association of Operative Cotton Spinners (Littleborough, 1987); W. Lazonick, ‘Industrial Relations and Technical Change: The Case of the Self Actor Mule’, Cambridge Journal of Economics 3 (1979): 231–262. 25 Mason, ‘Spinners and Minders’; Andrew Bullen, ‘The Founding of the Amalgamation’ (pp. 74–92) and ‘A Modern Spinners Union’ (pp. 93–114), in Fowler and Wyke (eds), 237

The Ashgate Companion to the History of Textile Workers As there had been a role reversal with the female spinner in domestic production replaced by the elite male spinner in factory-based mule spinning, so there was a parallel process in weaving. The ‘golden age’ of the handloom weaver 1780–1820 attracted labour to the industry, so much so that a surplus of labour was beginning to undermine the privileged position of the handloom weaver even before technological change in weaving. Nevertheless traditions of collective organization, sometimes associated with machine smashing, were established among handloom weavers. The adoption of the Lancashire power-loom in the 1830s and 1840s left the handloom weavers destitute. Crucially, though both male and female labour was recruited to work the power-looms, the majority of weavers were female. A federal union for the weavers was not achieved until the 1880s. This arose out of a major industrial dispute in north Lancashire – a dispute which persuaded employers as well as trade unions of the need for a period of peace in industrial relations as formal collective bargaining emerged at the same time, a pioneering development in British industrial relations.26 Yorkshire’s proximity to Lancashire partly explains the ease with which the technological advances in cotton were transferred across the Pennines and adapted for use in the wool industry. Yorkshire entrepreneurs grasped opportunities offered for the development of their industry and this coupled with the ready availability of coal consolidated Yorkshire’s position as the premier woollen textile and worsted producing area of England. The distinction between woollen and worsted in Yorkshire remained pertinent to the transformation from the domestic to the factory system. The preparatory systems involved different processes with the fulling mill a feature of the woollen sector and wool combing associated with worsted production. The greater mobilization of capital in worsted may explain why this sector adopted powered machinery sooner than woollen textiles in both spinning and weaving. The worsted industry adopted the Arkwright ring frame in spinning. The woollen industry introduced Crompton’s mule which lent itself to the spinning of the shorter fibres characteristic of that sector. The result was that worsted developed more rapidly in the nineteenth century in terms of the introduction of the factory system with woollen textiles following some 20 to 30 years behind.27 The differing impact of technological change in woollen textiles and worsted had a crucial impact on the labour force. The worsted industry’s adoption of the Arkwright technology saw the introduction of female labour, often young girls, in marked contrast to Lancashire where spinning was a male preserve.28 Power-looms were also lighter in worsted and as a consequence weaving was also dominated by women. Crompton’s mule was not adopted by the woollen sector until the Barefoot Aristocrats. 26 Andrew Bullen, The Lancashire Weavers’ Union (Rochdale, 1984) pp. 5–20. 27 David Hey, Yorkshire from AD 1000 (London, 1986), pp. 230–239; Hudson, The Genesis of Industrial Capital, pp. 25–52. 28 H.A. Clegg, A. Fox and A.F. Thompson, A History of British Trade Unions Since 1889, vol. 1 (Oxford, 1964) p. 34. 238

Great Britain 1830s and was quickly superseded by the self-actor following on the success of the technology in Lancashire. Nevertheless, male spinners were able to maintain their preserve as elite workers. Spinning trade unionism was slow to develop even after the adoption of mule spinning and this can partly be explained by the relatively small size of the spinning unit in woollen textiles compared to Lancashire and the lack of concentration of large numbers of workers in an enterprise.29 Hand-workers posed a greater challenge to their employers in Yorkshire than in Lancashire. Even before the advent of power-loom weaving in Yorkshire, industrial unrest had taken the form of Luddism, machine smashing, which was prevalent right up to the mid-1820s as a surplus of labour undermined wages. Collective organization in the wool industry was strongest among those groups of workers most susceptible to technological change such as wool-combers. The wool-combers, supported by the handloom weavers, made their major stand in the worsted industry in 1825 in Bradford. A four-month strike involving some 20,000 workers saw large sums of money raised nationally in support. The strike though was to resist wage cuts rather than the introduction of new technology.30 Its defeat spelt the end of effective trade unionism in West Yorkshire among hand-workers.31 Traditions of collective organization do not appear to have been passed from hand to factory workers. It was this factor that H.A. Turner identified as the main difference between Yorkshire and Lancashire trade union organization.32 Mass trade unionism among factory workers was not to be achieved in Yorkshire till 1914. Technological change and the structure of the industry in cotton, woollen textiles and worsted production were crucial in determining the effectiveness of collective organization of workers. In cotton Lancashire the elite male spinners established permanent mass trade unions by the 1870s followed a decade later by the organization of predominantly female power-loom weavers. In Yorkshire no similar pattern emerged in either worsted or woollen production. Female spinners were not able to establish effective trade union organization in worsted where there were relatively high concentrations of capital, nor were their male counterparts in woollen textiles who were much more dispersed. As the elite groups failed to organize so those lesser paid workers in weaving had no tradition of organization to follow.

29 Ibid. 30 Jack Reynolds, The Great Paternalist: Titus Salt and the Growth of Nineteenth Century Bradford (London, 1983), p. 30. 31 Theodore Koditschek, Class Formation and Urban Industrial Society: Bradford 1750–1850 (Cambridge, 1990), pp. 29–40, 351–365, 471–476; Karl Ittman, Work, Gender and Family in Victorian England (London, 1995), pp. 39–72. 32 H.A. Turner, Trade Union Growth, Structure and Policy: A Comparative Study of the Cotton Unions (London, 1962), pp. 173–179. 239

The Ashgate Companion to the History of Textile Workers

Mass unionism in textiles The 1851 census showed that the cotton industry employed over half a million workers, the majority – 90 per cent – being in Lancashire. The wool industry accounted for just under a quarter of a million workers of whom two-thirds were employed in West Yorkshire. The cotton industry was approximately double the size of the wool industry and by 1911 this ratio had increased as employment in cotton rose to over 600,000 operatives whereas the number of operatives in the wool industry, largely synonymous with West Yorkshire, stood at around 250,000, females forming just over 60 per cent of the labour force in both industries. A comparison of the figures for trade union membership in the two industries starkly demonstrates their differing experiences in collective organization. The Manchester meeting of the Trades Union Congress (TUC) in 1913 showed membership figures for the main cotton unions as weavers 182,000, card-room operatives 55,000 and spinners 23,000. This total trade union membership of 260,000 was augmented by approximately 25,000 piecers, whose trade union was not affiliated to the TUC. The General Union of Textile Workers, which covered all occupations in the wool industry, had a membership of 7,000. To this figure should be added a membership of around 7,000 for the Dyers’ union and 6,000 for a range of small craft societies.33 In cotton Lancashire around 50 per cent of the workforce was organized. In the wool industry in Yorkshire the figure stood at around 10 per cent. The contrast between the cotton and wool unions could not be clearer. Between 1889 and the First World War British labour was transformed by the growth of mass unionism. The experience of British textiles was to offer a contrast between the success of cotton operatives and the failure of wool workers to organize labour in their respective industries. The contrasting experience of cotton and wool surprised some contemporaries. Lancashire and Yorkshire retained many similarities in technology, economics and culture as well as geographical proximity. Many outsiders saw the towns and workforce of Lancashire and Yorkshire, on either sides of the Pennines, as similar, yet they responded differently to the growth of organized labour. The development of trade unionism in Lancashire cotton post-1880 drew on some of the lessons learnt from the experience of early trade unionism during the Industrial Revolution. For example, the need to develop trade unions that while based on both sectional and local interests together were centrally led with a leadership strong enough to override the membership’s wish to resist wage cuts when the market was unfavourable to such a policy. The most influential group in developing this policy was the spinners with their newly established Amalgamated Association of Operative Cotton Spinners. Following the triumph of the self-actor, the spinners developed this federal organization based on district associations. The spinners’ assistants, the piecers, were encouraged to form their own, subservient organization, in order to avoid them falling on the spinners’ funds during any dispute. The Amalgamation rules made strikes difficult without the support of 33 Trade Union Congress, Annual Report (London, 1913), pp. 9–10, 24–26. 240

Great Britain the leadership, thus centralizing control of the union among a small group of full-time officials. These officials were chosen by examination for their skill in wage calculation. The ability to calculate wages was particularly valued because the operatives were paid by a wage list that was highly complex and initially varied from town to town. The Oldham and Bolton wage lists were the most significant in spinning though they were based on different principles. The Bolton list maximized productivity gains to the spinner while employers paying the Oldham list were advantaged by increased production. The spinners’ union adopted the model of the trade union official as the wage calculator from an early weavers’ union. This model enabled the spinners to develop a strong trade union leadership before 1900 with James Mawdsley (General Secretary 1878–1902) and Thomas Ashton (President 1870–1911) recognized not just in Lancashire but nationally as outstanding figures in British labour. They also understood the need for a newspaper to report the growing labour movement and the two of them assisted in the founding of the Cotton Factory Times.34 This was followed by the founding of the United Textile Factory Workers’ Association to act as a political pressure group. Also critical to the spinners’ success was their policy of high membership subscriptions thus enabling the Amalgamation to develop a strong financial fund and with it an ability to survive lengthy disputes. The cotton weavers and card-room operatives despite lacking members with such high wages as the spinners adopted similar structures and policies. Federal trade union organization combined with strong central leadership of full-time officials, chosen for their ability in wage calculation, gave the weavers, despite their mass membership, very much the appearance of a craft or ‘old’ union. The Cardroom Amalgamation, following the advice of the spinners, similarly adopted this model. Thus to the outsider the cotton trade unions looked remarkable similar though this hid real occupational difference. The labour force of Lancashire cotton remained divided by both occupation and gender. The industry recruited adult male workers (mule spinners, strippers and grinders), female operatives (weavers, jack frame tenters) and juvenile and child labour (piecers and tenters). The half-time system by which successive compulsory education acts provided for children to both work and attend school enabled the employment of child labour right up until 1921. The earnings of different members of the labour force varied considerably with mule spinners representing the aristocracy of labour. Other operatives’ earnings ranged from the average industrial wage to wages bordering on contemporary definitions of poverty. Mule spinners earned over forty shillings (£2.00) a week in 1906 compared with a weaver’s wage of just over twenty shillings (£1.00). The prosperity of the labour force was not based on the adult male wage, but rather the collective earnings of the whole family, father, mother and children. This was an unusual feature of Lancashire which contrasted with the other main staples of British industry where prosperity was primarily determined by the adult 34 Eddie Cass, Alan Fowler and Terry Wyke ‘The Remarkable Rise and Long Decline of the Cotton Factory Times’, Media History 4 (2) (1998): 141–159. 241

The Ashgate Companion to the History of Textile Workers male wage. Prosperity in Lancashire matched that of any other working-class community in Britain. The proof of this was the success of Blackpool, the seaside resort of Lancashire cotton operatives and the only seaside town in the world based on working-class prosperity. 35 Female operatives in weaving were particularly highly paid, being the best paid women workers in Britain. Unusually for British industry there was no clear sexual division of labour in cotton weaving as both men and women performed the same work and ostensibly received the same wage based on the wage list. However, male weavers could receive more money in the weekly wage packet because of the types of cloth allocated to them. The existence at least theoretically of equal pay in weaving was novel. Lancashire weaving was also different in that a substantial element of the female labour force was married. The proportion of married women working varied from town to town and was higher in north than south Lancashire. Married women tended to continue in the industry beyond childbirth. Many women only left the industry when their own children became adult wage-earners. This practice flew in the face of Victorian ideology in which the woman’s place was in the home though for most female weavers the consequence was the double shift of factory work and housework. Contemporary critics suggested that high infant mortality in Lancashire was the consequence of this pattern of behaviour but the evidence is confusing. Nor were contemporary concerns quite as robust when it came to child labour though the practice did attract considerable criticism from trade unionists outside Lancashire.36 It is not always possible to state how far trade unions were responsible for raising living standards in a particular industry. The Lancashire cotton industry’s main statistician, G.H. Wood, suggested the rising wages in the industry between 1850 and 1914 reflected the impact of rising productivity and growth of wage lists as well as trade union pressure.37 Late Victorian and Edwardian cotton operatives had good reason therefore to believe that their trade unions had played a prominent part in their prosperity. Trade unions were not the only self-help voluntary organizations. Friendly societies and cooperative retail organizations were strong in Lancashire. Employers responded to the development of trade unionism by establishing their own organization. Strong employer and trade union organizations led to the growth of collective bargaining in the industry. The Brooklands Agreement of 1893 was the best-known example of collaboration between the two sides of industry though weaving had developed collective bargaining in the 1880s. Brooklands laid down a binding procedure to be followed in the event of any industrial dispute at a factory, local or county level. Collective bargaining combined with wage 35 Alan Fowler, Cotton Operatives and Work, 1900–1950: A Social History of Lancashire Cotton Operatives in the Twentieth Century (Aldershot, 2003), pp. 12–36, 109–12; John K. Walton, Blackpool (Edinburgh, 1998), pp. 46–106. 36 Fowler, Cotton Operatives and Work, pp. 43–65. 37 G.H. Wood, A History of Wages in the Cotton Trade During the Past Hundred Years (Manchester, 1910) pp. 139–146. 242

Great Britain lists ensured a degree of organization of labour relations that was much admired in British industry. The result was not only a more prosperous labour force but one which also saw itself as a partner in the industry. Symbolic of this was the relationship between James Mawdsley, the spinners’ leader, and [Sir] Charles Macara, the employers’ outstanding figure, the two architects of the Brooklands Agreement. The close relationship between capital and labour continued into the Edwardian period and is illustrated by the invitation to David Shackleton, the weavers’ leader, to join a delegation of leading cotton employers to travel the USA to view advances in technology in 1902. 38 The most significant point to make about workers in the wool industry compared to their Lancashire cotton counterparts is the different response to mass unionism. This difference has attracted comment from a number of historians as they have tried to make sense of the distinction between Lancashire and Yorkshire.39 The industry was associated with one of the well-known strikes during the period of mass unionism. The major strike in the Yorkshire wool industry was the Manningham mill strike of 1890–1891 at Lister’s mill in Bradford.40 The strike did not spread beyond the Manningham mill workforce yet has been identified as key moment in the history of British labour. The strike was an attempt to resist the usual employers’ strategy of wage cuts as a result of a loss of trade in the American market because of the McKinley Tariff. The union called a strike despite the fact that the Manningham mill labour force was largely unorganized. Initially the labour force was united in attempts to resist the wage cuts, but without strong union finances the inevitable happened when the spinners returned to work after a bitter struggle lasting 19 weeks. Samuel Lister was a leading employer and citizen in Bradford and the bitterness engendered by the strike led to a wider conflict particularly over the issue of free speech in the town. The politicizing of the strike was to have important consequences. The strike was closely associated with the growth of the Independent Labour Party founded in Bradford in 1893 and an important milestone on the road to the founding of a successful Labour Party. Yorkshire textile operatives played a critical role in the growth of labour politically but not industrially.41 Overall the Yorkshire wool industry’s performance between 1875 and 1914 was far more sluggish than Lancashire cotton. The ‘great depression’ of 1873–1896 was a period of falling prices and profits for British industry. A compounding factor for both woollen textiles and worsted production was the challenges faced in traditional export markets during this period. The wool industry felt the impact 38 Andrew Bullen, ‘The Making of Brooklands’, in Fowler and Wyke (eds), Barefoot Aristocrats, pp. 93–114; Ross M. Martin, The Lancashire Giant: David Shackleton Labour Leader and Civil Servant (Liverpool, 2000), p. 25. 39 Clapham, Woollen and Worsted Industries, pp. 204–213; Turner, Trade Union Growth, pp. 173–180; Clegg, Fox and Thompson, History of British Trade Unions Since 1889, pp. 33–34; Heaton, Yorkshire Woollen and Worsted Industries, pp. 320–321. 40 Cyril Pearce, The Manningham Mills Strike, Bradford December 1890–April 1891 (Hull, 1975). 41 E.P. Thompson, ‘Homage to Tom Maguire’, in Asa Briggs and John Saville (eds), Essays in Labour History (London, 1967), pp. 276–316. 243

The Ashgate Companion to the History of Textile Workers as tariff barriers were erected against its manufactures especially in the American market. There was also growing competition in the home market with an increase in imports from the Continent.42 There was some recovery in woollen textiles after 1900 but in general this economic background created unfavourable circumstances for the growth of permanent mass trade unions. Sir John Clapham offered the most detailed contemporary discussion of the failure of wool trade unionism before 1914 in his account of the industry written in 1907. He emphasized the ‘highly scattered’ nature of the labour force and the highly sectional nature of the trade. The cotton industry was also highly sectional but there was no equivalent of the distinction between the woollen and worsted sectors. Clapham also identified paternalism, the lack of a woollen spinners’ union comparable with the Lancashire spinners, the failure to organize women workers and the lack of uniform wage rates as barriers to the development of trade union organization.43 A number of small, male-dominated craft occupations did establish permanent organizations in wool, particularly the dyers. These were never to provide a focus for the spread of trade unionism throughout the industry and into other occupations. The adoption of different technologies for spinning by woollen and worsted manufacturers militated against the establishment of a spinners’ union. The Arkwright water frame commonly in use in the worsted sector utilized women’s labour. The small, male-dominated craft trade unions were overtly hostile to married women working outside the home, an unhelpful perspective in an industry where the overwhelming majority of workers were women and a substantial minority were married. The mule spinners formed a smaller proportion of the labour force than in Lancashire and subsequently had less influence or opportunity to provide a lead in the establishment of trade unionism. Without a strong spinners’ union with whom to ally, weavers also failed to organize effectively. Clapham’s consideration of employer paternalism as part of the explanation for the failure of trade unionism to develop in the wool industry has been challenged by subsequent historians. There were numerous small enterprises where workers would be in daily contact with their employer but there are also examples of large concentrations of capital, for example Lister’s mill at Manningham. Jowett argues that paternalism had peaked and was in decline by 1900.44 Wage lists were a crucial part of the organization of the cotton industry but appear to have been far less influential in Yorkshire. Workers were concerned to ensure that they were receiving the correct piece-rate for the yarns and cloths they produced. The complicated lists meant that most operatives were unable to 42 Jenkins and Ponting, British Wool Textile Industry 1770–1914, pp. 267–295; E. Sigsworth and J. Blackman, ‘The Woollen and Worsted Industries’, in D. Aldcroft (ed.), The Development of British Industry and Foreign Competition 1875–1914 (London, 1968), pp. 128–157. 43 Clapham, Woollen and Worsted Industries, pp. 174–220. 44 Tony Jowett, ‘The Retardation of Trade Unionism in the Yorkshire Worsted Textile Industry’, in J.A. Jowett and A.J. McIvor (eds), Employers and Labour in the English Textile Industries 1850–1939 (London, 1988), pp. 84–106, at pp. 96–97. 244

Great Britain calculate their own wages. Trade union officials in Lancashire spent considerable time and energy ensuring that members were paid the correct list price. This must have been a strong inducement to keep up the trade union membership dues. No such tradition existed in Yorkshire and even those lists in use often legitimized paying women less than men – not a situation conducive to worker solidarity.45 The result was that wage levels remained lower in Yorkshire than in Lancashire for both men and women. The main unions in cotton spinning and weaving were able to encourage the development of trade unionism among preparatory workers but a very different situation obtained in Yorkshire. As the industry turned to the import of raw wool, especially from the colonies, which was not required to be sorted to the same degree as home supplies, workers in the preparatory process became less skilled with a corresponding disinclination to organize collectively. The result was that there was not a group of skilled workers on whom to build the union. New unionism did influence Yorkshire in the 1890s and the General Union of Textile Workers, founded in the 1880s, was a trade union led by socialists with an ambition to organize all textile workers in Yorkshire. This union grew out of a strike by weavers in Huddersfield in the 1880s and was heavily involved in the disputes of the 1890s including the Bradford Manningham mill strike. Although the general union was essentially based on workers in weaving, largely in woollen textiles, the leadership was ambitious to extend trade unionism to the mass of workers in the industry. The most able of the Yorkshire leaders was Ben Turner who was closely involved with the launching of the Yorkshire Factory Times, an attempt to imitate the success of the Cotton Factory Times. Despite this advantage Yorkshire operatives were never able to repeat the success of Lancashire in creating permanent mass unions, at least not until 1914.46

Crisis and depression The outbreak of war in 1914 disrupted the pattern of international trade for both industries. The cotton industry was entirely dependent on shipments of raw material from either America or Egypt; the wool industry relied on supplies from Australia. Disruption of these supplies potentially turned the industries into turmoil. As raw material could not be imported, so also the industries faced problems in exporting finished goods to their traditional markets. 45 Joanna Bournat, ‘Lost Leaders: Women, Trade Unionism and the Case of the General Union of Textile Workers, 1875–1914’, in Angela John (ed.), Unequal Opportunities: Women’s Employment in England 1800–1918 (Oxford, 1986), pp. 207–234. 46 Jowett, ‘Retardation of Trade Unionism in the Yorkshire Worsted Textile Industry’; J.A. Jowett and K. Laybourn, ‘The Wool Textile Dispute of 1925’, The Journal of Local Studies 2 (1) (1982): 10–27; Ben Turner, About Myself 1863–1930 (London, 1930), pp. 86–124, 130–138. 245

The Ashgate Companion to the History of Textile Workers Lancashire’s premier place in the world textile markets had been dependent on free trade and a relatively cheap skilled labour force. These advantages did not survive the war as free trade was replaced by protection in many markets and the cheaper labour of Japan, though not India, undermined Lancashire prices. Cotton manufacturers remained committed to the traditional technology of the self-actor and the Lancashire loom while the newly developing industries of Asia and South America invested in modern technology, ring spindles and automatic looms.47 The wool industry had faced import tariffs in the American market before 1914 so the erection of new barriers as a result of the war had less of an impact in Yorkshire which had already been forced increasingly to rely on the home market. Neither was Yorkshire confronted with competition by the cheap labour of Asia as Yorkshire’s competitors were located on the European mainland. The disruption in markets, and in the supply of raw cotton, on the outbreak of war in 1914 led to an initial period of unemployment for many Lancashire cotton operatives followed by a period of sustained short-time working culminating in a crisis in 1917 which brought government intervention. Government, employers and trade unions cooperated to bring some order to the industry through a system which fairly distributed the amount of work available among the operatives by the use of a rota while providing unemployment pay for those out of work. However, government intervention took place in circumstances in which Lancashire cotton was no longer a national priority and its interests could be subordinated to those of India, the main market. The demand for the products of the wool industry remained stable during the war as wool was more central to the war effort than cotton Concerned with ensuring that supplies were maintained, rather than keeping order in the industry, state intervention led to tripartite collaboration. Both a Cotton Control Board and a Wool Control Board were established. Membership of the boards was extended to include trade union representation and state sanctioned support for collective agreements had important consequences for the development of trade unionism in the wool industry. Crucially, collective bargaining in the wool industry ensured that wages kept in line with the rise in the cost of living and trade union membership increased fivefold during the war. The cotton industry offered a contrast. War tends to create a tight labour market, advantageous to organized labour, and the cotton unions were unusual between 1914 and 1918 in not sharing these advantages to the extent of many other trade unions as real wages probably fell in contrast to the situation in the wool industry.48

47 Marguerite Dupree, ‘Foreign Competition and the Interwar Period’, in Rose (ed.), The Lancashire Cotton Industry, pp. 265–295. 48 Hugh Armstrong Clegg, A History of British Trade Unions Since 1889, vol. 2 (Oxford, 1985), pp 165–166; Chris Wrigley, ‘The First World War and State Intervention in Industrial Relations, 1914–18’, in Chris Wrigley (ed.), A History of British Industrial Relations 1914– 1939, vol. 2 (Brighton, 1987), p. 23–70, at pp. 43–44; Alan Fowler, ‘Impact of the First World War on the Lancashire Cotton Industry and its Workers’, in Chris Wrigley (ed.), The First World War and the International Economy (Cheltenham, 2000), pp. 76–98. 246

Great Britain The depression of the interwar years is associated by many economic historians more closely with the 1930s than the 1920s. However, in Britain the economy went into depression in 1921 after a very short post-war boom. The wool industry overall did rather better through the interwar slump than the cotton industry. Workers in both Lancashire and Yorkshire were subject to wage cuts and unemployment but the experience of organized labour in the two industries was different. The cotton industry experienced a return to normal collective bargaining at the end of the war heralding a return to the relationship established between capital and labour in the pre-war period. However, expectations of industrial harmony arising out of wartime collaboration were soon dashed. Disputes in 1918 and 1919 arose out of attempts to restore real wages during the post-war boom as well as to reduce the hours of labour. The cotton unions were largely successful. The immediate post-war period witnessed massive speculation in the cotton industry. A frenzy of recapitalization left the industry with significant debt when the post-war boom collapsed. As the depression deepened a further series of industrial disputes in 1921, 1929, 1931 and 1932 was the operatives’ response to employer attempts to reduce wages and alter working practices. Reduction of money wages (real wages rose overall) did not restore prosperity to the industry in the 1920s, leaving some sectors in spinning, where speculation had been greatest, in severe difficulties. The world depression of 1929–1931 resulted in the whole industry being thrown into crisis as employers demanded wage reductions, a strategy bitterly resisted by the trade unions. An alternative employer strategy was to attempt to restore competitiveness through an alteration of work practices. Traditionally, weavers on the Lancashire loom had operated four machines. Initially employers attempted to raise the number of looms each weaver worked to eight. Lack of success reduced the demand to the operation of six looms. The attempt to increase productivity of each individual worker was to have been accompanied by new wage lists which lowered the piecerate for individual cloths. The weekly wage therefore was to be upheld but surplus labour discarded. There was little or no attempt to retool the industry with the introduction of automatic looms. Productivity gains were to be made on existing capital goods which had long ago been bought and paid for. The resistance of the weaving unions created a crisis in industrial relations in the industry. It was only resolved by state intervention. The Cotton Manufacturing Industry Act of 1934 ensured that the wage list agreed between employers and unions was legally enforceable, a demand that the weavers had been making over a hundred years earlier.49 The wage agreement of 1935 reduced the cost advantage of ‘more looms’ by further reducing weavers’ wages on four looms but raising wages on six looms.

49 Alan Fowler, ‘Lancashire Cotton Trade Unionism in the Inter-war Years’, in J.A. Jowett and A.J. McIvor (eds), Employers and Labour in the English Textile Industries 1850–1939 (London, 1988), pp. 107–126; Fowler, Cotton Operatives and Work, pp. 74–107; Sue Bowden and David Higgins, ‘Productivity on the Cheap? The “More Looms” Experiment’ and the Lancashire Weaving Industry’, Business History 41 (3) (1999): 21–41. 247

The Ashgate Companion to the History of Textile Workers The result was that the ‘more looms’ initiatives was stifled. The state was also forced to intervene in the organization of the spinning sector of the cotton industry. The Cotton Spinning Act of 1936 allowed for compensation to be paid to employers scrapping spindles. No similar compensation payments were paid to operatives thrown out of work as a consequence of this reduction in capacity. The cotton trade unions used their strength and density of organization to challenge employer attempts to attack wages and conditions in the decades following the First World War and only turned to the state as a last resort. In Yorkshire, where the unions had always been weak, labour continued to look to the state throughout the period to preserve collective bargaining introduced during the war with government assistance. The Whitley Report produced in 1917 established what became known as Whitley Councils in a number of British industries where trade unionism was weak and collective bargaining largely absent. Both trade unions and employers in the wool industry were keen to continue the wartime collaboration into post-war society and the Whitley-inspired Northern Joint Industrial Council for Industry became the vehicle for collective bargaining. The General Union of Textile Workers was the main beneficiary of the creation of the Whitley Council in Yorkshire becoming, by 1920, not only a mass union but a key player in the industry’s future. The Council gained acceptance by workers as it promoted agreements for wages to follow the movement in the cost of living, leading to a series of wage advances in the post-war boom years. The initial success of the Northern Joint Industrial Council faded as the problem became managing industrial relations in a depressed industry. Wages were reduced in 1921 reflecting the fall in the cost of living with the onset of post-war depression. The General Union of Textile Workers was not strong enough to mount a resistance similar to that in cotton, being reliant on the Whitley Council to decide the level of wages in the wool industry. Collective bargaining through the Whitley Council failed to prevent the outbreak of industrial strife in 1925. Despite relative economic stability wool employers sought a 10 per cent reduction in wages. The trade union response was to seek a wage advance and the industry headed for conflict. The wool employers expected division within the trade union movement, especially between the general union and the small craft unions. However, the unions remained solid and government was forced to intervene with the establishment of a Court of Enquiry. It took the middle ground recommending neither an advance nor a reduction. While the unions saw this as a victory in having fought off wage cuts, the employers, convinced they could have won wage reductions if the government had not intervened, began to lose faith in collective bargaining. The employers launched another offensive in 1930 with a demand for further wage reductions arguing that markets could not be restored unless a more competitive price for labour was established. The unions appeared to accept the argument basing opposition on the extent of the cut rather than the principle, especially after a further government inquiry had found against them. This resulted

248

Great Britain in wage cuts for the operatives but also, and significantly, the collapse of collective bargaining in wool.50 Collective bargaining was restored to the industry in 1936 as economic conditions improved. This was only at the behest of government after an inquiry which recommended the restoration of a Joint Industrial Council as well as a wage advance. The restoration of collective bargaining coincided with a further amalgamation among wool trade unions. The National Union of Textile Workers (formerly the General Union) amalgamated with the main finishing workers union to create one major union for the industry.51 Recent writing has drawn attention to the superior performance of the wool industry in comparison with cotton during the interwar period. Bowden and Higgins recognize the wool industry did not have to pay for the consequences of the post-war capitalization the cotton industry suffered. Nevertheless they suggest the success of wool was partly due to the different labour policies of employers in the textile industries, in particular the wage flexibility which came with the end of collective bargaining in the 1930s. However, much of the comparative success of wool may have been due to the greater significance of the home market for the wool industry than for the cotton industry. The wool industry had adjusted to meet the loss of export markets before 1900 but the cotton industry had to face the loss of export markets as well as deal with the consequences of the slump.52 Trade unions in the cotton and wool industries shared similar experiences of employer offensives on wages and conditions in the interwar years. Cotton workers faced an added exposure to displacement as a result of industrial contraction when spindles were taken out of production and traditional work practices in weaving were attacked. The state was forced to intervene in both industries but the impetus for the intervention and its nature was different. The strength of mass unionism in Lancashire resulted in more serious disputes and it could be argued forced government to act to legalize wage lists in weaving and provide for orderly contraction in spinning. In contrast the weakness of the trade unions in wool resulted in the establishment of a Whitley council after the First World War to arbitrate between capital and labour over wages. Employers withdrew from the Council in the early 1930s but were forced back by Government in 1936.

50 Chris Wrigley, Cosy Co-operation under Strain: Industrial Relations in the Yorkshire Woollen Industry 1919–1930 (York, 1987); Jowett and Laybourn, ‘The Wool Textile Dispute of 1925’. 51 Hugh Armstrong Clegg, History of British Trade Unions since 1889, vol. 3: 1934–51 (Oxford, 1994), pp. 34–36. 52 Sue Bowden and David Higgins, ‘Much Ado About Nothing? Regional Business Networks and the Performance of the Cotton and Woollen Textile Industries, c 1919– 1939’, in John F. Wilson and Andrew Popp (eds), Industrial Clusters and Regional Business Networks in England 1750–1970 (Aldershot, 2003), pp. 90–111; Sue Bowden and David Higgins, ‘Quiet Successes and Loud Failures: The UK Textile Industries in the Inter-war Years’, Journal of Industrial History 3 (1) (2000): 91–111. 249

The Ashgate Companion to the History of Textile Workers

Post-war decline Textiles were a dominant sector of the British economy throughout the Industrial Revolution. Wool textiles dominated British industry before the nineteenth century and from the mid-eighteenth century Lancashire cotton emerged as Britain’s major industry and exporter, a position the industry maintained until the early twentieth century. This position of domination ceased in the second half of the twentieth century as textiles in Britain went into long-term economic decline. At the outbreak of the Second World War orderly contraction of the cotton industry began as a result of government initiatives to organize the Home Front. This was to enable movement of labour into the armed forces and munitions. A much reduced industry survived during the war staffed by married women and an ageing male labour force. This planned contraction ensured that those who stayed in the industry found wages and conditions improving.53 The modernization of industry was part of the agenda of the reforming post-war Labour government. A government working party looking at cotton in 1946 encouraged the industry to invest in the new technology of ring spinning and automatic looms. Wartime enquiries had suggested that this strategy might enable Lancashire to reduce the productivity gap with the USA. The report recommended the continuation of the wartime Cotton Board enabling the state to continue to play a role in the cotton industry while the industry itself still remained in private hands.54 The post-war demand for textiles created a seller’s market, especially with the absence of Japan from world markets till the 1950s. Profits and employment improved significantly in Lancashire cotton and there was even a major post-war shortage of labour. Despite these favourable circumstances and government encouragement there was little modernization of the industry as employers’ investment in new machinery was modest. The result was that by the early 1950s when international competition was restored textile Lancashire was in no better position to compete than in 1939.55 The industry went into terminal decline. The Cotton Industry Act 1959, enacted under a Conservative government, provided state subsidy to employers to scrap outdated looms and spindles and invest in ring spindles and automatic looms, machines long employed by competitors abroad. Many employers took advantage of the subsidy to destroy outdated machinery. Few, however, invested in new machinery. No state subsidy was available for operatives displaced by this policy but government made it clear that the industry itself would have to provide a redundancy programme for workers. Labour transferred out of Lancashire cotton in the 1950s and 1960s and the main role of trade unions was managing the decline. Lancashire cotton unions came to 53 John Singleton, ‘The Decline of the British Cotton Industry Since 1940’, in Rose (ed.), The Lancashire Cotton Industry, pp. 296–334, at pp. 299–307. 54 Board of Trade, Working Party Reports: Cotton (London, 1946). 55 Fowler, Cotton Operatives and Work, pp. 185–203; Singleton, ‘Decline of the British Cotton Industry since 1940’, pp. 303–310. 250

Great Britain share the employers’ view that protection offered the industry the best hope of survival. The industry’s decline did not produce conflict in industrial relations rather the industry developed a reputation for industrial harmony with both sides criticizing the government for not providing a more stable environment, which they were convinced meant protection.56 Government was prepared to be active in the reorganization of the cotton industry but, indicative of the dramatic fall in the importance of the industry to the British economy, the once powerful cotton lobby could not force other concessions. Cotton employers believed that the industry’s future could only be safeguarded by protection of the home market, a dramatic contrast with the industry’s traditional support for free trade. Neither the Conservative governments of the 1950s nor the Labour governments of the 1960s were prepared to grant protection. Protection was less of an issue for the wool industry which relied heavily on the home market. A government working party in 1947 resulted in little government action.57 While this working party noted the wool industry’s dependence on outdated machinery and recommended modernization, no further government action ensued. There was little modernization and little restructuring in the industry during the 1950s.58 Structural change in the cotton industry was achieved but not as a result of action by government, the industry or trade unions. During the 1960s the myriad manufacturers, large and small, disappeared as Courtaulds and ICI bought up large sections. But even these industrial giants failed to revive the industry’s flagging fortunes effectively.59 Major social change in working practices occurred in the textile industries in the 1960s with the introduction of the second and third shift, both in those sectors where there had been new investment in machinery and in mills and weaving sheds working the traditional machinery. A body of protective legislation had developed through the nineteenth century which prevented women from working at nights. As female labour was crucial to textile production, employers had been prevented from running a 24-hour factory. Bonus payments for working a second or third shift failed to attract male labour as a booming economy offered opportunities for workers outside textiles. The result was a migration of labour from South East Asia to the mills of Yorkshire and Lancashire attracted by the opportunities to work the third shift. The world market in textiles and the production of textiles developed inexorably during the second half of the twentieth century. In the long run no amount of restructuring, new investment or recruitment of a cheaper labour force could have enabled the British textile industries to have survived. Organized labour in the textile industries ceased to be significant in the British labour movement. The once mighty cotton unions disbanded to become part of the General and Municipal 56 Turner, Trade Union Growth, pp. 331–335. 57 Board of Trade, Working Party Reports: Wool (London, 1947). 58 G.C. Allen, British Industries and their Organisation (London, 1970), pp. 280–283. 59 Singleton, ‘Decline of the British Cotton Industry Since 1940’, pp. 310–315. 251

The Ashgate Companion to the History of Textile Workers Workers’ Union and the Transport and General Workers’ Union became the home of wool workers.

Conclusions The central experience of labour in British textiles in the period under discussion was that of industrialization. The Industrial Revolution transformed textile workers in Lancashire and Yorkshire from domestic workers to factory operatives (though it should be emphasized that wage labour was a feature of pre-industrial textile production). Industrialization had contrasting impacts on trade unions or combinations in wool and cotton. Trade union ‘traditions’ were passed from hand to factory workers as the cotton industry developed. The nineteenth century saw a series of attempts to create permanent trade unions, finally successful in the last quarter of the nineteenth century. Male cotton spinners were the first group of workers to demonstrate their industrial strength which provided a model for operatives in other sectors of the cotton industry. The experience of wool workers offers a marked contrast. Trade unionism came to an end in Yorkshire after workers had unsuccessfully opposed new machinery. From the 1820s to 1914 the industry was characterized largely by the absence of trade unionism. It was only with the impact of the First World War that permanent mass unionism emerged among workers in wool. Compounding factors militating against the establishment of permanent trade unions in wool included the division of the industry into worsted and woollen textile production and the adoption of differing technology, and the drawing in of female as well as male labour to spinning. State intervention in collective bargaining was a key factor in both cotton and wool during the twentieth century. The wool industry trade unions benefited from attempts by government to bring order to industrial relations and force the employers to collective bargaining during the First World War. The state continued to play a part in the industrial relations of the wool industry during the interwar years even though the employers withdrew from collective bargaining in the early 1930s when the unions resisted their traditional solution to trade depression – wage cuts. In fact the wool unions looked to government to force their belligerent employers to the negotiating table. The more powerful cotton unions were able to put up a greater resistance to employer demands for wage reductions during the slump even though the economic conditions of the interwar years negated much of their action. However, the weaving unions’ determination to ensure that their members did not suffer from new work practices coupled with opposition to wage cuts forced the state to intervene. The contrast between state intervention in the industrial relations of the wool and cotton industries is that in the former it was the result of trade union weakness while in the latter it was the result of trade union strength.

252

10 The long globalization and textile producers in India Tirthankar Roy

At two ends of the span of time being considered in this study, the seventeenth and eighteenth centuries on the one hand and the late twentieth on the other, India remained one of the world’s leading producers and exporters of textiles. This time span was punctuated by several major shifts. The transition from early modern trade to nineteenth-century industrialization, the transition from free trade to autarky in the mid-twentieth century and the return to globalization at the end of the millennium led to drastic and far-reaching changes in the textile industry. And yet, some fundamental elements proved remarkably durable. An earlier historiography stressed these disruptions relatively more. The concept of a nineteenth-century ‘de-industrialization’, for example, seems to drive a wedge between the eighteenth-century ‘proto-industrialization’ on the one hand, and the later growth of a cotton mill industry on the other. In effect, the past and the present of the textile industry were made discontinuous by the assumption that the artisan tradition had effectively ended in the region sometime about 1820, and what followed was something qualitatively different and unrelated. Historians of large-scale and modern industry and those studying the artisan tradition saw themselves as belonging to two different specializations, which rarely found common ground between them. A similar intellectual rupture occurred with 1947, when British colonial rule ended in the region, giving rise to a dirigiste development policy regime. The new regime wanted to reshape the textile industry radically. Policy analysts of post-independence India, who thought they were in charge of the new economy, rarely looked back, except to note how different their world was from colonialism. Forty years later, the interventionist project was in retreat, partly because intervention had failed to control forces of change that had long been inherent in the structure of the textile industry. One of the potential outcomes of a long-range history is that seemingly disruptive episodes begin to look a little less catastrophic and rather more like ebbs and flows on an essentially firm bed. Relationships and continuity become more visible. Recent tendencies in Indian textile history, especially what appears to be increasingly frequent dialogues between scholars studying the eighteenth century

Map 10.1 India 1939, including locations cited in the text

The long globalization and textile producers in India and those studying the twentieth century, suggest perhaps a desire to look for continuity across time and space. The basis for such a rethinking can be the belief that, in the long run, the textile producer could react to disruptions flexibly enough. The sources of such flexibility were various, but at least one principal source was the textile producers themselves, who possessed superior craft skills, knowledge of tastes and preferences, technology that could adapt itself to meet diverse needs, were willing to build cooperative communities, were willing to migrate in search of work and take entrepreneurial risks, and often tried to carve out a political space. In different ways, the eighteenth-century handloom weaver was all of these things as much as a garment subcontractor or owner of a power-loom weaving shop was at the end of the twentieth century. The notion of a producer who was adaptable, and who never became completely out of date, enables the drawing of lines that connect the eighteenth-century handloom with the late twentieth-century powerloom weaving shop, that connect the cotton mills of Bombay and Ahmedabad, which often relied on such native resources, with the artisan world around them and that connect India with other regions of the world. The focus of this chapter will fall on this interaction between macro episodes and micro-scale adaptation. Although it follows a chronological structure, and often presents extended review and reinterpretation of the episodes themselves, the point of emphasis is the response. I do not wish to suggest that the macro was wholly exogenous and that feedbacks did not occur. And yet, in an industry that was truly global, the producers were often indeed subject to forces that were beyond their control. In any case, I find the distinction useful from an expositional point of view. The material here is organized around five such episodes, each one of which redrew the relationship between India and the world economy in a particular way. These are early modern trade, nineteenth-century globalization, growth of modern industry, autarky and re-integration. In the concluding section, I shall return to a question in historiography: What features hold this narrative of an extended globalization together?

Early modern trade In 1750, India supplied about a quarter of the world’s industrial output, and possibly a larger percentage of world textile exports, and was a major protoindustrializing region before technology imposed a rearrangement of comparative advantage. Our knowledge about this episode is sadly partial, being too dependent on European trade records. Maritime trade, after all, was a small part of the larger textile economy before 1850, and we know less about the segment of textile

 Frank Perlin argues the case for treating coastal South Asia on the same level as eighteenth-century rural industrialization in Western Europe, ‘Proto-industrialisation in Precolonial South Asia’, Past and Present 98 (1983): 30–95. 255

The Ashgate Companion to the History of Textile Workers production that traded locally and overland. And yet, the significance of the trade for the regional economies, for integration of parts of the region in the international economy and in some case for political transition cannot be overlooked. The major exporting regions in India were Punjab, Gujarat, the Coromandel coast and Bengal. These regions had or could access good quality raw cotton, water in many places where dyeing was done and plenty of labour. The industry in the north was more urban than in Bengal or Coromandel. The rural weaver worked from within a household unit, the women and children of which supplied auxiliary services including spinning in some cases. There were exceptions in the towns of the northern Gangetic plains, in the form of large workshops owned by rich courtiers. Workers within these units typically consisted of all-male master– apprentice teams. In the larger of these workshops, more than one such team assembled under a single shed. We encounter these two general types – households and male master–apprentice teams – for a long time subsequently. Both tended to dissolve into various mutant types of wage labour-based workshops in the late twentieth century. Weaver settlements changed constantly as politics made rules of exchange and property rights unstable, and famines affected costs of production and of survival. A certain readiness to migrate was the most effective means the weavers possessed to deal with natural and manmade upheavals. Over these routine insecurities, there came about a long-term process of change. Contractual ties between the trader and the producer progressively strengthened in place of spot transactions, and a more diverse group of intermediaries began to be involved with the penetration of Europeans in Indian Ocean trade. The dispersed location of production and increasingly concentrated markets involved rather high transaction costs. The need to enforce quality and standardization, and the need to ensure timely supplies, were forever problems the European traders grappled



At the end of the eighteenth century, cloth exports from India amounted to about 50 million yards, whereas total production within India could not have been less than 1,800–2,000 million yards. For the former figure, see M.J. Twomey, ‘Employment in Nineteenth Century Indian Textiles’, Explorations in Economic History 20 (1983): 37–57.  On the seventeenth century especially, see K.N. Chaudhuri, ‘The Structure of Indian Textile Industry in the Seventeenth and Eighteenth Centuries’, reprinted from Indian Economic and Social History Review, in Tirthankar Roy (ed.), Cloth and Commerce: Textiles in Colonial India (Delhi, 1996), pp. 33–86; and Sanjay Subrahmanyam, ‘Rural Industry and Commercial Agriculture in Late Seventeenth Century South Eastern India’, Past and Present 126 (1990): 76–114. See also Om Prakash, European Commercial Enterprise in Pre-Colonial India (Cambridge, 1998), for a general economic history.  Chaudhuri (‘Structure of Indian Textile Industry’) suggests that the prevalence of wheeled traffic in the north facilitated inter-urban movement of goods, even if it tended to bypass rural produce. In the south pack animals and in the east the waterways made villages more accessible to trade and preserved rural production. That said, the urban centres of northern India, which were also centres of power, tended to evolve distinct cultures of work and industrial organization through interaction with west and central Asia, for example through artisans who migrated from west and central Asia. 256

The long globalization and textile producers in India with. No easy or permanent solution was found, leaving enough room for breach of agreement, fraud, default and, in turn, extortion and coercion. From the seventeenth century, a hierarchy began to take shape in the weaving villages. Historians distinguish between weavers who did not possess looms, weavers who possessed looms, caste-heads and village leaders, weaver-traders and traders by caste. It is likely that the position of headmen consolidated, for the Europeans needed powerful middlemen to communicate and to enforce deals, and they dealt with the caste-traders and the weaver-traders. As long as several companies and numerous ‘interlopers’ or traders who did not belong to the companies were working in the coastal villages, the weavers and the headmen enjoyed considerable bargaining power. As the English settled in politically, a sort of command-and-control system was devised to take over these functions. Both in Bengal and in Coromandel, the English company attempted to enforce measures to take closer control of the producers, by replacing headmen with paid officials. The process was not an abrupt one, and for some time the headmen were co-opted into the ‘reform’ process. By the end of the eighteenth century, the paid contractors in Coromandel were more usually recruited from the peasant castes. These outsiders, armed with some measure of governance power, brought in coercion and abuse, and were resisted by both the weavers and the brokers. The figure of Chinnum Jaggapah on the south-eastern coast, described in Swarnalatha’s recent work, is an example. The story is different only in detail in Bengal. Purchases through agents began to give way to procurement by salaried employees. The officials in charge of procurement began to acquire local authority and imposed arbitrary exactions. The Bengali weaver was not as prone as the southern counterpart to migrate, possibly because the incidence of land-holding by the weaver was greater in the east. But that very feature exposed these part-time peasants to pressures from the revenue system. In recent scholarship, the relationship between handloom weavers and the English East India Company in southern India has been explored from a perspective of disempowerment. Parthasarathi documents a story of weavers losing bargaining power and customary rights to the Company as it consolidated itself politically in interior Tamilnadu. Swarnalatha deals with a similar tendency  S. Arasaratnam, ‘Weavers, Merchants and Company: The Handloom Industry in Southeastern India, 1750–90’, reprinted from Indian Economic and Social History Review, in Roy (ed.), Cloth and Commerce, pp. 85–114; and J. Brennig, ‘Textile Producers and Production in Late Seventeenth Century Coromandel’, Indian Economic and Social History Review 23 (1986): 333–355.  Arasaratnam, ‘Weavers, Merchants and Company’; H. Hossain, ‘The Alienation of Weavers: Impact of the Conflict between the Revenue and Commercial Interests of the East India Company’, reprinted from Indian Economic and Social History Review, in Roy (ed.), Cloth and Commerce, pp. 115–141.  P. Swarnalatha, ‘Revolt, Testimony, Petition: Artisanal Protests in Colonial Andhra’, International Review of Social History 46 (2001): 107–29.  Prasannan Parthasarathi, The Transition to a Colonial Economy: Weavers, Merchants and Kings in South India 1720–1800 (Cambridge, 2001). 257

The Ashgate Companion to the History of Textile Workers on the northern Coromandel. Both go further than other major works in the field in analysing weavers’ response to their progressive disempowerment, in particular by studying migration and various forms of resistance and protests. Their work suggests increasing institutionalization of the form of protest, as consolidation of British power led not only to innovation in organization, but also to innovation in legal instruments for contract enforcement.

The Industrial Revolution and the artisans English yarn first reached Indian ports late in the second decade of the nineteenth century. Quantity and value of imports rose rapidly from 1820. In perhaps the mainstream view, this was a period of devastation in Indian industry. Joined with a politically induced agrarian depression, the ‘de-industrialization’ created a general economic crisis of low demand and falling output. The picture of devastation is now being revised. Globalization in the nineteenth century saw an unprecedented increase in the scale of world commodity trade, shortly followed by movements in capital and labour. These processes affected India in two principal ways. The price of manufactures, chiefly cloth, declined greatly, whereas the demand for agricultural goods exported from India increased, leading to sustained upward pressure on agricultural prices. This pattern was consistent with what trade theory would predict for a region with abundant land and shortage of capital or skilled labour needed for modern industry to develop.10 One would imagine that this dual revolution would push many artisans to land. Such predictions need to be made cautiously. A great deal of the pessimistic assessment has used the evidence of price depression as evidence of economic depression, which is an incorrect procedure for the early nineteenth century. The price depression in the nineteenth century was driven largely by technological change, which had the effect of large outward shifts in the supply curve. Such a process implies a demand expansion, and not a demand contraction. Further, large segments of artisan production were protected from competition of low-price imports by traditional preferences or the particular skills that these 

The idea of an agrarian depression in the second quarter of the nineteenth century has been articulated mainly in regional histories on the Bombay-Deccan and Madras. For an older discussion, see P.J. Thomas and B. Natarajan, ‘Economic Depression in the Madras Presidency (1825–54)’, Economic History Review 7 (1936): 67–75; and a recent discussion, D.A. Washbrook, ‘India, 1818–1860: The Two Faces of Colonialism’, in A. Porter (ed.), The Oxford History of the British Empire, Vol. III: The Nineteenth Century (Oxford, 1999), pp. 395–421. On de-industrialization, see A.K. Bagchi, ‘De-industrialisation in India in the Nineteenth Century: Some Theoretical Implications’, Journal of Development Studies 12 (1976): 135–64. 10 For an application of trade theory to the study of convergence in the Atlantic economy, see K.H. O’Rourke and J.G. Williamson, Globalization and History: The Evolution of a Nineteenth-century Atlantic Economy (Cambridge, Mass., 1999). 258

The long globalization and textile producers in India goods embodied. Such segments would gain from the price revolution by accessing cheaper imported yarn or dyes, and via increased purchasing power of the consumers. In the 1960s, Morris pointed out the possibility that low-price imports had the potency to affect artisan industry positively.11 Indeed, descriptions of regional experiences do not necessarily reveal a picture of devastation and misery in the early nineteenth-century textile industry. There was undoubtedly a fall in the production of cloth and yarn. But the fall was more evenly spread over time than has been believed. There never occurred a dramatic crisis, but only a persistent depressive tendency that was increasingly being compensated as the century wore on. Data gathered by Specker shows that but for famines and epidemics, nearly every attempt at a loom count in the early nineteenth-century Madras Presidency found the number of looms on the increase.12 Studies on central India suggest that the decline in spinning was rather mild in 1820–1850, and accelerated only after 1865, about which time the major railway routes opened.13 As late as 1870, the share of imports in total yarn used in Madras Presidency was 37 per cent. In 1890 the percentage was 55. Bengal, being more accessible by sea than central India or the Deccan, may well show a different pattern, one which conforms more closely to the classical de-industrialization timetable. Exports to Calcutta were much larger in volume than exports to Bombay or Madras until late in the nineteenth century. But Bengal was probably an exception. And yet, there is undeniable empirical evidence suggesting that the community of hand-spinners gave up spinning on a large scale, and this factor alone may account for a loss of industrial employment to the extent of 4–5 million persons according to the best benchmark estimates available. Such astronomical figures of job loss left little or no impact on national income statistics of the late nineteenth century. Data gathered from official statistics suggest that most urban wages were stable, even rising in the late nineteenth century. Agricultural real wages declined marginally during the 1890s, but this finding is contradicted by another contemporary source on wages.14 With job loss in rural India of such order, how did real wages remain stable? A possible answer is that spinning was an extremely labour-intensive activity (on average, one fully employed weaver needed 2–3 spinners working full time), that continued more or less solely because of the rhythm of monsoon agriculture, which made available an extraordinary quantity of surplus labour time in rural 11 Morris D. Morris, ‘Towards a Reinterpretation of Nineteenth Century Indian Economic History’, Journal of Economic History 23 (1963): 606–618. 12 Konrad Specker, ‘Madras Handlooms in the Nineteenth Century’, reprinted from Indian Economic and Social History Review, in Roy (ed.), Cloth and Commerce, pp. 175–217. 13 P. Harnetty, ‘Deindustrialization Revisited: The Handloom Weavers of the Central Provinces of India’, Modern Asian Studies 25 (1991): 455–510; S. Guha, ‘The Handloom Industry of Central India, 1825–1950’, reprinted from Indian Economic and Social History Review, in Roy (ed.), Cloth and Commerce, pp. 218–241. 14 For more discussion on wages, see T. Roy, Rethinking Economic Change in India: Labour and Livelihood (London, 2005), ch. 4, pp. 86–105. 259

The Ashgate Companion to the History of Textile Workers India. A lot of the spinning labour was unpaid, when performed by women in weaving households, or underpaid, when performed by agricultural labour castes on the side. The ruin of hand-spinning did see massive loss of jobs, but rather little loss of income. Furthermore, there is the possibility that some of this surplus labour could find alternative employment in agriculture, which did not yet face a land shortage. A similar point can be made about the timetable and spatial spread of decline in weaving. In Bengal the decline probably peaked earlier than in peninsular India and a large part of interior northern India. In the early nineteenth-century south India, imported cloth was only a minor factor in the experience of weaving. Famines, loss of export market and decline of the elites were much more important factors. In central India it was only after the major railway routes opened c.1870 that the impact on cloth production became serious. The Maratha regimes immediately preceding the British in central India (about 1775–1815) patronized urban handicrafts on a large scale, giving rise to a number of weaver settlements in the region. Some of these clusters languished after the change of regime, but some of the others more fortunately served by mass markets and transportation routes survived and expanded in the late nineteenth century. In southern and central India, both spinning and weaving felt the pressure of import from the decade of the Cotton Famine (1860s). But by then, several positive forces working on handloom cloth demand had indeed begun to consolidate. Agricultural exports and agrarian commercialization strengthened local demand, and export of textiles from south India to the new settlements of expatriate Indians began to increase. With these positive forces at work, the period 1870–1940 saw expansion rather than contraction in weaving. Throughout this period, the quantity of cotton yarn taken by the handloom industry was increasing, even though the number engaged in this industry fell. Handlooms accounted for about 25 per cent of the cotton cloth produced annually in the first half of the twentieth century. Market-share of handloom cotton cloth was roughly stable between the 1890s and the 1930s. Production of cotton cloth in handlooms expanded 30 per cent between 1900 and 1939. Through this period, total cloth consumption was growing, and Indian cloth was steadily substituting imported cloth. In cloths made of silk and other fibres, handlooms dominated. These other fibres were generally more expensive than the average quality of cotton. Taking all fibres together except wool, in the 1930s handlooms’ market share in total cloth consumption in value may have been about 50 per cent.15 What were the sources of survival? First of all, competition between handloom and power-loom was restricted. In the mid-nineteenth century, two types of cloth faced competition from mill-made cloth: ‘coarse-medium’ cotton, and printed and bleached cotton. By contrast, cloths that used very coarse or very fine cotton yarn, or complex designs woven on the loom, or non-cotton yarn tended to use the handloom. These were so labour-intensive that the mills did not enter them by choice or used non-cotton fibres that the mills did not want to handle, or employed 15 T. Roy, Artisans and Industrialization: Indian Weaving in the Twentieth Century (Delhi, 1993), ch. 2, pp. 42–71. 260

The long globalization and textile producers in India flexible specialization mainly by using a range of decorative technologies. Second, the market was growing. Average cloth consumption increased between 1870 and 1940, partly as a result of a fall in cloth prices, and partly because of agricultural growth in some areas. It is suggested that growing nationalist sentiments in the interwar period, the ideology of ‘Swadeshi’, also helped the weaver. But, in fact, the Swadeshi campaign against foreign yarn and cloth, by favouring the Indian mills, hurt the weaver who consumed a lot of foreign yarn and sometimes competed with the mills. Further, its impact was variable over space; in Bengal where Swadeshi sentiment was the most intense, handloom weaving actually declined. Perhaps most significant of all, labour became more productive via a process of institutional change in the early twentieth century. In the early twentieth century, the household was in decline and giving way to a variety of wage-based units. By 1920, a significant number of handloom factories had appeared in the textile towns of western and southern India. In the course of this shift, the average hours worked increased, for wage workers worked longer hours per day than did family workers, and the work site could incorporate a range of somewhat larger-scale tools and equipment. This endogenous drive to efficiency was due to a number of changes that were happening at the same time: migration of labour and capital, opportunities for mercantile accumulation and the availability of new technical options. Capital and labour involved in the handloom industry became increasingly mobile. There was migration from rural regions towards new points of trade, and towards the railways and spinning mills. The most important example is a migration into textile towns in western India such as Sholapur, Malegaon, Bhiwandi, Burhanpur and Surat. The workers as well as the capitalists were handloom weavers who came from regions like the eastern districts of the United Provinces, with decaying urban economies and demographic pressure upon the land, and the famine-prone arid districts in the Hyderabad state.16 Today, in nearly every major town of western India, the older quarters house settlements of handloom weaver communities, many among whom continue textile business but rarely on handlooms. Not only was the handloom market a growing one, long-distance trade in cloth and raw materials expanded too. Descriptions of trade in major handloom sari brands and of regional trade networks suggest growth in trade networks.17 Cotton yarn, earlier made locally by hand, was either imported or manufactured in mill towns, and distributed in handloom towns and villages by merchants specializing in these new trades. Dyes, gold-thread and silk yarn were earlier obtained locally, and now increasingly imported. These tendencies created more capital for some groups 16 See on weavers’ migration, Douglas Haynes and Tirthankar Roy, ‘Conceiving Mobility: Migration of Handloom Weavers in Precolonial and Colonial India’, Indian Economic and Social History Review 36 (1999): 35–68. 17 On the development of a regional trade network with reference to handlooms, see Douglas Haynes, ‘Market Formation in Khandesh, c. 1820–1930’, Indian Economic and Social History Review 36 (1999): 275–302. 261

The Ashgate Companion to the History of Textile Workers of weavers. Merchant profit as a category, and merchants themselves became more noticeable in sources on weaving from the interwar period. Rates of profit were high by any standard around the 1920s. Profits were squeezed somewhat in the post-Depression period, but did not decline to the extent that wages did. The early twentieth century saw increasing adoption of several new types of tools and processes.18 Two things combined to encourage technological dynamism: the efforts of provincial governments in popularizing these instruments and endogenous capitalist tendencies. The increasing wealth and knowledge of capitalist weavers and the increasing stability of their markets made them more willing to try out new tools. The older vintage of loom was the throw-shuttle type, set up in a pit dug up in the living room of a weaver’s home. From this system, there was change towards the fly-shuttle loom or loom mounted on a wooden frame. The frame loom took up less space, could weave longer lengths of yarn and thus became popular with the handloom factories. Warp preparation changed from various means by which the warp was stretched out, to the use of a warp beam such that longer lengths of thread could be woven. Warp preparation was previously a side activity of women in weaving localities. This form of collective labour was replaced by the warping mill. Another type of technical change was the switch to synthetic dyestuff in place of vegetable and animal substances, with the result that dyeing became less knowledge-intensive and less a specialized occupation. A final stage in this process of endogenous technological change was the ‘powerloom’ factory. From the frame loom, the idea of a power-driven frame loom was a small step. Mills always discarded power-looms at scrap rates. Better-off weavers bought such looms, reconditioned them and started to replace handlooms with power-driven looms for products where such a switch was possible. The first such weaver-owned power-loom factories appeared in handloom towns about 1900 and used fuel oil. They spread faster from the 1930s when many such towns received electricity. By 1940, there were 15,000 such looms. Persons with a handloom background had started these enterprises. The ground had thus been prepared for what was to become in the next few decades India’s largest industry.19 Although the distance between the household and the workshop was often a small one in a weaving town, a subtle but gradual change had begun in relations of production. Employment practices in the old systems contained an element of ‘regard’ that took different forms – caste, community, masterhood – according to the context. The community at times enabled different ways of pooling and allocating labour as with the example of warping above. This whole range of socio-economic systems began to give way to casual labour, factory labour and migrants in new 18 D. Haynes, ‘The Logic of the Artisan Firm in a Capitalist Economy: Handloom Weavers and Technological Change in Western India 1880–1947’, in B. Stein and S. Subrahmanyam (eds), Institutions and Economic Change in South Asia (Delhi: 1996); and T. Roy, ‘Acceptance of Innovations in Early Twentieth Century Indian Weaving’, Economic History Review 55 (2002): 507–532. 19 D. Haynes, ‘Artisan Cloth-producers and the Emergence of Powerloom Manufacture in Western India 1920–1950’, Past and Present 172 (2001): 170–198. 262

The long globalization and textile producers in India industrial towns. Wage relations continued for long to be embedded in non-wage ones. Class formation, however, was far advanced in the power-loom towns by the end of the twentieth century. Who gained from this process of increasing industriousness? There were three distinct types of income-earners about the middle of the twentieth century: pure wage workers, households and merchants or capitalists. Statistics on ‘wages’ and ‘earnings’ in handloom work can be ambiguous. In an industry undergoing a change from households to factories, ‘wages’ sometimes referred to the income of families working on putting-out contracts, and sometimes to wages of hired labour. There is no easy way to disentangle these two types from the data. We do know that the volume of handloom cloth production more than doubled between 1900 and 1940. There is no complete price series for handloom cloth, but since the index of industrial prices shows considerable increase, the value of production increased by a higher ratio. And yet, the poorest weavers do not appear to have gained from this growth at all. Sivasubramonian concluded that weavers’ money wages ‘declined from a high level [at 1900] to very low levels by the end of the thirties’.20 Major official sources too were generally pessimistic about wages. There was some increase in the interwar period, but a reversal in the post-Depression years.21 If we move from ‘wages’ to ‘earnings’, and from lower to higher qualities, then there is evidence of improvement. Several datasets on earnings from silk, fine cotton and exportable goods over different pairs of years in the long span 1880–1940 suggest, after adjustments made for price movements, a rising tendency in general.22 A similar point emerges through data on profitability. In the interwar period, prices of cotton cloth generally fell, but money wage and yarn price fell somewhat faster than product price in handlooms. Value added and profitability increased in 1924–1940.23 It seems reasonable to conclude that inequality increased in handloom weaving. Some households were becoming richer than the average, and capitalists gained more than the wage worker. Inequality among the households can be explained in terms of the variable intensity of competition from power-loom cloth. Not all weavers were equally affected, or equally capable of handling the crisis. Cotton was worse off than silk, plain weaving worse off than designed cloth, families of weavers in arid famine-prone agrarian zones were worse off than their counterparts in prosperous areas. The capital-labour gap can be explained in terms of new avenues of profit-making that not everyone could exploit, and which are discussed above with reference to mercantile opportunities and merchant profits. That being said, growing inequality followed the tracks of pre-existing differentiation based on product, status, skills and honour. Handloom weavers 20 S. Sivasubramonian, National Income of India in the Twentieth Century (Delhi, 2000), p. 282. 21 Roy, Artisans and Industrialization, ch. 1, pp. 1–25. 22 T. Roy, Traditional Industry in the Economy of Colonial India (Cambridge, 1999). 23 Roy, Artisans and Industrialization, ch. 1, pp. 1–25. 263

The Ashgate Companion to the History of Textile Workers were always differentiated. At the top were groups that earned quite good income by contemporary standards, who manufactured silk or cotton cloths involving complicated designs, often used elaborate looms, were urban, took part in organized long-distance trade and were in some way or other involved in a patronage network. Very often, such weaver elites saw themselves as aliens in their local society. These ‘origin myths’, which served a purpose of emphasizing difference and otherness, were in a few cases well founded, and reflected the fact that the lords of the land always tried to import and settle skilled artisans in their domains in the interest of promoting commerce and meeting their own consumption needs. Several examples of migration, mythical or otherwise, have been studied recently.24 At the other end of the scale were the rural artisans who supplied simple routine articles to their peasant neighbours. In the case of coarse cotton textiles, artisans and agricultural labour tended to coalesce, especially in central and western India. A large group of people performed both agricultural labour and crafts according to season. Rarely were the articles they made traded beyond the immediate neighbourhood. Some of these rural artisans, for example, were known to produce and sell high quality hand-spun yarn. Such forms of manufacturing, however, did not survive on a large scale beyond the interwar period. Generally speaking, the weaver handling silk or very fine cotton cloth had income between three and five times that of a weaver making cloths variously called ‘coarse’ or ‘ordinary’. Between these easily identifiable extremes there were a great many types of weaver who combined rural and urban, local and long-distance trade, high and low skill, more capital or little capital. They strove to achieve elitehood, but were often pushed by circumstances to regress to labour. These long-standing hierarchies influenced how groups of weavers responded to the upheaval of regime change and globalization. The rural labour castes engaged in coarse weaving on the side tended to leave the industry for mill labour and semi-skilled services. The urban communities and silk weavers responded differently. When the silk weavers had to leave weaving, they tended to shift to skilled professions and trade rather than labour. But they also left weaving to a much lesser extent. In the interwar censuses, the percentages engaged in weaving among silk or fine cotton weaver castes was considerably higher than in castes that combined weaving with agricultural labour. The persistence with weaving in the former case meant making decisions to invest in new equipment, enter new market segments or leave depressed regions to resettle in towns with a flourishing handloom trade. And those artisans who made these decisions could do so because they had easier access to a cooperative community, knowledge of markets and techniques, and the money to make investments. In another way, institutional change enhanced inequality, by inducing the exit of women from artisan textiles. Women’s work was closely related to the strength of the household. Any weakening of the household could diminish women’s presence in the workforce. In cotton handloom weaving, the household was universal about 1900. Women and girls performed warping of the yarn, winding the thread on to 24 Haynes and Roy, ‘Conceiving Mobility’. 264

The long globalization and textile producers in India bobbins for use as weft yarn, and shared with men the task of sizing, that is applying starch paste on the warp thread. Weavers as a rule were men, sometimes assisted by young adults, almost invariably boys. About one in ten women in weaver families knew how to weave. Usually, they learnt weaving from observation rather than via systematic apprenticeship such as the boys went through. In fields allied to textiles, such as sericulture, women performed vital tasks in the manufacturing process. In the textile industry, thus, men and women usually participated as complementary elements in a gendered division of labour. Wherever men’s and women’s work could be connected in this way, women’s work participation tended to be higher and so did the presence of household industry. Thus, women formed 40 per cent of the textile workforce, and a similar percentage in household industry. The situation changed in the twentieth century. The household was in decline in handloom weaving, in some cases being replaced by factories working with migrant workers. The general effect has been adverse for women’s participation in manufacturing. The textile scenario falls in a more general pattern. For nearly a century spanned by the Indian censuses, the percentage of women in the industrial (and textile) workforce steadily declined (1881 to 1971), beginning to rise slowly after 1971. Why would the weakness of the family as a work unit lead to withdrawal of women from the industrial workforce? In part because institutional change demanded and encouraged migration, which was always gender-biased; in part because, no matter how important women’s work was inside a household, men tended to have greater access to capital and marketing. In other words, men faced a higher opportunity cost than women of leaving the industry altogether. Faced with market pressure, men often looked for alternatives within manufacturing, but women usually left manufacturing.25

25 For more discussion on women’s work in small-scale industry, see Roy, Rethinking Economic Change, chs 7–8, pp. 134–164. 265

The Ashgate Companion to the History of Textile Workers Table 10.1 Scale of the Indian textile industry, 1920 and 1940 1920 Handloom industry Workers Looms (Cotton looms) Idle looms Production (million yards of cotton cloth) Percentage of domestic cotton cloth consumption Mill industry Workers Looms Production (million yards of cotton cloth) Percentage of total domestic cotton cloth consumption a

1940

2,407,300a 2,025,000b (…) (…) 931 25

(…) 2,193,262 (1,417,200) 265,464 1703 28

332,200 117,558 1,529 44

625,000c 199,000c 3,738 62

Census estimate of ‘actual workers’.

There was a handloom census in 1921, which did not cover several major states and provinces. The 1940 estimate is more comprehensive. I have multiplied the 1940 loomage with the 1920:1940 ratio of looms in regions covered in both years to derive an estimate for 1920.

b

c

Refers to 1943.

Source: India, Fact-Finding Committee (Handlooms and Mills) (Delhi, 1942), ch. II.

The Industrial Revolution and the mill worker As suggested above, the partial decay of manufacturing in the course of India’s integration into the nineteenth-century international economy occurred because the region lacked some of the resources needed for modern industry to develop, mainly financial capital and skilled labour. I then added one qualification. The decay did affect the artisans, but not all of them. In fact, those who survived gained from the globalization-induced price revolution. It is necessary to add another qualification to this story; India’s colonial connection could be used to partially overcome the supposed disadvantage in capital and skilled labour, giving rise to a successful cotton mill industry from the end of the nineteenth century. At the end of colonial rule, the mill industry gave direct or indirect employment to nearly half a million people, and had a substantial presence in world trade in textiles. Factory labour was a new form of work in India in the middle of the nineteenth century. It represents a shift from cultivation or the crafts to the mechanized factory located in large metropolitan cities. Most workers in modern 266

The long globalization and textile producers in India industry were immigrants into Bombay or Calcutta, and faced very different work, workplace, habitat and social milieu. A great deal of ‘labour history’ scholarship of the region has been engaged with what can be called the transplantation problem. Who were the mill workers? They came from the western coast, the Deccan and western United Provinces in the case of the cotton mills of Bombay, and north Bihar, eastern United Provinces and Orissa in the case of the jute mills of Calcutta. These streams of migrants fell into the larger pattern of internal migration in colonial India, which consisted of two broad streams. The first went from north India towards Calcutta, Assam and Bombay. The second went from south India to Bombay. Earlier research on labour migration suggested that the ‘reserve price’ of labour was low in the source regions because of poor agricultural conditions.26 These regions were either overpopulated or arid. Is it poverty, then, that explains migration into mills? More recent work suggests that poverty at source does not explain migration completely. For one thing, migration was ‘segmented’, migrants of specific background moved into specific occupations.27 For another, migrants came from all sections of the peasant society, and were not necessarily the poorest classes. In both Bombay and Calcutta, there were individuals who owned land back in the villages they came from. They were usually adult males, who had left their families behind in the villages to look after the rural work. Their families migrated if rural resources became exhausted. City work was not a total break from, nor a clear alternative to, rural work.28 Understanding such migrations requires a different framework. The concept of an ‘industrious revolution’ can supply some of the elements of such a framework. Along with reduction of poverty and reduction of the risks entailed with exclusive dependence on land, there was perhaps another desire at work, to attain higher levels of household consumption by working harder in the wage labour market. In turn, such decisions might involve reallocation of household labour time, whereby women formerly engaged in commercial work retreated to domestic duties, and men (because they were more open to migration) increased their participation in wage work.29 How were the migrants recruited into the mill labour force? Much of the movement of labour into the modern sectors relied on ‘contractors’ and foremen, called jobbers and sardars in the cotton and jute mills. Sardars undertook to gather a number of workers usually from areas where they had personal connections. They sometimes even managed the workers on the factory floor. Such intermediaries facilitated supply of labour, provided the workers with a comforting link between 26 Lalita Chakravarty, ‘Emergence of an Industrial Labour Force in a Dual Economy – British India, 1880–1920’, Indian Economic and Social History Review 15 (1978): 249–328. 27 Arjan de Haan, ‘Migration in Eastern India: A Segmented Labour Market’, Indian Economic and Social History Review 32 (1995): 51–94. 28 Rajnarayan Chandavarkar, The Origins of Industrial Capitalism in India: Business Strategies and the Working Classes in Bombay, 1900–1940 (Cambridge, 1994), pp. 124–167; Samita Sen, Women and Labour in Late Colonial India: The Bengal Jute Industry (Cambridge, 1999), pp. 21–53. 29 J. DeVries, ‘The Industrial Revolution and the Industrious Revolution’, Journal of Economic History 54 (1994): 249–270. 267

The Ashgate Companion to the History of Textile Workers their old and new homes, represented them in collective bargaining and sometimes trained them at work. Morris suggests that as the workforce was linguistically diverse, communication with the managers involved an intermediary.30 The employers often found it hard to gather a large number of hands without the help of agents. Labour markets with sufficient depth to supply mill-hands on large scale did not exist in the mid-nineteenth century, except briefly during famines. Contractors helped avoid a costly search for workers and the transaction cost of having to monitor a socially diverse workforce. The power that these intermediaries enjoyed as a valuable link between the owners and the workers was often misused. Because the employers and the workers could not communicate directly, the contractors could collect fees from both sides. That said, on both sides, the need for the unreliable intermediary diminished progressively. After 1900, voluntary internal migration increased. Eventually several types of transaction costs involved in hiring labour fell. A range of skills became available for hire in one place. Potential workers came to the mill gates rather than wait for the sardar to take them there. Mill managers in the interwar years reported that they hired their daily hands from the ‘mill gates’, breaking with past practice, whereas it was only the new factories and new divisions that still relied on the agent.31 Were the mill workers better off by coming to the city? Detailed research on the early mill worker’s family income is scarce. Nevertheless, mill wages were three to five times above that of an agricultural labourer, and possibly double that of the average peasant household annual income in about 1875. The city-dwellers never suffered the threat of famine in the way the rural population did. Caste oppression and caste as barrier to entry in new jobs were a weaker force in the cities. For individual workers, the opportunities of occupational and income mobility were far greater in the cities than in the villages. On the other hand, until 1920 or so, the average mill wage was too little to think of growing roots in the city, bringing families in and giving up connections with land. But conditions were changing. Real wages of mill workers, though nearly stagnant for more than half a century since mill expansion accelerated in the 1860s, began to rise steadily from early in the interwar period.32 Agricultural incomes and wages, on the other hand, were stagnant in this period. Collective action played an important role in this unusual wage movement. Employment legislation (Factories Act) and trade union legislation (Industrial Disputes Act) evolved slowly during the colonial period. They began in the 30 M.D. Morris, ‘The Recruitment of an Industrial Labor Force in India, with British and American Comparisons’, Comparative Studies in Society and History 2 (1960): 305–328. 31 Report of the Royal Commission on Labour in India (London, HMSO), evidence vol. III, part I, Central Provinces and United Provinces, pp. 8, 144. 32 K. Mukerji, ‘Trend in Real Wages in Cotton Textile Mills in Bombay City and Island, From 1900 to 1951’, Artha Vijnana 1 (1959): 82–95; ‘Trends in Real Wages in Jute Textile Industry from 1900 to 1951’, Artha Vijnana 2 (1960): 57–69; ‘Trend in Real Wages in Cotton Textile Industry in Ahmedabad from 1900 to 1951’, Artha Vijnana 3 (1961): 124–134. 268

The long globalization and textile producers in India nineteenth century, but were somewhat effective and widely known codes only in the interwar period. A singular feature of labour practices in Bombay was the delayed and somewhat retarded development of bargaining institutions, a feature that took a toll in the long run in a particularly unstable industrial relations scenario. Bombay witnessed strikes before the mid-1920s, but these tended to be spontaneous, sporadic and sudden outbursts, which sometimes generated a scale of response that surprised the employers and the workers. These strikes were in reaction to issues such as wages and working hours. But these occasions ended as suddenly as they arose, and on several such occasions, observers noted the workers’ inability to combine in a systematic way or sustain an action. The first successful attempt at a large-scale trade union in Bombay was that of N.M. Joshi, who established the Bombay Textile Labour Union in 1926. In the next decade, at least eight other major unions started work, while a few older and small ones merged with the larger organizations or simply became defunct. The major issue in the 1920s was wage standardization across mills, a problem that reflected the haphazard way labour practices had developed in the city’s textile mills. In the Depression-hit early 1930s, several major strikes and closures put the industry under great stress, and drew the government deeper into the ‘conciliation’ process. One legacy of this involvement left an indelible mark on the bargaining process. This was the Bombay Industrial Relations Act 1946, which recognized the Congressaffiliated Rashtriya Mill Mazdoor Sangh (RMMS) as the sole representative of textile workers in the city. In these 20-odd years of organized trade union movement, the difficulties of conducting unions were legendary – membership fluctuated widely, as most workers correlated their participation with settlement of immediate grievances. The migratory character of the workforce and their rural roots, contemporary observers believed, reinforced the syndrome of ‘opportunistic’ participation. The organizations consequently relied too heavily on individual initiatives. Employers, on their side, allegedly ‘victimized’ union workers when they could. These abuses were counteracted only by comprehensive legislation in 1938. And finally, as the movement for workers’ rights inevitably became connected with anticolonial struggles and sentiments, the government did not necessarily look upon mobilization efforts kindly. Nevertheless, trade union membership increased from 100,000 in 1927 to 400,000 (or one-third of the factory workforce) in 1938. Not everyone gained from an increasingly regulated and militant labour regime. In the late nineteenth-century mill, women worked on the shop floor in those tasks that traditionally engaged women in the handloom-weaver household, namely, winding, sizing and sometimes warping. As with men, a team of women workers was supervised by an intermediary, in this case female. She was the recruiter, supervisor and medium of communication between the managers and the workers. Almost invariably, only those women whose husbands were employed there came to the mill. Yet, fairly strict gender segregation was observed. Despite an increasing trend in women’s migration, the share of women in the factories declined late in the interwar period. Women tended to be replaced by migrant males. Employment legislation made special provisions for women that often discouraged employers 269

The Ashgate Companion to the History of Textile Workers from employing them when male labourers were available for the same work. The prohibition on night work for women and maternity provisions were the new regulations fiercely debated in contemporary literature. Sen has recently argued in the context of the jute industry that with increasing withdrawal from paid and men’s work women’s work tended to get devalued.33 Historians of Indian mill labour have asked the question, why the workers did not take part in collective action more often, from an earlier time and more effectively. If class formation was muted, might this be due to cultural reasons? In one view, the jute mill worker was indeed mentally a migrant peasant, and therefore politically too raw for collective action.34 The ‘peasant-worker’, in that view, was held back from becoming a conscious member of the working class by various kinds of divisive ‘primordial ties’ like caste, region, religion, community, gender and language. The mill town was a melting pot, but the ingredients in this pot never mixed well together. This image of a worker carrying a primitive mentality, originally constructed in colonial administrative discourse, has been questioned more recently.35 In more recent works on labour history, ideas and the environment are not seen as either distinct or unchangeable, but evolving in mutual interaction. The fact that despite diverse backgrounds the workers shared common grievances over wages, hours, workload and safety did bring them together at times. And yet, they understood and expressed these grievances through a language of custom, fairness and honour that did not originate in a capitalist relationship, but derived from the networks of caste, religion or region.36 Class and social affiliations were not necessarily either at odds or identical, but an unstable mix that had the possibility to gel.

From autarky to reintegration India gained independence in 1947, with a commitment to socialist type planning and an accent on capital-intensive industrialization, pursued under an importsubstituting trade regime. The struggle for nationalism in the interwar period had seen three potential pressure groups taking shape: Indian business, factory labour and a Gandhian group that wanted protection for the handloom weavers and rural artisans. The policy regime that followed independence can be seen as an attempt to accommodate these three interests. Capitalists received protection 33 Sen, Women and Labour, ‘Introduction’. 34 D. Chakrabarty, Rethinking Working Class-History: Bengal 1890–1940 (Princeton, 1989), pp. 3–13. 35 Much of new labour history research examines the theme of competing and fluid identities. See, for example, Chandavarkar, Origins of Industrial Capitalism; Chitra Joshi, Lost Worlds: Indian Labour and its Forgotten Histories (Delhi: 2001); S. Basu, Does Class Matter? Colonial Capital and Workers’ Resistance in Bengal, 1890–1937 (Delhi, 2004). 36 Basu, Does Class Matter?, ch. 1, pp. 1–36. 270

The long globalization and textile producers in India from international competition, but agreed to abide by fairly stringent labour laws that offered unprecedented security of jobs in the formal sector. At the same time, a Textile Policy imposed restriction on expansion in the cotton mills, and extended a variety of fiscal and other concessions to the handlooms. Ironically, the policy did not prevent the decline of the handloom industry, though it did adversely affect the textile mills. The power-looms or weaving factories took up the place vacated by their joint retreat, despite the fact that the power-looms were officially unrecognized and illegal. As an evolution out of the handloom industry; the power-loom industry could clandestinely capture many of the concessions. And operating in the informal labour market, the power-looms could access low-wage labour. Some of the older trends in the handloom industry – accumulation of capital and commercialization of labour – merged into the transformation of the power-loom industry in the 1970s and the 1980s. The uneasy coalition between capital and labour remained intact during the first three decades of the new regime, when protected domestic markets gave mill owners sufficient scope for earning profits. Given protection in the final market, the average firm could bear excess labour costs and afford a bit of over-manning. This arrangement being state-mediated, the government was committed to intervening in dispute settlements. Accordingly institutions and laws were set up that involved the government as a powerful third party In the case of textile mills, special laws had already awarded monopoly to an ‘official’ trade union, the RMMS. By the end of the 1970s, a great many of the mills faced a saturated or slow growing home market in consumer goods. Protection in the goods market was proving to be not enough compensation for losing flexibility in the labour market. The mills’ progressive bankruptcy and the monopoly union that neither the capitalists nor the majority of the labourers trusted any more complicated the picture. In the 1970s, subcontracting with the power-looms could be a way of bypassing the Textile Policy, or employment of casual labour a way to bypass the labour laws. But these practices only made labour more restless in the 1970s. As formal and informal means of negotiation collapsed, a militant unionist who made exaggerated demands on the employers and was prepared to back those demands with endurance filled the void. The result was the 1982–1983 textile strike in Bombay, by some indices the biggest industrial action the world has seen.37 Many employers by then stood to gain from a strike. It gave them a chance to reduce inventory, rationalize production, even to exit. The strike eventually died out, ruining many workers, but it enabled a different kind of alliance between the capitalists and the government that bore fruit in a substantially pro-market and anti-regulation Textile Policy in 1985. The philosophy behind the change of regime became more general in India’s wide-ranging liberalization in the early 1990s. The Textile Policy eased restrictions on the mills, enabled them to import equipment for a much needed modernization and recognized power-looms.

37 The best detailed account of the strike is H.Van Wersch, The Bombay Textile Strike 1982– 83 (Delhi, 1991). 271

The Ashgate Companion to the History of Textile Workers For the surviving cotton mills, the new regime was a mixed blessing. It reduced import protection but did not change labour laws. The reform dice was loaded in favour of new firms that benefited from access to export markets and imported technologies, but did not have a labour overhead to deal with. Incumbent firms like the mills shared the gains unequally. Some became bankrupt, worsening an already serious case of unemployment of textile mill workers. Those that survived had to find ways of dealing with the labour overhead, and they did so without the help of formal legal reforms. Some of the methods adopted were contractual employment, voluntary retirement, ban on recruitment and outsourcing. On the other hand, new industries and firms spawned completely different industrial relations. New firms in textiles included large power-loom factories, garment export factories and numerous spinning mills – all a product of the reforms. In very few such firms was there a presence of the national trade unions, even any pan-industry union at all. Many firms in cotton spinning started up in the countryside of south India, outside labour agglomerations. The labour component came from impoverished rural regions and included women. The workers entered jobs that were easily available, but tended to be highly insecure. As one would expect, therefore, the effect of market opening has been mixed. In general, the market reforms led to sharp acceleration in employment in labourintensive industries, while leading to a deceleration in wages. Export opportunities played a role in the expansion, whereas internal labour mobility and access to informal and female labour accounted for the slowdown in wages. Stagnation in real wages, together with a rise in productivity, implies increase in profitability. Lured by these prospects, in these highly labour-intensive and low-investment firms, an estimated one million jobs were created via new investment in the course of an export boom in the period 1985–1996, nearly all of it in the informal sector. Individually, many of these new firms are vulnerable, but so are the workers that these firms hire. While inequality between capital and labour was perhaps again rising, inequality between men and women workers seemed to narrow somewhat in the last quarter of the twentieth century. The proportion of women in the manufacturing workforce began to rise again from the 1970s, after a decline that lasted nearly a century. The rise cannot be attributed to a revival of the family household as a work unit, which continued to decline, but rather to expansion of firms that demanded women as wage workers. A great many of the export firms employed women workers. At the same time, the return of women to paid work involved a cultural change enabling greater work participation outside the home. The cultural transition had begun before the present globalization. The handloom industry of south India furnishes examples of women adapting to the wage labour market to a much greater degree than elsewhere. Already in interwar south India, in towns that were growing rapidly as urban weaving clusters, more women started weaving for wages. In Salem town about 1950, a significant percentage of women weavers could be found in small handloom factories. Today, women are common in handloom factories, and in Salem itself, export-oriented garment units employ a large number of young women workers. 272

The long globalization and textile producers in India

Conclusion: research issues In the broad sense of a significant participation in world trade, the Indian textile industry has been a ‘globalized’ industry in the last 350 years. The early nineteenth century and again in the period 1950–1985, when the level of participation in the international economy receded, would appear as brief interludes rather than turning points. The global world of the present time, of course, is a vastly different one from that of the late eighteenth century. Capital, for example, is far more mobile now than before, and therefore textile manufacturing more dispersed, and more competitive. And yet, there are significant similarities too. One of the lessons of this survey is that the dominance of the small firm can be seen as a constant feature across time. If we take a long view, the development of the large firm would seem to be an anomaly, remarkably long-lived historically, but eventually revealing itself to be a special case rather than the norm. The post-independence Textile Policy and the return to competition aided the recent decline of the large firm, but structural changes did not favour it either. There are few economies of scale in textile production. Craftsmanship and knowledge of designs, key resources in flexible differentiation, are often personalized resources and do not require a lot of capital to be deployed. While the large firm may have been weakening due to intervention, the small firm has strengthened via an endogenous drive to enhance efficiency and productivity. Future research needs to deal with this slow, decisive and many-sided transition in the small firm and its constituents – from family to wage labour, women to men and back to women again, settled to migrant, personalized recruitment to market-based hiring and contractual employment, personalized knowledge to market purchased skills, informal to formal training, simple tools to complex tools and so on. The related transformation of the labour market has not necessarily involved a drastic move from the home to the mill shop floor. More often, it has involved a rather graduated shift from the home to the small handloom/powerloom workshop. Yet, it invariably means significant and sometimes permanent adjustments in labour-time allocation by the artisan household. A second running theme has been the effect of this subtle transition on inequality. Clearly, not all participants could access opportunities or bear costs equally. Exclusions and differentiations were built into this process. Any account of a commercializing labour pool shows that some regions and individuals adapt better to wage work than others. The distinction between the more mercantile more capitalist elements on the one hand, and those leaning towards labour on the other, is an important one in understanding the nature of the emerging inequality patterns. It is equally important to understand family decisions to allocate and withdraw labour, to migrate or not to migrate. Along with capital and family, a third axis of inequality formation was differential political capacity of the textile producer. The mill workers have long been divided by where they worked, and by the casual–permanent or outsider–insider status. Mills in the older metropolises have been sites of popular political movements, in which labour has played a key role. Workers in the more rural ‘up-country’ mills 273

The Ashgate Companion to the History of Textile Workers have been far less politically active. The competition between the city and the countryside in large-scale textile production became intense in the interwar period, and the differential political capacity of the two worlds of workers mattered to how each fared in this competition and how it affected wages and employment. A similar differentiation again emerged at the end of the century, as the incumbent and insider workers’ attempts to form political lobbies were constantly undercut by new firms that accessed the ubiquitous ‘reserve army’. In the artisan sector, and possibly in the mills as well, real wages have not increased significantly in the last 150 years, despite episodes of rising output and incomes. There is evidence from time to time of a rise in non-wage incomes, an increasing share of which consisted of merchant profits. By and large, growth has taken the form of sudden outbursts of jobs, and not that of spurts in wages. At the other end, the potentials for such expansion were due to the relatively low capital needs of the small firm, and low barriers to entry. That very fact, however, constrained profitability too, explaining why examples of artisan origins of big business remain scarce in a region in which small firms survive to an extraordinary extent.

274

11 The Italian textile industry, 1600–2000: labour, sectors and products Giovanni Luigi Fontana, Walter Panciera and Giorgio Riello

According to the Austrian statesman Metternich, ‘Italy is a geographic expression, and you will find it only on paper.’ It is difficult to talk about Italy in its entirety, especially for the period preceding the unification of the country in 1861. Therefore, as a matter of necessity, our analysis will concentrate on the north of the country and will draw mostly on evidence from localities in its north-eastern regions. Our analysis does not aim to be comprehensive or to survey a wide historiography, which would stretch well beyond the remit of the project as a whole. We aim instead to provide some insights into the relationship between labour (a wide label that includes the social and economic organization of production, human capital, gender relations, technology and so on) and textile manufacturing for the extended period between the mid-seventeenth century and the present. Special attention will be given to areas of specialization in production, charting in particular the development of specific textile districts over time. Human capital (entrepreneurship, attitudes to innovation, capital networks and so on) forms a second nucleus of concern, in particular in relation to the specific organizational solutions adopted in textile manufacturing in the last three centuries. The first part of this chapter focuses on the period between the mid-seventeenth and the beginning of the nineteenth century, and concentrates on the reorganization of textile manufacturing following the economic and demographic crisis of the 1630s. The second part examines the period of industrialization and focuses on the specific case of the Lanificio Rossi, one of the most significant examples of textile production in the country and one of the most important testing grounds for a new set of labour relations that emerged in the nineteenth century. Finally, the third part is dedicated to textile production in late nineteenth- and twentieth-century Italy and considers the dynamics of the sector in the difficult interwar period and

Map 11.1 Italy

Map 11.2 Northern Italy

The Italian textile industry the relationship that has emerged since the 1950s between textile and ready-made clothing production.

The pre-industrial period: between crisis and change By the middle of the seventeenth century, the Italian textile industry had long lost the primacy that it had enjoyed over other European countries since the later Middle Ages. In the fifteenth and sixteenth centuries new and dynamic textile centres had emerged in France and England. A century later these countries dominated the international markets for textiles both in northern Europe and in the Mediterranean basin. In Italy, economic decline – intertwined with complex political, social and demographic factors – had negative effects on many cities that had long been engaged in textile production. Historians agree that the early seventeenth-century crisis of the Italian economy caused, for the first time since the early Middle Ages, a real ‘de-industrialization’. The main example of this economic involution was the decline of wool manufacturing in Florence. The quarter of a century following the great plague of 1630–1631 was the most difficult period for the seventeenth-century Italian economy, especially in the north of the country where the plague had hit hardest. The population of towns and cities declined sharply and, although recovery was quick, the rate of urbanization remained at a low 18–19 per cent for the next centuries. It should be noted, however, that this rate was one of the highest in Europe, although its stability was symptomatic of the wider urban crisis that was affecting the entire country. It was evident, for instance, that the medieval guilds, and in particular the rich and powerful textile ones, were experiencing a phase of economic and political decline. Moreover, Italy had lost its central role in the international markets. This dark picture must be contextualized. In most cases decline was relative rather than absolute. Firstly, the poor economic performance of the peninsula was evident only when compared with the exceptional growth in northern Europe, particularly in the Netherlands and England. It has been estimated that real per-capita income in Italy was affected badly by the crisis of the 1630s, but by the end of the seventeenth century it had already returned to pre-1570 levels. Secondly, the production of manufactured goods did not disappear suddenly. This was particularly true for the north of the country. In recent years historians have Carlo Maria Cipolla, ‘Il declino economico dell’Italia’, in Le tre rivoluzioni e altri saggi di storia economica e sociale (Bologna, 1989), pp. 85–104; Paolo Malanima, La fine del primato. Crisi e riconversione nell’Italia del Seicento (Milan, 1998), pp. 7–142; Paolo Malanima, La decadenza di un’economia cittadina: l’industria di Firenze nei secoli XVI–XVIII (Bologna, 1982); Edoardo Demo, L’anima della città. L’industria tessile a Verona e Vicenza (1400–1550) (Milan, 2001), pp. 175–206.  Malanima, La fine del primato, pp. 207–215.  Ibid., pp. 220–223. 

277

The Ashgate Companion to the History of Textile Workers emphasized that prior to industrialization in the nineteenth century, Italy was neither underdeveloped nor solely reliant on agriculture. The expansion of the cultivation of mulberry trees and increase in the production of silk can be seen as a positive feature of the Italian textile economy in the early seventeenth century. The production of silk cocoons and the subsequent operation of reeling the silk thread were instrumental in securing steady incomes for peasant families and good returns on capital for rentiers. The spinning and twisting of silk, and its weaving – mostly local to the area of Como in the northern part of Lombardy – were the highest value-added activities in textile manufacturing and were devoted entirely to the production of merchandise for export. During the latter half of the seventeenth century a significant shift in the economic structure of textile production took place. An increasing share of the labour and capital was directed towards the production of silken textiles, especially in the area between the northern plains and the Alps, from Turin in the north-west to Friuli in the north-east of the country. These new production activities were sometimes located in established textile centres, for example silk yarn was twisted in Bergamo and silk cloth was woven in Venice. More often, however, silk started new protoindustrial traditions, as happened in Bologna, the province of Turin, the city of Como and the Bassano area of the Veneto. This was also the period when the most sophisticated machines in pre-industrial Europe were built, including hydraulic spinning/twisting machines that employed dozens of people in so-called protofactories. During the eighteenth century the production stage of silk throwing was also subject to innovation, especially in Piedmont. Although it remained a totally un-mechanized stage in the production of silk, throwing was increasingly performed in mills rather than in the peasants’ houses. The structural shift towards silk did not mean an absolute decline in woollen manufacturing or in the production and processing of vegetable fibres. The latter, in particular, was subject to profound transformations in response to foreign competition and the restructuring of international exchange networks. It is difficult to provide an overall image of the dynamics of change in textile production in Italy, as local idiosyncrasies and specific conditions make Italy a land of extreme differences. We can, however, identify three basic trends for the two centuries between 1650 and 1850. Firstly, textile production was localized in small towns and  See, for example, Alain Dewerpe, L’industrie aux champs. Essai sur la proto-industrialisation en Italie du Nord (1800–1880) (Rome, 1985).  Luca Molà, Reinhold C. Mueller and Claudio Zanier (eds), La seta in Italia dal Medioevo al Seicento. Dal baco al drappo (Venice, 2000); Giovanni Federico, ‘Per una storia dell’industria serica italiana’, Annali di Storia dell’impresa 4 (1988): 112–130.  Carlo Poni, ‘All’origine del sistema di fabbrica: tecnologia e organizzazione produttiva dei mulini da seta nell’Italia settentrionale (secc. XVII–XVIII)’, Rivista Storica Italiana 88 (1976): 444–496; Flavio Crippa, ‘Il torcitoio circolare da seta: evoluzione, macchine superstiti, restauri’, Quaderni Storici 73 (1990): 169–212.  Francesco Battistini, ‘Origini e fortuna di un’innovazione: la “bacinella alla piemontese” per la trattura della seta (sec. XV–XVIII)’, Nuova Rivista Storica 81 (1997): 19–100; Giuseppe Chicco, La seta in Piemonte, 1650–1800 (Milan, 1995). 278

The Italian textile industry in the countryside. Such a localization was in line with a specialization in mediumto low-quality cloth. Secondly, the organization of labour was rather eclectic, with women performing an increasing share of the production process. And finally, mechanization remained low and confined to specific production stages, especially finishing.

The recovery of woollen manufacturing Today, cities like Prato, Biella, Schio and Valdagno are among the most important centres of woollen manufacturing in Italy and in Europe. Their engagement with textiles started in the period after 1670, when new specializations rapidly emerged from ancient artisan traditions. Schio specialized in an imitation of the woollen cloth produced in Verviers in present-day Belgium and in Wiltshire in England by using cheap local wool. These were the so-called panni a uso estero (cloths as done abroad), medium-quality carded woollens, also called mischi (medleys). At the end of the eighteenth century, the province of Vicenza, and Schio in particular, produced more than 20,000 pieces of such cloth every year (which was equivalent to around 400,000 linear metres). There were 150 firms involved in the industry, employing 12,000 people, half of which were women working at home. The late seventeenth-century growth in woollen textile production in the area of Schio and Valdagno was the antecedent of the largest nineteenth-century textile firm in the country – owned by the Rossi, originally a family of wealthy cattle breeders, who engaged in the sale of raw wool. Prato, another important textile centre, near Florence, developed a specialization in the so-called ‘mezzelane’, low-quality mixed linen and woollen fabrics of popular appeal. The city also produced a limited range of traditional pure woollen cloth and, from the end of the seventeenth century, it specialized in the production of woollen caps all’uso levantino (fez). Only after 1850 did Prato start to use recovered wool (otherwise known as lana meccanica) produced from recycled rags.10 Finally Biella, in Piedmont, specialized in the production of low-quality mixed woollens by relying on a substantially more backward and decentralized production system.11 

Corine Maitte, ‘A chacun sa trame: Prato, Biella et Schio entre XVIIIe et XIXe siècle’, in De la fibre à la fripe. Le textile dans la France méridionale moderne et contemporaine (XVIe– XXe siècle) (Montpellier, 1998), pp. 61–92.  Walter Panciera, I lanifici dell’alto vicentino nel XVIII secolo (Vicenza, 1988); Giovanni Luigi Fontana, ‘L’industria laniera scledense da Niccolò Tron ad Alessandro Rossi’, in Giovanni Luigi Fontana (ed.), Schio e Alessandro Rossi. Imprenditorialità, politica, cultura e paesaggi sociali del secondo Ottocento (2 vols, Rome, 1985–1986), vol. 1, pp. 71–255. 10 Paolo Malanima, ‘Le attività industriali’, in Elena Fasano Guarini (ed.), Prato, Storia di una città, vol. 2: Un microcosmo in movimento (1494–1815) (Florence, 1986), pp. 217–277; Corine Maitte, La trame incertaine. Le monde textile de Prato. XVIIIe–XIXe siècles (Villeneuve d’Ascq, 2001). 11 Franco Ramella, Terra e telai. Sistemi di parentela e manifattura nel biellese dell’Ottocento (Turin, 1984), pp. 27–53; Mauro Ambrosoli, ‘The Market for Textile Industry in 279

The Ashgate Companion to the History of Textile Workers Many other minor centres of production took part in the rebirth of the woollen industry at the end of the seventeenth century. The woollen mills of Val Gandino, in the Bergamo province of Lombardy, were smaller in terms of the number of workmen and total output only to those in the province of Vicenza. They produced cheap imitations of more famous and higher-quality products in a way not dissimilar from present-day Chinese textile production. Another important textile centre was the city of Padua and its immediate hinterland in the Veneto. It specialized in the production of mixed cloths (the droghetti and folladini) and in knitting.12 There were also isolated mills, such as the manufacture in Ormea in Piedmont, which produced cloth imitating English woollens, and other local specializations such as the black wools, known as tarantole, produced in the southern valleys of Maiella in Abruzzo, and the woollen cloth of Matelica in the Marche region, and of Arpino, near Naples.13 By the end of the eighteenth century, the production of woollen cloth in Italy was back at the levels it had been at before the crisis of the 1630s. The qualitative level of production was certainly inferior to that of the previous century, but productivity and competitiveness remained high. The woollen sector was particularly receptive and dynamic, especially in the Veneto region. However, another crisis was not far off, this time caused by the profound political and economic upheavals of the first half of the nineteenth century.

The organization of labour in the woollen mills If there was a common element in the organization of the production of woollen textiles in early modern Italy, it was probably the decentralization of carding and spinning. As mechanical carding and spinning were not adopted, the work had to Eighteenth Century Piedmont: Quality Control and Economic Policy’, Rivista di Storia Economica 16 (2000): 343–363. 12 Walter Panciera, L’arte matrice. I lanifici della Repubblica di Venezia nei secoli XVII e XVIII (Treviso, 1996), pp. 115–133 and 209–234; Domenico Sella, ‘Le attività manifatturiere nelle valli bergamasche’, in Aldo De Maddalena, Marzio A. Romani and Marco Cattini (eds), Storia economica e sociale di Bergamo, Il tempo della Serenissima. III. Un Seicento in controtendenza (Bergamo, 2000), pp. 83–97; Walter Panciera, ‘Il lanificio bergamasco nel XVII secolo: lavoro, consumi e mercato, in De Maddalena, Romani, Cattini (eds), Storia di Bergamo, pp. 99–132. 13 Patrizia Chierici, ‘Le fabbriche di pannilana nel Piemonte d’Antico Regime’, in Giovanni Luigi Fontana and Gérard Gayot (eds), Wool: Products and Markets (18th–20th century) (Padua, 2004), pp. 595–606; Costantino Felice, ‘Tra artigianato e industria. Un profilo di lungo periodo’, in Massimo Costantini and Costantino Felice (eds), Abruzzo. Economia e territorio in una prospettiva storica (Vasto, 1998), pp. 120–156; Sergio Anselmi, ‘L’industria della lana a Matelica’, in Economia e vita sociale di una regione italiana tra Sette e Ottocento (Urbino, 1971); Silvio De Majo, ‘Ascesa e declino della protoindustria in Campania: i casi di Arpino e Cava dei tirreni (XVIII–XIX secolo)’, in Giovanni Luigi Fontana (ed.), Le vie dell’industrializzazione europea. Sistemi a confronto (Bologna, 1997), pp. 1141–1160. 280

The Italian textile industry be done entirely by hand, and employed a vast number of unskilled women and children working from homes often distant from the manufactories. This explains why carding and spinning continued to be as high as 20 per cent of the total cost of the finished cloth, thus making production not only difficult to control but also expensive. In such a production environment, labour could at times be scarce, especially in those areas where a number of different textile enterprises were active within a relatively small territory. Attempts to regulate seasonal labour failed as was the case in the Vicenza district, where a number of firms tried to obtain privileged access to the labour of female spinners in specific communities of the Venetian Republic.14 As in many other parts of Europe, women’s wages – based mostly on piecerates – were far inferior to those of men. Complaints about the seasonal cycle of the labour supply (especially during the summer months) and the embezzlement of raw materials and finished products were frequent throughout the period considered. Historians can glean from the records a certain degree of dissatisfaction from merchants and entrepreneurs about the decentralized organization of production. However, it was the only viable option, as mechanization levels remained low. It was not until 1806 that the government of the Napoleonic Kingdom of Italy encouraged the adoption of machines for the carding and spinning of wool. It was not love at first sight. The Gelmi & Bosio textile manufacturers in Gandino in the Bergamo area received new machinery in 1808, but only started using it five years later and in 1818 they disposed of it by selling it to the Rossi of Schio.15 The final stages of production were far more mechanized and employed specialized artisans in centralized manufactories as early as the seventeenth and eighteenth centuries. This was the case with fulling, through the use of fulling stocks, or mechanical teaselling. In both the teaselling, and in the dyeing of the wool and woollen yarn, the artisans employed were highly skilled men. This was also true for the finishing stages and pressing, which required the use of expensive machinery. In places such as Schio and Prato, finishing was often done by small independent firms. Weaving was the stage of production that most varied the quality of the final product and was affected by the type of loom used. In places such as Schio and Verona, large looms with flying shuttles were adopted to produce good quality cloth during the second half of the eighteenth century. Here workers were employed in socalled ‘proto-factories’ where they could be controlled easily and their productivity constantly monitored. This explains why their salaries were rather high: in the 14 Panciera, I lanifici dell’alto vicentino, pp. 99–100. 15 Giovanni Luigi Fontana, ‘Formazione imprenditoriale all’estero e quadri stranieri nell’innovazione tecnico-produttiva: il caso del lanificio Rossi’, in Carlo G. Lacaita, Angelo Ventura and E. Decleva (eds), Innovazione e modernizzazione in Italia fra Ottocento e Novecento (Milan, 1995), pp. 297–395, at pp. 309–310; Walter Panciera, ‘Imprenditori, tecnici e macchine: l’atteggiamento verso l’innovazione nel Settecento veneto’, in Filiberto Agostini (ed.), L’area alto-adriatica dal riformismo veneziano all’età napoleonica (Venice, 1998), pp. 133–152, at pp. 134–137. 281

The Ashgate Companion to the History of Textile Workers Veneto, for instance, they earned nearly double what a worker employed at home earned. Their social status was such that they often portrayed themselves as small autonomous entrepreneurs.16 This eighteenth-century ‘aristocracy of labour’ was formed only of men, not only because of their perceived gender superiority, but also because the combing of wool yarn required substantial physical strength. At the opposite end of the spectrum, the production of low-quality woollens and mixed cloths was characterized by domestic and household production. Here too, the organization of labour could vary greatly. In the province of Bergamo, for instance, weaving was done by men at home and, in the smaller communities, also by women. In the Prato area two-thirds of all the looms used in the production of mixed cloth were located within the city walls, while pure wool cloth was produced in the nearby valley of Bisenzio. In both cases weaving was done by women, who earned less than the wages paid to male weavers in the Vicenza area.17 Finally, in the Biella area weaving remained an activity performed by peasant families until well into the nineteenth century, especially in the Valle Mosso, an area of poor agrarian potential. Here it was the head of the family who coordinated the household textile production and was often the interface between home and the market. Weaving, organized according to the Kaufsystem or Verlagssystem and controlled directly by merchants, developed in parallel with centralized production during the eighteenth and early nineteenth century. The Fabbrica Reale in Ormea in Piedmont and the privileged manufactories owned by the noble Venetian Niccolò Tron in Schio and Follina are just some examples of centralized production systems. Most stages of production, from the preparation of the raw material to the finishing of the cloth, were done under one roof, even if the spinning and weaving remained entirely un-mechanized. The construction of imposing buildings – some of which are still standing, such as the splendid Lanificio Garbin in Schio, completed in 1828 – followed the need for increasing control over production and labour, rather than any technological innovation.18

Hemp, linen and cotton: between the household and the market The processing of vegetable fibres, particularly hemp, was widespread and linked strictly with agricultural production and for use within rural households. The work carried out within the household system often did not produce any exchange beyond the boundaries of the family or the local community. Domestic spinning and weaving were discontinuous activities that complemented the income generated by working on the land and were undertaken mostly by women and children in the long winter months.19 During the nineteenth century the production of linen for 16 Panciera, L’arte matrice, pp. 313–319. 17 Maitte, La trame, pp. 126–137. 18 See Danilo Gasparini and Walter Panciera (eds), I lanifici di Follina. Economia, società e lavoro tra medioevo ed età contemporanea (Verona, 2000). 19 Bruno Caizzi, Storia dell’industria italiana dal XVIII secolo ai giorni nostri (Turin, 1965), 282

The Italian textile industry domestic consumption was replaced by the external purchase of similar products, in particular those that had been produced industrially. Even before the unification of the country in 1861, most hemp cloth was already being produced in factories and the sector was second in value only to silk.20 By the second half of the eighteenth century, the share of land given to flax cultivation was substantial. This was not only the result of the increasing domestic production of linen; large amounts of flax and hemp were also needed to sustain the increasingly complex and articulated putting-out systems co-ordinated by merchant entrepreneurs. The medieval urban economies were experiencing a crisis and only in a few cities in Lombardy, including Cremona, Monza and Busto Arsizio, was the production of cotton textiles still able to continue.21 Elsewhere, urban textile guilds barely survived, while new ‘proto-industrial’ centres, free from guild regulations and able to exploit domestic labour in the countryside, were the busy centres from which merchants were operating. This was the case in Bologna, Chieri in Piedmont, Pistoia, Prato and Milan. Even in the undeveloped south of the country, some areas experienced a phase of growth, for example Cava dei Tirreni near Naples.22 Lombardy remained the most advanced area in textile manufacturing, especially in the processing of vegetable fibres. Here the first spinning machines were introduced in the 1790s. From the beginning of the eighteenth century to 1813 the largest pre-industrial undertaking, the manufacture owned by Giacomo Linussio, was at the forefront of technological and organizational innovation. At its peak the Linussio business employed more than 1,000 male and female weavers, mostly in villages, hamlets and the mountains of Friuli. At the centre of the system was a manufacture that employed 300 workers in finishing and dyeing 40,000 pieces of linen and hemp cloth per year.23 The success of the Linussio business gave a stimulus to similar undertakings in the area of Carnia in the north-east of the country. From the middle of the eighteenth century, Venice and its immediate hinterland specialized instead in the production of printed cottons and fustians. It is said that this sector employed 10,000 workers: female spinners were scattered all over the Terraferma (the region

pp. 130–132. 20 Paolo Malanima, L’economia italiana. Dalla crescita medievale alla crescita contemporanea (Bologna, 2002), pp. 189–191. 21 Maureen Fennell Mazzaoui, The Italian Cotton Industry in the Later Middle Ages, 1100–1600 (Cambridge, 1981); Sergio Zaninelli, L’industria del cotone in Lombardia dalla fine del Settecento alla unificazione del paese (Turin, 1967). 22 Caizzi, Storia dell’industria, pp. 146–147; Alberto Guenzi, La «fabbrica» delle tele fra città e campagna; gruppi professionali e governo dell’economia a Bologna nel secolo XVIII (Bologna, 1987); Silvio De Majo, ‘Il sistema protoindustriale di Cava dei Tirreni nell’Ottocento’, in Paolo Macry and Angelo Massafra (eds), Fra storia e storiografia. Scritti in onore di Pasquale Villani (Bologna, 1994). 23 Giovanni Ganzer, Jacopo Linussio. Un manager del Settecento (Udine 1986); Luciana Morassi, ‘Iniziative tessili nel Friuli del Settecento’, in Fontana (ed.), Le vie dell’industrializzazione, pp. 597–608. 283

The Ashgate Companion to the History of Textile Workers of the Veneto), while the weaving was done mostly in Venice itself.24 Cotton textile manufacturing was short-lived. By the beginning of the nineteenth century the sector had declined under competition from more advanced European producers. Only in Friuli did the production of cotton textiles emerge again after 1849.25

Factory regulations and neo-corporativism The information that we possess on the condition of labour in the pre-industrial textile sector is too fragmentary to provide general conclusions.26 Exploitation seems to be one of the common features. Girls and boys started working at the age of 6–7 and laboured for up to 14 hours a day. This was in line with the conditions of textile workers in other parts of Europe. In Italy, however, a decline in working conditions took place. If urban textile production guaranteed minimal protection for salaried labour, the expanding proto-industrial system based on decentralized production in the countryside probably created harsher conditions in the sector. It was also a world in which conflict could hardly surface. This explains why in seventeenthand eighteenth-century Italy protests from textile workers were extremely rare. Production was carried out in the countryside and in small centres where paternalism dominated the labour relations between entrepreneurs and workers. Catholicism provided the backbone of local assistance through community-based charity and help.27 Protests and violent reactions surfaced only rarely. The only instances of protest that have been identified by historians occurred in the urban woollen mills. Such protests adopted classic corporative forms that appealed to the civic political and juridical authorities. In 1704, for instance, 600 woollen workers (lanari) from Padua petitioned the city authorities. The combers employed in the workshops of the urban merchants protested against their employers’ attempts to limit the mobility of labour and reduce piecework rates. Fifty years previously, the lanari, unable to form an independent production organization (Arte), had established a religious brotherhood able to act as a legal entity. This is why they managed to convince the city council to intervene in their defence against the threat of declining wages in 1704. Similarly, the weavers in the town of Bassano, 30 miles north of Padua, were able to have their occupation recognized as a guild.28 24 Walter Panciera, ‘L’economia: imprenditoria, corporazioni, lavoro’, in Piero Del Negro and Paolo Preto (eds), Storia di Venezia, VIII, L’ultima fase della Serenissima (Rome, 1998), pp. 517–521. 25 Giovanni Zalin, ‘Tra promozione esogena e iniziativa endogena. I distretti cotonieri del Friuli (1840–1929)’, in Fontana (ed.), Le vie dell’industrializzazione, pp. 621–681. 26 Luigi Dal Pane, Storia del lavoro in Italia. Dagli inizi del secolo XVIII al 1815 (Milan, 19582), pp. 296–303. 27 See Angelo Torre, Il consumo di devozioni. Religione e comunità nelle campagne dell’Ancien Régime (Venice, 1995). 28 Walter Panciera, ‘Padova 1704: “l’antica Unione de’ Poveri Laneri” contro “la ricca Università dell’Arte della Lana”’, Quaderni Storici 87 (1994): 629–653; Francesco Vianello, 284

The Italian textile industry These are rather exceptional cases of a successful defence of working conditions. In the Veneto and Lombardy, especially during the second half of the eighteenth century, merchants won power by imposing their own rules and regulations, often framed in a neo-corporative language. Many towns and cities saw the appearance of new officials entrusted with the control of the quality of goods and the organization of labour. This happened in Verona in 1763, in Bergamo in 1769 and in Schio just 10 years later. Such legislation was applied with discretion as it faced strong opposition within a production system that was characterized by extreme flexibility. Eventually, however, these bodies of legislation were abolished or became redundant. A similar analysis can be made of the so-called ‘factory regulation’ (regolamenti). This was a tool to control free labour that was adopted occasionally in the Veneto during the eighteenth century. The first regolamento was probably applied in 1722, in a large woollen mill in Treviso owned by a Venetian trading company. It was compiled by a certain Lambert de Micheroux, a dyer from Liege, and was a basic system of rules about workmen’s tasks and the pay rates associated with each production stage. The factory regulation adopted in 1750, nearly 30 years later, by the Tron-Stahl manufacture in Follina, near Treviso, was based on a more complex system of nine chapters. These established fines for infringing factory discipline, for drunkenness and for escaping through the factory windows! Other regolamenti followed in subsequent years, including the very harsh proposal in 1784 that established capital punishment for the theft of cloth. This was proposed by the Lancastrian Thomas Bamford, consultant to the Republic of Venice, but was never approved by the Republic’s senate. It is interesting to observe how this and other similar proposals were suggested by technicians and textile experts from northern Europe, but were rarely implemented.29 In Italy there was a preference for ad hoc solutions, mediated normally by paternalism and local settlements, rather than state legislation or any comprehensive attempt to regulate labour as a productive factor within a national manufacturing system.

The early phases of industrialization Historians suggest that the process of industrialization in Italy happened in a series of consecutive ‘waves’, and was concentrated in specific areas of the country, especially northern Italy, where long-standing manufacturing traditions helped economic growth. These were the same areas that had enjoyed wide-ranging European commercial contacts over the previous centuries. One must remember Seta fine e panni grossi. Manifatture e commerci nel vicentino, 1570–1700 (Milan, 2004), pp. 197–226. 29 Danilo Gasparini, ‘“Fortune negotij et artefici” a Treviso in età moderna (sec. XV–XVIII)’, in Doretta Davanzo Poli (ed.), Tessuti antichi (Treviso, 1994), pp. 325–350, at pp. 348–350; Panciera, L’arte matrice, pp. 319–323. 285

The Ashgate Companion to the History of Textile Workers the context of Italy before 1860: the country did not have a ‘national market’ (which actually emerged well after Unification in 1861). Instead, its industries benefited from trade and commercial contacts with the most developed parts of the Europe. The export of raw silk and silk yarn, for instance, acted as a stimulus for the functional specialization of textile manufacturing. Moreover, as Luciano Cafagna underlines, France and Belgium provided much needed ‘cultural exchange (read import market), as well as technical and entrepreneurial skills’ for the Italian textile industries.30 A process that we can call ‘modernization’, especially in the organization of labour, started during the Napoleonic period and the Restoration and was confined to ‘pioneer industries’ located in the ‘core region of Italian industrialization’, the banana-shaped area in between the agrarian plains of northern Italy and the Alps, from Piedmont in the west to Friuli in the east. This was an area of plentiful hydraulic resources and was characterized by a rich proto-industrial tradition as discussed in the previous section. This area was not just a place of proto-industry: it was also rich in mulberry trees and silk worms, two of the key raw materials of the early industrial age. Silk is an ‘industrial crop’ that combines agriculture and manufacturing. It was the perfect sector to absorb the surplus of labour from agriculture. Silk therefore acted as a means for a smooth transition of labour from agriculture to industry, a key mechanism in the ‘wave’-like process of industrialization that characterized Italy in the nineteenth century. The country did not experience the formation of an urban working class, but its workforce maintained strong links with the countryside, living in non-urban areas within traditional socio-economic familial structures. The labour market remained flexible, and revertible in case of necessity, without impacting on established agrarian forms of employment.31 The employment of lowwaged workers was characteristic of the entire nineteenth-century Italian textile sector, but was particularly intense in silk manufacturing where numerous silk mills in Lombardy, Piedmont and the Veneto were owned by the same landowners who cultivated silk worms and planted mulberry trees. Silk was key to the economies of the most advanced regions of Italy (Piedmont and Lombardy), and raw silk, together with high-quality linen, was widely exported.32 Between 1800 and 1853 the production of raw silk tripled in the Lombardo-Veneto region (at the time part of the Austrian Empire). The reliance on foreign markets was accompanied by a dependence on foreign capital.33 It was the cash advances made by French and English silk merchants that financed the 30 Luciano Cafagna, Dualismo e sviluppo nella storia d’Italia (Venice, 1989), pp. 359–360. 31 Ibid., pp. 361–367; Paul R. Corner, Contadini e industrializzazione. Società rurale e impresa in Italia dal 1840 al 1940 (Rome and Bari, 1993). 32 Sergio Zaninelli, Storia dell’industria lombarda. Vol. I: Dal Settecento all’unità politica (Milan, 1988); Valerio Castronovo, Economia e società in Piemonte dall’Unità al 1914 (Milan, 1969); Roberto Tremelloni, L’industria tessile italiana (Turin, 1937), pp. 30–31. 33 Giovanni Federico, Il filo d’oro. L’industria mondiale della seta dalla restaurazione alla grande crisi (Venice, 1994). 286

The Italian textile industry employment of large numbers of women and children from the countryside and the occasional specialized skilled workman.

The mechanization and the organization of labour in the cotton sector The cotton industry also employed a cheap rural workforce. Cotton, however, was not linked to the agrarian production of the raw material. Its manufacture developed particularly strongly in areas that had land of poor agrarian value but were rich in hydraulic resources. The industry was located in the valleys and mountain plains of Lombardy, in particular in the province of Varese, in Gallarate, Busto Arsizio and Legnano, and in the valleys of Bergamo and Brescia. In 1860 the provinces of Milan and Como alone had 80 per cent of all the cotton spindles in Italy. The remaining spindles were located across Piedmont, particularly in Chieri, in the area between Alessandria and the borders of Liguria and Novara, and around the Lago Maggiore. In contrast to silk manufacturing, it was so-called ‘merchant capital’ that was key in the development of the Italian cotton industry. During the Napoleonic period foreign technicians had united with Lombard and Piedmontese merchant entrepreneurs to create the first mechanized factories for the production of fustians and other cotton textiles.34 Four cotton mills erected in Olona (Borgomaneri, Martin and Schoch, Ponti, and Eraldo Krumm) ‘experimented with new forms of the organization of production that incorporated continuities (of the knowledge of products, the use of hand-operated jennies, and the traditional web of merchant entrepreneurs) but also discontinuities (factories, mechanization, the discipline of labour, inorganic power and complex forms of organization)’.35 Notwithstanding the permanence of putting-out, the cotton sector was characterized already in the first half of the nineteenth century by a high degree of centralization of the workforce as well as by complex financial forms among which was shareholding. The Manufacture Annecy and Pont in Piedmont was established in 1829 and already in 1845 used 410 mechanical looms.36 It integrated spinning, weaving and finishing under the same roof and gave work to 1,600 men and women.37 Lombardy was the most advanced region: Gian Giacomo Muller introduced in his cotton mill in Intra the first mechanical spinning machine in 1808. In 1816 the Count Porro Lambertenghi introduced steam-powered machines in his cotton mill in Como. During the first decades of the nineteenth century some of the most important dynasties in Italian capitalism emerged in the cotton

34 Valerio Castronovo, L’industria italiana dall’Ottocento a oggi (Milan, 1980), pp. 11–13. 35 Silvia A. Conca Messina, Cotone e imprese. Commerci, credito e tecnologie nell’età dei mercanti-industriali. Valle Olona 1815–1860 (Venice, 2004), pp. 230–231. 36 Antonio Fossati, Lavoro e produzione in Italia dalla metà del secolo XVIII alla seconda guerra mondiale (Turin, 1951), p. 96. 37 Castronovo, L’industria cotoniera in Piemonte (Turin, 1965), pp. 63–69. 287

The Ashgate Companion to the History of Textile Workers sector: Ponti, Borghi and Krumm, Cantoni, Vanzina, Crivelli, Airoldi and Turati.38 Between 1839 and 1840 the above-mentioned firms in Olona decided to replace all their machinery. Just a dozen years later they had developed into a modern industry that employed technology similar to that adopted in Alsace. The absence of a national mechanical industry did not therefore preclude the possibility of adopting the newest technology available. This happened thanks to the presence of several foreign entrepreneurs, especially Swiss and German. They were active not just in the north of Italy but also in the south of the country: Schlaefer Wenner & Co., Davide Wonwiller, Escher and Meyer had factories in Campania.39 What Messina defines as the ‘historical moment key to the formation of the criteria necessary to the success of the factory system’ both in the cotton and wool textile sectors, is to be found in the period between 1860 and 1880, during the first 20 years of the newly unified Italian state. ‘Training through the production process (learning by doing), the making of the first pieces of machinery, and a factory organization centred around mechanical workshops allowed the cotton textile industry to grow, even in a context of technological dependence’.40 By the end of the nineteenth century, the mechanization of the sector had reached sufficient levels to allow for the vertical integration of many of the production processes. The concentration of production in factories changed the conditions of labour. The domestic system of production declined, but ‘although the 50 years between 1810 and 1860 were characterized by a dissociation from handicraft production, and conversely by the adoption of increasingly complex machinery, some of the phases of production within the industrial system remained heavily dependent on the skills of spinners, their agility and dexterity in handling the yarn, moving the shuttle … their practical deftness did not disappear, even in the major factories’.41 The organization of labour, however, gradually assumed stricter and more indispensable rules. The directors and technical staff were responsible for the control of technologies and people, the planning of the time schedules, the organization of labour and the attribution of specific responsibilities (carpenters and mechanics, for instance, were key figures in such organizations). Internal hierarchies grew in a more structured way compared to the past. The limited availability of a skilled workforce in Italy made it necessary to recruit from abroad. It became essential to be able to select personnel and balance the discipline of labour on the factory floor with an endemic scarcity of both technical expertise and a skilled workforce. The majority of the workforce comprised women and children who worked up to 15 hours a day in dusty and unhealthy environments. The conditions of work in 38 Fossati, Lavoro e produzione in Italia, p. 98; Zaninelli, L’industria del cotone in Lombardia; Roberto Romano, La modernizzazione periferica. L’Alto Milanese e la formazione di una società industriale 1750–1914 (Milan, 1990). 39 Fossati, Lavoro e produzione in Italia, p. 98; John Davis, Società e imprenditori nel Regno borbonico 1815–1860 (Bari, 1979); Luigi De Rosa, La rivoluzione industriale in Italia e il Mezzogiorno (Bari, 1973). 40 Conca Messina, Cotone e imprese, p. 247. 41 Ibid., p. 243. 288

The Italian textile industry the cotton mills, as was the case in Britain, became the subject of much debate, and public denunciations. The terrible conditions of children’s work and the health of the industrial workforce became contentious issues.42 Debates closely followed the example of the public discussions in France and Britain. Giuseppe Sacchi reported in the Journal Annali Universali di Statistica in 1842 that in the province of Milan, around 500 children less than 12 years old were made to work between 12 and 15 hours a day for a salary as little as a quarter to half a lira. More than half of the workforce employed in these mills was aged between 9 and 14 years old.43 The division of labour between men, women and children did not follow uniform criteria. Men were normally employed in the preliminary stages of cleaning and preparation of the raw cotton; they could be given responsibility for single machines and they were always in charge of spinning high-count yarns. Women were employed in the spinning of low-count yarns and children helped adults in several different tasks, including the cleaning of machinery (often while moving) and the repair of broken yarn.44 The cotton industry was at variance with other textile sectors because of its high degree of concentration of the workforce, its use of complex technology and the presence of an elite of thoroughly trained workmen. The sector also had the most complex financial structure, with companies offering shares, but not yet adopting limited liability. The most important early cotton mills were owned by what became some of Italy’s best known nineteenth-century entrepreneurial dynasties such as the Cantoni family in Lombardy. Some of the entrepreneurs in the cotton sectors (even in the south of Italy) were Swiss and German.45

The factory system and the woollen textile sector At the beginning of the nineteenth century the rural workforce employed in the woollen textile sector remained large. Even though proto-factories were not uncommon, they operated within extensive systems of subcontracting and domestic work.46 In some cases production was coordinated not by merchant entrepreneurs but by institutions of different kinds. This was the case in Rome and the surrounding region of Lazio where several pii istituti (associations for poor relief) existed, such as the hospice of S. Michele that gave work to more than 1,000 workers in the production of tapestries. Other parts of Italy, especially towns such as Fabriano, 42 Sergio Zaninelli, ‘Il lavoro dei fanciulli nell’industria lombarda alla metà del secolo XIX’, in Alberto Cova (ed.), L’economia lombarda nella prima metà dell’Ottocento. Alcuni temi di ricerca (Milan, 1981); Zaninelli, L’industria del cotone in Lombardia and Storia dell’industria lombarda, vol. I. 43 Giuseppe Sacchi, ‘Sullo stato dei fanciulli impiegati nelle manifatture’, Annali universali di statistica 71 (1842): 186–204. 44 Conca Messina, Cotone e imprese, pp. 244–247. 45 Castronovo, L’industria italiana, p. 13. 46 Ibid., pp. 9–10. 289

The Ashgate Companion to the History of Textile Workers Alatri, Foligno, Spoleto and Matelica in Umbria and the Marche, specialized in products that could have local, national or even international markets.47 The wool industry was well protected by a high tariff system and only sporadically showed signs of dynamism in a context that was dominated by traditional production and obsolete working practices that were mostly performed by hand. Due to the agrarian workforce, salaries remained extremely low. Some small improvements appeared in the Napoleonic period. The self-acting spinning machine (for cotton) was used in Italy for the first time in 1808 by Gian Giacomo Muller at his factory at Intra and in 1815 four machines were purchased for the manufacturing of wool by the Ministry of Interiors in Milan and sold to local industrialists.48 The partial unification of the country under the Napoleonic regime allowed the creation of a larger market with fewer import and export duties. The results were mixed. On the one hand (for Lombardy and the Veneto especially) this was a unique opportunity to produce for new markets and military commissions in a context of legal uniformity across what had previously been different states. On the other hand, however, Italy suffered from the Continental blockade of 1805, and also had to compete against French cotton and wool textiles, especially after the signing of the disadvantageous treaty of commerce with France in 1808 that inflicted a severe blow on the export of silk to France.49 The Restoration period saw attempts at the mechanization of production. The area of Biella in Piedmont emerged as a leader in the production of heavy woollen cloth in Italy. More modest results were achieved in other areas that had been heavily dependent on military commissions and this, combined with the beginning of the Austrian domination, saw their secure market disappear overnight. This was particularly the case for the wool mills in Ormea and Mondovì, for several mills in the area of Como and Gandino and for Schio in the Veneto. In Tuscany, the major tradition of Florentine wool manufacturing survived in Prato, due to the making of red Levantine berets that had been started in 1788 by Vincenzo Mazzoni. During the nineteenth century Prato became the third most important centre for wool textile manufacturing after Biella and Schio.50 The aforementioned Prato, Biella and Schio were, in the 1820s, the three areas that aspired to compete with the well-developed European centres of wool production. Particularly relevant to their development were two entrepreneurial families: the Sella in Biella and the Rossi in Schio.51 Pietro Sella (1784–1835) was a member of a family of industrialists, scientists, politicians and philanthropists. 47 Roberto Dodi, Del lanificio in Italia e all’estero (Rome and Biella, 1943), pp. 183–185. 48 Ibid., p. 187. 49 Tremelloni, L’industria tessile, pp. 29–30. 50 Bruno Caizzi, Storia dell’industria italiana dal XVIII secolo ai nostri giorni (Turin, 1965). 51 Valerio Castronovo, L’industria laniera in Piemonte nel secolo XIX (Turin 1964), pp. 59–63; Giovanni L. Fontana, ‘Formazione imprenditoriale all’estero’, pp. 297–375; Giovanni L. Fontana, ‘L’Europe de la laine: transfers de techniques, savoir-faire et cultures d’entreprise entre Verviers, Biella et Schio’, in Fontana and Gayot (eds), Wool: Products and Markets, pp. 687–759. 290

The Italian textile industry He had travelled widely in France, Belgium and England, before buying a series of machines for cleaning, carding and spinning wool and the finishing of woollen textiles in 1816. He created a true ‘mechanical revolution’ in his wool mill in Strona in Piedmont and had to face conflicts with the Sardinian government who regarded his innovations with deep concern.52 In the 1830s the Lanificio Fratelli Sella at Croce Mosso gave work to more than 1,000 workers both in the main production plant and in the surrounding countryside.53 Pietro Sella’s nephew Venanzio developed the use of chemistry in wool dyeing and during the 1840s the Sella mill specialized in the production of high-quality products thanks to the use of mule-jennies. In 1841 there were 17 cotton mills in Biella, all of which were highly mechanized in spinning: they had 816 looms and employed 5,329 workers. Mechanization in weaving arrived only in the 1870s. Francesco Rossi (1782–1845), together with his business partner Eleonoro Pasini, founded the Lanificio Rossi in 1817–1819. His son Alessandro (1819–1898) developed the business to make it one of the most important firms in Italy in the nineteenth century. Francesco Rossi was keen to adopt the latest technology. In 1811 he had already enquired about carding and spinning machines from the mechanical firm of James and John Collier in Paris. We do not know how advanced the mechanization of spinning or carding was in other local firms, such as at the wool mill of G.B. Garbin in Schio which at the time was the most advanced in the area.54 However, Francesco Rossi eventually bought machinery from Gandino where the first carding and spinning machines had been introduced in 1820. In the same year G.B. Mazzoni in Prato started the mechanization of both cotton and wool manufacturing. Notwithstanding the partial mechanization of the industry, by the mid-nineteenth century centralization of production was just starting, and was already encountering opposition from workers, especially weavers. Production was far too varying to allow general mechanization and the resources that each mill could draw upon were based on familial links. The low cost of labour (25 per cent less than in other European countries) was not sufficient to compete in international markets and therefore protectionism was seen as the only defence.55 The Lanificio Rossi was something of an exception: when Alessandro Rossi succeeded his father in 1842, he quickly modernized and expanded all the production plants, devised new products and restructured the organization of production and distribution. In 1855 the Lanificio was a fully integrated production structure. In 1862 Rossi erected the largest industrial plant in nineteenth-century Italy, the so-called ‘Fabbrica Alta’ (Tall Mill), a six-storey building designed by the 52 Castronovo, L’industria laniera; Raffaella Gobbo, ‘The Transfer of Knowledge between Verviers and Biella: Based on Documents taken from the Files of the Sella Wool Mill in Croce Mosso’, in Fontana and Gayot (eds), Wool: Products and Markets, pp. 747–761. 53 Castronovo, L’industria laniera. 54 Giovanni Luigi Fontana and R. Marchesini, ‘La città della lana. Storia per immagini’, in Fontana (ed.), Schio e Alessandro Rossi, vol. 2, pp. 821–1424, fig. 51. 55 Castronovo, L’industria italiana, p. 11. 291

The Ashgate Companion to the History of Textile Workers architect August Vivroux from Verviers. At the time of the International Exhibition in London in 1872, the Lanificio employed 800 workers (paid between 0.5 and 2 liras per day), had 200 looms and an annual turnover of 2 million liras, allowing profits amounting to 750,000 liras a year.56 The Lanificio Rossi became an example to follow for other firms such as the Lanificio Marzotto in Valdagno (which later became one of the major European woollen producers), and for a group of smaller firms such as P. Cazzola, G.B. Conte and A. Ferrarin.57

Alessandro Rossi and nineteenth-century paternalism Rossi was not just ‘the most perfect type of Italian industrialist’.58 His vision of the sort of industrialization that Italy had to embrace and the way in which labour relations had to be formulated had a lasting impact throughout the Italian textile sector and beyond.59 His overall understanding of labour and labour relations discarded state intervention in favour of a paternalistic form of entrepreneurial behaviour, leaving room for capital and management to create a relationship with the workforce suited to idiosyncratic local conditions. He opposed, for instance, direct intervention of the state in the regulation of working hours and the work of women and children.60 It was the moral duty of the entrepreneur, rather than the imposition of the state, he believed, to provide the organizing principle of development. He argued in his Appello agli Industriali Italiani (Letter to the Italian Industrialists) of 1867 that ‘the firm has not only to be a source of profit and employment’, but had to be seen by all parties as ‘a common spring of culture and development’.61 At the centre of such a philosophy of labour was a system of facilities and services to workers, provided by the firm itself, that were ‘completed to produce material 56 Giovanni Luigi Fontana, Mercanti, pionieri e capitani d’industria. Imprenditori e imprese nel Vicentino tra ‘700 e ‘900 (Vicenza, 1993), pp. 41–75. 57 Giorgio Roverato, Una casa industriale.i Marzotto (Milano, 1986); Fontana, Mercanti, pionieri e capitani d’industria, pp. 111–322. For a wider analysis of Rossi and his vision of industrialization and capital, see Giovanni Luigi Fontana and Giorgio Riello, ‘Seamless Industrialisation: The Lanificio Rossi and the Modernisation of the Wool Textile Industry in Nineteenth-century Italy’, Textile History 36 (2) (2005): 170–197. 58 Quoted in Cafagna, Dualismo e sviluppo, p. 225. See Fontana (ed.), Schio e Alessandro Rossi; Piero Bairati, ‘Studi su Alessandro Rossi’, Rivista Storica Italiana 100 (1) (1988): 212–219. 59 Bairati, ‘Studi su Alessandro Rossi’. Rossi was also the author of 250 books, pamphlets and speeches composed between 1865 and his death in 1898. 60 Marco Doria, ‘Gli imprenditori tra vincoli strutturali e nuove opportunità’, in Franco Amatori, Duccio Bigazzi, Renato Giannetti and Luciano Segreto (eds), Storia d’Italia: annali Einaudi, vol. XV: L’industria (Turin, 1999), p. 646; Guido Baglioni, L’ideologia della borghesia industriale nell’Italia liberale (Turin, 1974), p. 143. 61 Alessandro Rossi, L’industria italiana nei suoi rapporti coll’esposizione internazionale di Parigi. Lettera (Florence, 1867). See also Mariella Berra, L’etica del Lavoro nella Cultura Italiana dall’Unità a Giolitti (Milan, 1981), pp. 40–44. 292

The Italian textile industry and moral improvement for all workers’ to which Alessandro Rossi personally contributed half of his salary as chief director of the company. The initial activities organized by Rossi were financially limited and included prizes to workers (1858– 1859), a friendly society (1861), a short-loan system for rents (1864), an evening school for illiterate workers (1866), a nursery school (1867), a block of flats for workers (1865–1870), the Jacquard Theatre and Garden (1869) and numerous social clubs.62 With the restructuring of the family firms into a major Italian joint-stock company in the early 1870s, the intervention became more radical, innovative and financially consistent with a total expenditure of over half a million liras over the decade between 1873 and 1883, a compelling example from the rich tradition of European industrial paternalism. This period saw the completion of the ‘Nuovo Quartiere’ in Schio, workmen’s houses, a dormitory for female workers in Pieve, a new nursery school building, a primary school, an industrial school in Vicenza and a school of agronomy and pomology. By the time of the ‘workmen’s jubilee’ to celebrate Alessandro Rossi’s seventieth birthday in 1889, the firm had 21 clubs as well as circulating libraries, cooperatives, common kitchens and ovens, and a company band.63 Rossi was a successful entrepreneur in bringing together business logic and social engineering. He strongly believed in the maintenance of an ‘organic link’ with rural life that had characterized the early stages of Italian industrialization. Such a link was not found in the classic employment patterns of the textile worker who was at times an agricultural labourer, but in the preservation of small (and manageable) communities.64 Rossi was well aware of the vast literature on the English worker who was reduced to living in destitute conditions and existing in squalid and miserable circumstances in anonymous towns. Like many fellow visitors to Britain in the early 1840s, Rossi had been struck by the evils of urbanization.65 The traditional way of living that secured social stability, respect for institutions 62 Fontana and Marchesini, ‘La città della lana’, figs 469–523. 63 Franco Mancuso, ‘Schio, “Nuova Schio” e Alessandro Rossi’, Storia Urbana II (1977): 46–98; E. and L. Mariani Travi, Il paesaggio italiano della rivoluzione industriale: Crespi d’Adda e Schio (Bari, 1979); B. Ricatti, Antonio Caregaro Negrin, Un architetto vicentino tra eclettismo e liberty (Padua 1980); Franco Barbieri, ‘La “Nuova Schio” di Alessandro Rossi’, in Roberto Gabetti, Carlo Olmo et al., Villaggi operai in Italia. La Val Padana e Crespi d’Adda (Turin, 1981), pp. 229–248; R. Marchesini, ‘Le società di Alessandro Rossi e il “Nuovo Quartiere”’, in Fontana (ed.), Schio e Alessandro Rossi, vol. 1, pp. 319–357; B. Ricatti Tavone, ‘Antonio Caregaro Negrin architetto-urbanista di A. Rossi’, in Fontana (ed.), Schio e Alessandro Rossi, vol. 1, pp. 687–730; Franco Barbieri, ‘Dal primo al secondo progetto della ‘Nuova Schio’. Verso Crespi d’Adda’, in Fontana (ed.), Schio e Alessandro Rossi, vol. 1, pp. 731–744; Fontana and Marchesini, ‘La città della lana’, figs 539–606; Giovanni Luigi Fontana, ‘Schio, “Nuova Schio” e il Lanificio Rossi: costruzione e riuso di un caso esemplare’, Annali di storia dell’impresa 13 (2002): 153–187. 64 Gualberto Gualerni, Economia aperta. Un approccio storico all’economia e politica industriale in Italia: 1860–1996 (Turin, 1999), p. 17. 65 Fontana, ‘Formazione imprenditoriale all’estero’, pp. 363–375. 293

The Ashgate Companion to the History of Textile Workers and personal industriousness did not need to be suppressed by the new ‘industrial society’. Technology, industrial development and the economic modernization of his firm – and, by extension, of the entire nation – could be achieved without undermining the cultural and social position of the worker. By the early 1870s the influence of the Lanificio on the life of Schio and its hinterland was becoming inescapable. The town had to be transformed to accommodate the new industrial needs and future plans for development, before chaos took charge of the urban structure. The entrepreneur was responsible to the community as a whole for providing facilities, services and housing, thus taking the place of modern state intervention. A good deal of paternalism was mixed with a conservative and rather utopian idea of social relations in the industrializing societies.66 However, international – rather than local – concerns informed Rossi’s philosophy of intervention. His understanding of the social problems faced by small and large industrial towns in Belgium, England, France, Germany, Spain and the United States was both direct (through travels and personal acquaintances) and indirect (through reading, statistical information and several journeys undertaken by his sons and his personal secretary Egidio Rossi).67 Saltaire in England, or the worker’s villages built in Italy by Crespi and Leumann were well-known examples of solutions to social and urban ‘emergencies’. However, these foreign and domestic experiences were neither uncritically accepted nor dogmatically refused. In Rossi’s view, industrial ‘villages’ were not to be developed as separate entities, but had to grow within existing urban structures. Rather than planning a new ‘modern’ industrial town, Rossi was keen to preserve the traditional social and cultural equilibrium of a rural and artisan world by connecting the new and the old. He did so by accepting the help of entrepreneurs, architects and technicians with whom he shared similar concerns and sensibilities. Again the European influence was at work. The Nuova Schio was not only the result of social engineering but also a way to promote the firm. Partial plans were shown at the Société Industrielle de Verviers and at the Milan Exhibition in 1872. Prints similar to those published by Titus Salt were sent to all major European entrepreneurs, chambers of commerce and international industrial societies.68 Alessandro Rossi’s experience in Schio has to be considered as more than just a ‘social laboratory’. The structures, societies and good wages of his workers were clearly part of a paternalist view of society and industrial development.69 66 Valerio Castronovo, La storia economica d’Italia. Dall’Ottocento ai giorni nostri (Turin, 1995), pp. 56–57; Castronovo, L’industria italiana, p. 40. 67 Giovanni Luigi Fontana, ‘Dar casa agli operai. Logiche d’impresa e ingegneria sociale nell’industrializzazione moderna’, in Carolina Lussana (ed.), Dalmine dall’impresa alla città. Committenza industriale e architettura (Dalmine, 2003), pp. 23–54. 68 Archivio Lanificio Rossi, b. 8: ‘Letter from Robert Centner to Alessandro Rossi’, 15 December 1870; 12 February 1871; 15 May 1871. 69 See Silvio Lanaro, ‘Genealogia di un modello’, in Silvio Lanaro (ed.), Le regioni dall’unità a oggi. Il Veneto (Turin, 1984), pp. 3–96. 294

The Italian textile industry However they also existed as the final embodiment of the ideas that Rossi relentlessly advertised to the intellectual, financial and political elite in the Veneto, in Lombardy and also in Parliament in Rome.70 The success of his creation was clear evidence that Italy’s industrialization could respect the rights of workers, increasing their standards of living without destroying their social environment. The Lanificio was proof that the process of industrialization was not necessarily the carrier of social evils, moral and physical degradation and industrial conflict that had been the result of English industrialization. Rossi rejected the English model of industrialization, which produced ‘horrible and destitute plebs, without any dignity and any human sentiment’, in favour of the American model of industrial development, in which ethical and moral elements were combined with a new role for industrial undertakings: industry was a solution to poverty and could instil industriousness in the workforce.71 As Valerio Castronovo has underlined, the bourgeois virtues of self-help and reliance on individual merit were linked to the more Catholic principles of a structured community based on loyalty but also on hierarchy.72

The textile industry in the late nineteenth and twentieth century Vittorio Ellena was the author of the first Italian industrial statistical survey in 1876. His investigation was rather biased and excluded important industries including artisan activities such as clothing production. Textile production, however, featured prominently in this early survey. It revealed that out of the 380,000 workers employed in textile manufacturing only 103,000 were men, with women and children accounting for 188,000 and 90,000 respectively. Silk was still the dominant textile industry and it employed 200,000 people, more than half of the entire workforce in the industry, although only 15,000 of them were men. Ellena’s survey also revealed the importance of textile manufacturing within Italian industrial production as a whole.73 The subsequent survey carried out in 1881 showed that out of 3.3 million workers employed in manufacturing, nearly 1.4 million belonged to the textile sector and a further 1.1 million were employed in the more artisan-based clothing production. The statistics also showed that powerlooms favoured the employment of unskilled women and children, but that the use of machines was confined to some of the largest firms, while most of the industrial 70 Luciano Segreto, ‘Storia d’Italia e storia dell’industria’, in Amatori, Bigazzi, Giannetti and Segreto (eds), Storia d’Italia, p. 13. 71 Quoted in Baglioni, Ideologia, p. 141. 72 Valerio Castronovo, Grandi e piccoli borghesi. La via italiana al capitalismo (Rome and Bari, 1988), p. 5. 73 Stefano Musso, Storia del lavoro in Italia dall’Unità a oggi (Venice, 2002), pp. 27–28. 295

The Ashgate Companion to the History of Textile Workers system was reliant on domestic subcontracting.74 It was only during the last two decades of the nineteenth century that domestic manufacturing declined, with a reduction of more than half in the number of people employed, especially in the south of Italy.75 These early industrial statistical surveys also reveal the emergence of strong areas of industrialization where a new working-class culture was taking shape. The vertical and horizontal integration of production resulted in the success of a smaller number of larger firms. The aforementioned Lanificio Rossi in Schio employed more than 5,000 workers in the 1880s; the nearby Lanificio Marzotto in Valdagno reached 2,000 employees in 1900 and the cotton mills owned by the Crespi family had 1,300 workers at the turn of the century.76 Such a restructuring of the sector also coincided with a time of intense economic growth for the country, the so-called ‘first industrial miracle’ from 1910 to 1914.77 The share of the workforce employed in textile manufacturing decreased for the first time as other sectors developed. The number of women and children employed in textile production also decreased as labour became more protected and domestic industries generally declined. Working hours in the 1860s had reached peaks of 14 hours or more a day. During the following decades they decreased, mostly as a consequence of a shift from seasonal to full-time work. However, in the 1900s it was still not uncommon to work 12 hours a day with frequent night shifts. Wages were still low (a weakness that was associated with the relatively high employment of women) compared to other sectors, despite productivity rising dramatically. The total workforce increased by only 25 per cent between 1867 and 1900, while the overall production of yarn and cloth increased fourfold.78 In 1876, 650 cotton mills provided work for 54,000 men, women and children. In 1900 the number of cotton mills had only slightly increased to 727 but the number of workers had more than doubled to 135,000. In the same period the number of cotton looms increased threefold (from 0.7 to 2.1 million) and the value of manufactures increased sixfold (from 50 to 300 million liras) notwithstanding the drastic reduction in prices. Production per worker increased from 940 to 2,250 liras per year. The cotton sector expanded in the 1890s (forming a National Association in 1893), but was badly affected by a production crisis in 1907. The industry had developed rapidly due to low wages, following the example of other textile sectors, but was unable to compete successfully with foreign firms, especially the English mills. Productivity still remained low because of the lack of specialization in the

74 Ibid., p. 30. 75 Ibid., pp. 29–30. 76 Fontana (ed.), Schio e Alessandro Rossi; Roverato, Una casa industriale; Roberto Romano, I Crespi. Origini, fortuna e tramonto di una dinastia lombarda (Milan, 1985). 77 Giorgio Mori, ‘L’economia italiana dagli anni Ottanta alla prima guerra mondiale’, in Giorgio Mori (ed.), Storia dell’industria elettrica in Italia, vol. I: Le origini 1882–1914 (Rome-Bari, 1992). 78 Tremelloni, L’industria tessile, pp. 74 and 77. 296

The Italian textile industry workforce, the low efficiency of its female component and endemic problems of distribution, all features shared by the wool textile sector.79 The situation of the silk industry was similarly difficult. The 1887 tariffs exacerbated the problems of an industry troubled by strong price fluctuations and a definite difficulty with the mechanization of production. Silk manufacturing also suffered because of a diversion of capital to the cotton industry that prevented the industry from adopting the latest mechanical devices. The competition of silk products from Asia further impacted on Italian production. However, silk continued to be a key sector of the national economy. In 1906 its exports amounted to 695 million liras, equivalent to more than 36 per cent of all exports.80 However, the 1887 tariff system had positive effects on wool manufacturing. The sector expanded and mechanized with the number of power-looms rising from 2,600 in 1887 to 6,500 in 1894. During the same period the number of handlooms decreased from 6,000 to 3,800. The total value of production increased from 74 million liras in 1866 to 100 million in 1894, to 275 million in 1909 and 350 million in 1913.81 In 1914 the sector was worth half a billion lira and employed 50,000 workers.82 In contrast to silk, wool textiles were mostly destined for internal consumption, although successful attempts were made to sell them in Latin America, India and the Near East. The linen (flax and hemp) industry had to cope with the problem of retting (rotting) the raw material: this was a long and expensive process that could be shortened only through the use of chemical agents. In 1903 there were 309 firms in the linen sector (including jute) employing 25,000 workers, mostly in weaving as the majority of yarn was imported. In contrast, most of the firms in the hemp industry (which was mainly used to produce ropes) specialized in spinning local raw material. The first comprehensive industrial census in 1911 included all firms employing at least one worker within specialized premises. The textile sector was still the most important in the Italian economy with 22 per cent of all workers, followed by the mechanical industry with 16 per cent and the food-producing sector with 13 per cent. Textile factories employed around 20 per cent of all workers in the survey and still included some of the largest firms in the country (together with the rising sugar-processing sector). In the textile industry, 98 per cent of the workers were employed in firms with more than 10 employees. The average size of a textile firm (across all sectors) was 139 employees, an indicator of the definitive establishment of a strong group of large and medium-sized firms that co-ordinated the activities of smaller subcontractors. Development was uneven across the national territory: Piedmont and Lombardy still accounted for three-quarters of all workers in textile

79 Ibid., pp. 82–90. 80 For an overview of the Italian silk sector, see Giovanni Federico, An Economic History of the Silk Industry, 1830–1930 (Cambridge, 1997). 81 Tremelloni, L’industria tessile, pp. 97–98. 82 Raimondo Targetti, Scritti di economia laniera (Rome 1942), pp. 85–86. 297

The Ashgate Companion to the History of Textile Workers manufacturing.83 The area of Schio and Valdagno in the Veneto were exceptions to this geographical pattern, as more than 50 per cent of the firms with more than 50 workers were concentrated in this area.84 However, it was the north-west of the country that was to be the centre of the new industrialization. Factories had moved towards the plains and urban areas with the adoption of electricity. It was here that the first working class appeared. In the decade before the First World War entrepreneurs and industrial workers became the core contributors to political, economic and social debates. More than any other class, they symbolized the climate of change that was sweeping across the country and, in contrasting ways, proposed visions and projects for the reorganization of society at large. The opinions of other social groups, such as the self-employed, the white collar workers and the professional classes, were less well represented in this period. The very figure of the entrepreneur changed profoundly towards the turn of the century. While during the previous decades, in particular in the 1860s and 1870s, large firms had been accompanied by a vast number of smaller entrepreneurs and artisans, in the early twentieth century the dynamic – often prosperous – entrepreneur was epitomized by his ability to propose a new vision of economic and social progress.85 The textile sector was not part of this new vision. Highly innovative textile entrepreneurs, such as Rossi, were replaced by a younger group of conservative industrialists who perpetuated a vision of labour relations based on paternalism, of a father-figure that often rejected managerial solutions, opening the way to forms of ‘authoritarian paternalism’ in the interwar period.86

The interwar period The First World War mobilized new forms of labour: women and young men, but also non-industrial labour, which was heavily employed in textile production for the war effort. In a context of labour scarcity, wages were high but workers were subjected to long hours (overtime was compulsory), military discipline and high production targets. At the end of the war, the situation was rather strained and was followed by the so-called ‘biennio rosso’ (the two red years) with strikes and protests in all the major industrial cities. The outcome of these was a reduction of 83 Musso, Storia del lavoro, pp. 35–37. 84 Giorgio Roverato, ‘La terza regione industriale’, in Silvio Lanaro (ed.), Il Veneto. Storia d’Italia. Le regioni dall’Unità a oggi (Turin, 1984), pp. 163–230, at p. 186. 85 Paride Rugafiori, Imprenditori e manager nella storia d’Italia (Roma and Bari, 1999). 86 Duccio Bigazzi, ‘Le permanenze del paternalismo: le politiche sociali degli imprenditori in Italia tra Ottocento e Novecento’, in Ricerche di storia in onore di Franco Della Peruta (Milan, 1996), vol. 2; Roverato, Una casa industriale, pp. 321–352; A. Erseghe, G. Ferrari and M. Ricci, Francesco Bonfanti architetto: i progetti per la città sociale di Gaetano Marzotto 1927–1946 (Milan, 1986); Giorgio Roverato, ‘Valdagno e la “Città sociale” di Gaetano Marzotto Jr: tra utopia conservatrice e moderno welfare aziendale’, Annali di storia dell’impresa 13 (2002): 133–151. 298

The Italian textile industry the working day to eight hours and better wages, but these achievements were soon challenged by the advent of Fascism in 1922. The general improvement of the condition of workers happened during a difficult period for the Italian textile industry. The silk industry, which had traditionally been the most important sector of the economy, was, at the end of the First World War, 10 per cent smaller than in 1910.87 The textile industry had large stocks of unsold goods, nearly obsolete machinery and had to operate in a context of severe inflation, making the supply of raw materials haphazard. The years 1920 and 1921 saw the most profound crisis for the industry since Unification: entrepreneurs hardly had the resources to continue, trade unions were in turmoil and markets were stagnant. The application of a new tariff system to protect linen production and an increase in the export of cottons and raw silk are seen as turning points for the beginning of a new period of expansion that lasted until the end of the decade.88 It was the cotton industry in particular that most thrived during this period. It became the most important of the textile industries in terms of capital (more than 3 billion liras), number of workers (150,000), machinery (60 per cent of all spindles and 50 per cent of all looms), consumption of power (286,000 horse-power), total value of production (c.2.5–3 billion liras) and foreign transactions (0.7 billion liras for imported raw materials and 0.5 billion liras of exported finished cloth). The sector had 300 limited liability companies in 1932, compared with just 200 in 1901, and their total capital had increased threefold.89 The wool sector was much smaller than the cotton industry. It consumed just a quarter of the raw material and had half of the number of employees. The mid-1920s was a period of strong growth especially because of the substitution of imported cloth and yarn with domestic products. An important feature of the wool industry is that four large groups (Marzotto, Lanerossi, Tollegno and Borgosesia) accounted for 80 per cent of the total capital invested in the sector. The areas of Prato and Biella were still active but were characterized by smaller firms producing woollens rather than worsteds. In 1927 the industry provided work for 80,000 people. This might appear to be a relatively small figure, but Italy was the fourth largest exporter of woollen and worsted textiles after the UK, France and Germany.90 It was also, within the textile industry, the sector that best resisted the economic crisis of 1929. The reliance on home-produced commodities during the period of autarchy in the mid-1930s acted as a further stimulus for the wool sector. New fibres or mixes of wool with artificial fibres were accompanied by new production methods. This period of relative austerity coincided with substantial technical innovation for wool manufacturing. The production and manufacturing of artificial fibres appeared soon after the First World War and had its greatest expansion in the 1920s. This was an industry with strong links to chemical production that required considerable investment in 87 Tremelloni, L’industria tessile, p. 112. 88 Ibid., p. 231. 89 Ibid., pp. 153–62. 90 Comitato nazionale per la produtività – G.L.I.T., Il problema tessile italiano (1927), p. 55. 299

The Ashgate Companion to the History of Textile Workers research, product development and marketing. It was therefore concentrated in a few industrial groups. The early development of this industry in Italy was quite astonishing: in 1935 Italy produced 15 per cent of all European synthetics and was the third largest producer in the world after the USA and Japan. This was achieved with just 31 production plants employing 24,000 workers. We have no space here to enter into a detailed analysis of the dynamics of each of the sectors that formed the textile industry. The general trend across the industry in the interwar period was industrial concentration. In 1927 the textile industry counted more companies with more than 50 employees than any other sector. Just over 2,000 firms employed 580,000 workers. Ninety per cent of all the firms in the sector had more than 50 employees (compared to 85 per cent for the metal-processing sector and 68 per cent in the chemical sector). Textile firms were becoming not just bigger but also more bureaucratic and were reliant upon a web of collateral and commercial services. If before the war profits accounted for two and a half to three times the wage bill, after the war this ratio was significantly reduced. However, this trend was interrupted by the advent of Fascism. Real wages declined after 1922, moving away from the more egalitarian attitude that had dominated in the years immediately following the First World War.91 By 1931 the total workforce in textile production was six times bigger than that of 1901.92 The 1930s generally saw a contraction of the number of workers not just in textile production but also in clothing manufacturing. In 1937, for the first time in the history of the Italian economy, the mechanical industry overtook the textile industry in terms of the number of employees (21 per cent compared to 18 per cent). The Fascist era was also a period of further economic divergence between the so-called ‘triangolo industriale’ (the ‘industrial triangle’ between Genoa, Turin and Milan) and the rest of the country. Female work was the casualty of this process of change and the contingencies of the regime. Female participation in waged labour suffered as a result of the Fascist patriarchal models that wanted women to be relegated to domestic tasks. But female labour was also in a weak position because it was a time of change in the general economy, which was shifting from light production to heavier forms of industry such as chemical and mechanical products. As a result, there was a form of ‘involution’ and more and more women were employed within their domestic walls. When Italy entered the Second World War in May 1940, the laws regulating female employment were repealed, with the exception of areas where there was male unemployment.93 The period of political turmoil following September 1943 and the heavy bombardment of all the major industrial areas in the country caused a general standstill in the economy until the end of the war.

91 Musso, Storia del lavoro, p. 173. 92 Tremelloni, L’industria tessile, p. 229. 93 Musso, Storia del lavoro, p. 48. 300

The Italian textile industry The crisis of the textile industry and the rise of textile-clothing production At the end of the Second World War the Italian textile industry faced many challenges: it had to create a minimum stock, reorganize production, buy new machinery and quickly look for new suppliers and new markets. These were demanding tasks considering that capital was scarce, there was little money in circulation and inflation was spiralling, and the economic climate had suffered as a result of the years of civil unrest. However, there were also some opportunities. The overall international context was much more positive than it had been either at the end of the First World War or during the Fascist regime. American aid, together with governmental plans, an expanding internal demand and the revival of international commerce, made recovery faster than expected. The role of labour was profoundly different from its role in the period before the war: it accounted for a much higher percentage of the total cost of production due to the fiscal burden of the newly introduced national insurance and pension schemes. At the same time this change contributed to the final demise of the classic paternalism that had characterized the largest businesses in the sector, for instance Marzotto.94 The situation concerning the renewal of the technological set-up, especially in the wool sector, was more complicated. There were opportunities to create new products both for men’s and women’s wear.95 A new link was emerging between the industrial production of textiles and the rising fashion industry. Designers and stylists became new key figures in textile production as early as the mid-1950s. The second half of the 1950s therefore saw a strong recovery in the entire textile sector, which had somewhat languished over the previous five years. The opening of the European Common Market in 1959 coincided with a general improvement in standards of living throughout Europe. Consumption of textiles and clothing was increasing fast, and Italy’s textile exportation levels were rising, especially in the German market. However, the wool sector went into irreversible decline. Lanerossi, which for over a century had been the most important firm in the sector, was commissioned and became state-owned with a loss of jobs. The area of Schio-Valdagno also suffered from the re-sizing of the Marzotto group, which led to further labour unrest. Marzotto seemed to be destined to a sad end similar to that of Lanerossi, but by the early 1980s the group was able to recover. In 1985 it acquired Finbasetti and in 1987 the state-run Lanerossi, which had been its most important competitor for over a century.96 The story of Marzotto is, however, exemplary of a wider dynamic within the sector. Its winning strategy was to diversify production by investing heavily in the emerging textile-clothing sector. Moreover, it modernized both the technologies and the organization of production in its wool textile segment and continued to 94 Roverato, Una casa industriale, pp. 383–386. 95 Ibid., pp. 432–433. 96 Ibid., p. 456. 301

The Ashgate Companion to the History of Textile Workers invest in both cotton and linen manufacturing, thus counteracting the decline in wool manufacturing. It had to de-specialize the region of production (Valdagno and its surroundings) in the face of the negative forecast for the future of what had been the product specialization of the area since the early nineteenth century. This allowed the local economy to respond in more flexible ways to the crises of the wool sector. Marzotto also chose to focus on high-value-added products and in 1951–1952 it decided to begin production of ready-made garments, a promising sector in the immediate post-war period. However, the choice made by Marzotto was not followed by other wool textile firms. Marzotto created its own strategy and dominated the market. Towards the end of the 1950s it opened new production plants specializing in clothing manufacture in another area of the province of Vicenza where Marzotto was located, and also in the south of Italy in the area of Salerno. It further diversified its production by manufacturing acrylic fibres and at a later time by producing blankets.97 The 1970s were a period of general crisis for the hitherto dynamic clothingtextiles sector caused by the sudden rise in the cost of energy coupled with the spiralling cost of labour. This had a particularly negative effect on labour-intensive sectors. Economists and politicians were confident that the demise of textile manufacturing was unavoidable. Italy had reached industrial maturity and all areas of the production that had characterized the first phase of industrialization were destined to become key sectors of the economies in the ‘emerging’ countries in Africa and Asia. The crisis in many of the traditional areas of textile production seemed to confirm this prediction. Most of the firms were too small and undercapitalized to survive international competition from areas with low wage levels. This prediction has never become reality, at least not fully. The reasons why textile and clothing production is still important in Italy are to be found in both supply- and demand-led explanations. On the supply side, many firms, and not just the larger ones, decided to integrate production. It is not unusual to find firms that carry out all stages of production from the raw material to the finished items. There has also been a certain degree of cross-sector contamination within textile manufacturing: some of the technologies and raw materials are interchangeable. Italy is still one of the most innovative countries in the creation of mixed fibres (natural and synthetic). This mixing of products and fibres is in line with what consumers have required since the 1980s: a mixture of different items that had traditionally been provided by a diverse range of production set-ups (woollens for suits, linens and cotton for shirts, tailoring for men, dressmaking for women, hosiery for jumpers and stockings, and so on). The rise of casual wear has led to a certain degree of interchangeability between different clothing items, creating new competition, but also new alliances. 97 Giovanni Luigi Fontana and Giorgio Roverato, ‘Processi di settorializzazione e di distrettualizzazione nei sistemi economici locali. Il caso veneto’, in Franco Amatori and Andrea Colli (eds), Comunità di imprese. Sistemi locali in Italia tra Ottocento e Novecento (Venezia, 2001), pp. 527–617, at pp. 561–562. 302

The Italian textile industry Our analysis must, in conclusion, mention the more general economic and extraeconomic effects of the ‘Made in Italy’ factor. The success of high-quality Italian design and clothing in world markets has created new opportunities for sectors that do not specifically fulfil the criteria of ‘Made in Italy’. Italy has been successful not just in spreading style and cashing in on famous brands, but also in producing lower-quality products that have sold well in the shadow of the more characteristic and high-class products of the ‘Made in Italy’ brand. Today, it is this vast spectrum of cheap textile products and clothing that means Italy is still competitive against Asian producers.98

98 Ibid., pp. 558–560. 303

This page has been left blank intentionally

12 Japan Janet Hunter and Helen Macnaughtan

Japan has a long history of textile production, and the forms and nature of textile production have undergone major changes over the years since 1650. In order best to comprehend the consequent changes in the nature of textile work and textile workers, this chapter will be divided into three chronological sections: the years from the seventeenth century up until the 1850s, characterized by pre-industrial and proto-industrial forms of production; the period from the 1850s up to 1937, marked by Japan’s increasing integration into the international economy, accompanied by the growth of capitalist factory production and successful international competition; and the years 1937–2000, which were marked firstly by a shift of resources out of textile production for war-related reasons, followed by recovery and the relative decline of the significance of textile production in the Japanese economy, and in Japan’s international trade.

1650–1850s As might be expected in a low income country, in pre-industrial Japan clothing accounted for a considerable part of household expenditure even among the wealthiest families, and textiles was by far the largest sector of manufacturing. Many Japanese families continued to produce their own clothing, but the period witnessed the spread of the market economy, and with monetization and commercialization farmers and non-farmers alike became more dependent on market purchase of textiles. The central regime recognized the importance of the industry, and sought to confirm the control of guilds over the textile trade, and to some extent over production as well. However, the power of the central government remained limited, with several hundred quasi-autonomous domains able to pursue relatively independent economic policies. As domain finances became more precarious from the eighteenth century, some domain lords sought to control or promote textile production and trade to strengthen the local economy and their own income. Such measures were often beneficial to local textile industries, releasing producers and

Map 12.1 Map of Japan, showing major areas of textile production

Japan traders from stultifying central or guild controls. However, the authorities had little inclination or power to intervene in the actual production process.

Production The most important textiles produced in Japan from the seventeenth century were silk and cotton. Sheep were not indigenous to Japan, and no wool was produced before the 1870s. At the start of the period silk, worn mainly by the aristocracy, was mostly imported, but import restrictions from the early 1600s stimulated the growth of domestic production. A growing number of peasant families, particularly in the central regions, were involved in sericulture, the production of cocoons. Cotton was produced in Japan from the sixteenth century and gradually replaced hemp as the main clothing material for non-aristocratic families. Commercial production of cotton took off from the mid-eighteenth century. Data on production is very patchy, and growth was focused on particular areas, but it is estimated that when Japan was opened up to foreign trade in the 1850s–1860s up to three million farm household members working in cotton production were affected by the inflow of imports, some 10 per cent of the population. Until the mid-nineteenth century trade restrictions meant that virtually all textile production was for the domestic market. The growth of commercial textile production was closely tied to the growth of the economy, and to the spread of urbanization and consumer demand. Sumptuary laws attempted to restrict the use of silk, but the concentration of the ruling classes and wealthy merchants in cities and towns across the country increased conspicuous consumption, sustaining a buoyant demand for silk cloth. Demand for cotton products from the commoner classes likewise increased, and the Osaka region became a centre of cotton production. A lively trade in cotton and silk textiles developed across the country from the eighteenth century, with exports from the Osaka region to the capital, Edo (now Tokyo), particularly prominent. The growth of production was sustained by a number of indigenous technological advances, which increased both output and productivity. Some improvements came in the sericulture process, but probably more important were advances in weaving technology, in particular the spread of the so-called ‘tall loom’ across the country by the mid-nineteenth century, offering greater efficiency than the traditional looms. 

Yūjirō Ōguchi, ‘The Finances of the Tokugawa Shogunate’ (pp. 192–212), and Shinsaku Nishikawa and Masatoshi Amano, ‘Domains and their Economic Policies’ (pp. 247– 267), both in Akira Hayami, Osamu Saitō and Ronald P. Toby (eds), The Emergence of Economic Society in Japan, 1600–1859 (Oxford, 2004).  Satoshi Nakamura, ‘The Development of Rural Industry’, in Chie Nakane and Shinzaburo Oishi (eds), Tokugawa Japan: The Social and Economic Antecedents of Modern Japan (Tokyo, 1990), pp. 81–96, at p. 95.  William B. Hauser, Economic and Institutional Change in Tokugawa Japan: Osaka and the Kinai Cotton Trade (Cambridge, 1974), chs 3 and 4.  Nakamura, ‘Development of Rural Industry’, p. 88. 307

The Ashgate Companion to the History of Textile Workers In many areas textile production experienced greater specialization and division of labour, for example separation of the spinning, weaving and finishing processes. There was, moreover, extensive regional specialization, with differentiation among the products of different localities and the ability to cater for niche markets. However, almost all production remained handicraft production, and it was only well into the nineteenth century that a few producers began to use water power. The expansion of textile demand and production was not an unbroken upward trend. Both industries suffered from fluctuations in the economy, and silk, as a luxury commodity, was particularly badly hit by the economic problems generated by successive crop failures in the early part of the nineteenth century.

Organization of production During the seventeenth and eighteenth centuries much textile processing was urbanbased. Unprocessed or ginned cotton was bought in by specialist spinning and weaving artisans, whose products were then marketed by groups of wholesalers. It has been suggested that despite official support the significance of guilds had declined relative to the medieval period, and that urban textile producers often enjoyed a relatively high degree of autonomy. Guilds and protective associations were important, but more in the retailing and distribution of textiles than in their production. Such guild involvement did indirectly shape the options open to producers, by laying down the institutional framework of textile production. Silk weavers, too, concentrated in towns to supply the ruling warrior class. The silk weavers of Nishijin, in Kyoto, who produced high-class damasks and brocades, became particularly famous. Because silkworm cocoons were highly perishable, however, the process of drying and reeling the thread was often carried out at or near the place of cocoon production, often by the sericulturists themselves. The production of silk thread and some cloth therefore became an activity integral to many farm family economies. Over time textile production shifted increasingly to rural areas. Guild restrictions in the towns hindered the development of new technologies and raised the costs of production, so new producers seeking to satisfy the growing demand located their operations in rural areas where labour was cheaper. Urban artisans lost their technological superiority in the important finishing processes. By the start of the nineteenth century there were weaving specialists in many localities, some of them employing wage labour and running operations akin to handicraft factories. Elsewhere a putting-out system operated, with a wholesaler supplying thread and maybe equipment to weavers from farming families, and paying them  Hauser, Economic and Institutional Change, ch. 4.  Haruko Wakita, ‘Ports, Markets and Medieval Urbanism in the Osaka Region’, in James L. McClain and Osamu Wakita (eds), Osaka, the Merchants’ Capital of Early Modern Japan (Ithaca and London, 1999), pp. 22–43.  Tamara K. Hareven, The Silk Weavers of Kyoto (Berkeley, 2002), pp. 55ff. 308

Japan on a piecework basis. In many regions farm family members produced on their own account, selling their products direct to traders or wholesalers. The net result was twofold. Firstly, the momentum of textile production shifted away from older urban areas, in the process generating ‘new’ manufacturing areas, much of whose population was no longer dependent exclusively, or even mainly, on income from agriculture. Secondly, a very high proportion of farming families included one or more members engaged in textile production (including raw material production) on a full- or part-time basis, a pattern known as by-employment. This process of rural industrialization has been identified by some scholars as proto-industrialization, although use of the term can be controversial. What is apparent, however, is that by the nineteenth century Japan had an extensive textile industry catering to a large, well-functioning domestic market. Systems of organization were diverse, including cottage industry, quasi-factories (manufactories), domestic outwork and individual production, but elements of proto-capitalism were present. However, production remained on a handicraft basis.

Textile workers This organization of production resulted in a persistent diversity within textile workers as a whole throughout the pre-modern period. At one end of the spectrum were the specialist and highly skilled artisan weavers in places such as Nishijin. Acquiring skill through a long apprenticeship, they remained in the industry throughout their lives. As apprentices they were subservient to the master of the house, but could later set up on their own under the supervision of the guild. Whole families were often employed in these hereditary weaving enterprises. At the other end of the spectrum were those who produced textiles as farming side work. The number of farm family members so engaged, both in the production of raw material (silk cocoons or cotton) and in its processing, steadily increased over the period. In most cases the work was adapted to the farming cycle, but as rural industry expanded many families and individuals detached themselves from agricultural dependence and engaged in textile production full time. Many became highly specialized in different parts of the textile production process, and workshops developed in addition to work being carried out within the home. Ginning, carding and dyeing joined spinning and weaving as specialist activities. Data on textile workers from this period is largely local, so it is difficult to make general observations about the characteristics of textile workers. We know that in both gender and age composition they were very diverse, although males predominated in artisan production, particularly at the luxury end of the market,  See Nakamura, ‘Development of Rural Industry’; Hiroshi Shimbo and Akira Hasegawa, ‘The Dynamics of Market Economy and Production’, in Hayami, Saitō and Toby (eds), Emergence of Economic Society in Japan, pp. 159–191.  Hareven, Silk Weavers of Kyoto, chs 3 and 4. 309

The Ashgate Companion to the History of Textile Workers and a majority of textile workers from farm families are likely to have been women. Sometimes male and female workers would specialize on different tasks, with women, for example, producing ordinary cloth, while their menfolk produced rope and matting. In some regions the acquisition of weaving skills by women became a prerequisite for marriage, stimulating further the importance of textile production as an important income-earner. Unlike in the later period most textile workers would have continued to live in their own homes, although apprenticed and indentured workers resided in the dwelling of the employer. These workers are more likely to have experienced a more structured employer–employee relationship. Employment relations in commercial and manufacturing enterprises in this period tended to be categorized by the principle of oyabun-kobun, a fictive parent–child relationship established for mutual economic benefit. The oyabun, the parent figure, was responsible for the welfare and guidance of the kobun, who in return had a duty of obligation and obedience. Such formalized relationships extended to some of the rural specialist textile producers, but were weaker in many rural areas, and played little part in textile production within the family, which was subject to the norms of farm family operation rather than any broader code of commercial practice. Workforce diversity meant that a single occupational identity failed to emerge, but artisan weavers associated with the making of specific brands of cloth or thread had a strong occupational identity as producers of those commodities. The growth of national markets and specialist products acted to highlight the reputation of particular groups of artisans well beyond their localities. Such local groupings on occasions developed their own guilds and organizations to protect their collective interests.10

1850s–1937 Textiles was the lead sector in Japanese industrialization. It spearheaded the development of capitalist factories, and sustained Japan’s balance of payments over these years. The momentum started with silk reeling, then passed to cotton, and later wool and rayon. Until the 1930s there were more women than men in Japanese factories. Industries such as cotton and wool also symbolized the processing strategy (import of raw materials, export of processed goods) that characterized Japan’s twentieth-century growth. Within Japan the individual sectors of textile production were crucial for local economies across the country. The diversity of operation and organization even into the 1930s reflects the whole spectrum of patterns within Japanese industrialization. Until production was wound down with the onset of war in the late 1930s, textile production was pre-eminent in Japanese manufacturing. In recognition of this importance Japanese governments maintained a watching brief over the industry’s progress. A hands-off policy was adhered to except where 10 Nakamura, ‘Development of Rural Industry’, pp. 85ff. 310

Japan the national interest was seen to be at stake, and it was bureaucratic concerns over the poor conditions of textile workers that stimulated the introduction of protective legislation limiting hours of work, banning night work for women and prohibiting the employment of very young workers. After the First World War further new measures affected the operation of the industry, including the introduction of health insurance. Although textiles spearheaded factory development, a great deal of textile production continued to take place in the countryside, though textile by-employment declined rapidly. With a majority of textile workers from rural areas, the link with agriculture remained important throughout these years. The agriculture–industry relationship has been identified as a key factor in Japan’s industrialization, and the textile industry was pivotal to that relationship. The rural connection was key to the employment relations strategies and working conditions that emerged in the textile industry, giving it a paramount position in the Japanese industrialization process.

Growth of textile production Manufacturing of silk and cotton yarn and cloth remained by far the most important areas of textile production throughout the late nineteenth to the early twentieth century. Hemp and linen cloth continued to be produced on a small scale. Wool was manufactured in Japan for the first time in the 1870s, stimulated by demand for uniforms. Production expanded on the basis of imported raw material. Artificial fibres such as rayon were produced with some success from the 1930s. In 1935 Japan was the world’s second largest producer of both cotton yarn and cloth, and rayon yarn and fibre, and was also the world’s sixth largest consumer of raw wool. The same year Japan accounted for 64 per cent of world production of raw silk.11 These years witnessed the growth of factory production, and of capitalist forms of enterprise, but textile production retained considerable diversity in scale and organization. Agricultural by-employment in textiles continued, and in many sectors very small scale or family-run enterprises coexisted with larger factories. The context of the changes was a major expansion in production stimulated by balance of payments and trade imperatives consequent on Japan’s integration into the international economy from the 1850s. While accurate data for the period before the First World War is limited, the scale of the expansion that took place can be gauged from Tables 12.1 and 12.2.

11 Osamu Saito and Masayuki Tanimoto, ‘The Transformation of Traditional Industry’, in Hayami, Saitô and Toby (eds), Emergence of Economic Society in Japan, pp. 268–300, at pp. 268ff; Toshiaki Ushijima and Takeshi Abe, ‘Mengyō’, in Shunsaku Nishikawa, Kōnosuke Odaka and Osamu Saitō (eds), Nihon Keizai no 200 Nen [200 Years of Japan’s Economy] (Tokyo, 1996), pp. 225–255, at pp. 225ff. 311

The Ashgate Companion to the History of Textile Workers Table 12.1 Output of machine reeled silk (in 1,000 kg), 1895–1940 Output

Output

1895

3,083

1920

20,478

1900

3,901

1925

29,445

1905

4,655

1930

38,654

1910

8,725

1935

41,252

1915

12,787

1940

38,345

Source: S. Fujino et al. (eds), Sen’i Sangyō, vol.11 of Chōki Keizai Tōkei (Tokyo, 1979), p. 290.

Table 12.2 Output of integrated cotton mills (spinning and weaving), 1895–1935 Output (million ¥)

Output (million ¥)

1895

32

1920

1,091

1900

70

1925

1,213

1905

129

1930

641

1910

179

1935

1,136

1915

260

Source: S. Fujino et al. (eds), Sen’i Sangyō, p. 243.

Through the second half of the nineteenth century much textile production was geared to the domestic market, but increased opportunities for trade with the West brought huge demand for raw (reeled) silk from Western countries hit by silkworm disease (pébrine). As French and Italian silk producers recovered from this blight European demand declined, but Japanese producers, recognizing the importance of quality of output as well as quantity, successfully oriented their exports towards the United States, which by 1914 was taking nearly 70 per cent of Japan’s raw silk exports. High quality silk cloth, much of it from specialist artisan weavers, was also exported. Japan’s main competitor in silk, China, was outstripped. Japan’s cotton producers came under pressure to substitute for imported yarn, and from the 1880s production expanded through the use of imported raw cotton. Cotton yarn and cloth was exported from the 1890s, when the Sino-Japanese War of 1894–1895 stimulated the export of yarn to neighbouring Asian countries such as China and Korea. Yarn 312

Japan exports expanded to other Asian countries, including British colonies in South East and South Asia, and by the interwar period major advances were being made into parts of the Middle East, Africa and Latin America. By this time cotton cloth, too, was rapidly advancing into overseas markets, and the whole cotton industry was competing successfully with more established producers such as Britain. In the 1930s Japanese exports of wool and artificial silk (rayon) also expanded. The export significance of Japanese textiles is shown in Table 12.3. Table 12.3 Share of textiles in total Japanese exports (by value, %) Cotton yarn

Cotton cloth

Raw silk

Silk cloth Total export value (million ¥)

1890–1894

0.2

0.9

38.0

4.7

86

1895–1899

8.4

1.7

31.7

7.2

160

1900–1904

9.3

2.4

29.2

10.6

265

1905–1909

7.7

3.8

28.8

8.0

394

1910–1914

10.9

5.1

30.8

6.4

531

1915–1919

7.2

9.9

25.3

5.8

1500

1920–1924

6.8

16.3

34.9

7.2

1619

1925–1929

2.8

19.1

37.9

6.4

2093

1930–1934

1.1

20.3

23.0

3.7

1612

1935–1939

1.8

16.1

14.4

2.1

2927

Source: Y. Andō, Kindai Nihon Keizai Shi Yōran (Tokyo, 1975), pp. 22, 118.

The export expansion was characterized by a gradual move up-market into higher value-added goods. In the case of silk, Japan’s success in sustaining quality of output was regarded as a key to successful competition against China. Cotton yarn production moved gradually into higher counts (that is, greater fineness). Export success was never guaranteed by supply conditions, however, and short-term slumps from a downturn in the business cycle could hit producers hard. These were particularly obvious in the wake of the Sino-Japanese and Russo-Japanese wars, as well as after the First World War. Silk, as a luxury commodity, received a blow from the Great Depression after 1929 from which it never recovered. Expansion in response to external demand was also constrained by shortage of capital, much of it required for imports of technology. During the First World War, for example,

313

The Ashgate Companion to the History of Textile Workers cotton spinners were unable to import from Britain the equipment they needed to cater to the growth in demand for their products.12 One key advance in textile production in this period was the application of steam power, and later electricity. Production of all textiles was increasingly characterized by mechanization and the use of inanimate forms of energy. Gas was used for lighting from the 1880s, facilitating production during the hours of darkness. Technology transfer coexisted with developments based on indigenous technology. The ring spindles that dominated cotton spinning mills from the 1880s replicated Western practice, as did the technology of wool production. By contrast silk reeling technology was based on traditional Japanese practice, though both Western practices and domestic modifications were implemented in the interests of greater efficiency and productivity. Incremental technical changes were numerous, but a number of major technological advances epitomize the broader significance of technical change in this period. Some new technologies furthered the division of labour, such as the imported ring spinning machinery, or the division of the boiling and reeling processes in the silk reeling industry. Major advances, including the imported flying shuttle and jacquard apparatus, the indigenous high draft spinning equipment and multi-end reeling machine, the power-loom and, in the 1930s, the automatic loom, speeded up the process of production. The seriplane technology introduced in silk reeling was aimed at maintaining quality of output.13 Together, these advances made a significant contribution to the growth of productivity, particularly the productivity of capital. Advances in labour productivity were also made, but in most cases were contingent upon the greater application of capital, even in small enterprises. This change had two major effects. Firstly, although worker skill remained of considerable importance in both silk reeling and specialist artisan weaving of cloth, there was a significant element of de-skilling in many areas of textile production. Secondly, the emphasis on investment in capital equipment made it progressively more difficult for small and part-time textile producers to compete with larger operations, and hastened the spread of mechanized factory production and the capitalist employment relations that were associated with it.

Structure and organization: production, employers and employees As noted earlier, there was a diversity of forms of organization of textile production in this period, with large capitalist factories employing several thousand workers at one end of the spectrum, and agricultural by-employment, with farm family 12 Janet Hunter, ‘Britain and the Japanese Economy during the First World War’, in P. Towle and N.M. Kosuge (eds), Britain and Japan in the Twentieth Century: One Hundred Years of Trade and Prejudice (London, 2007), pp. 15–32. 13 For technology, see Masanori Nakamura (ed.), Technology Change and Female Labour in Japan (Tokyo, 1994), ch. 1; Ryōshin Minami and Yukihiko Kiyokawa (eds), Nihon no Kōgyōka to Gijutsu Hatten (Japanese Industrialization and the Development of Technology) (Tokyo, 1987), pp. 43–149. 314

Japan members engaged in production on a part-time, often putting-out basis at the other. In between were many small-scale workshops, many of them family owned. The balance within the industry shifted more towards larger scale capitalist enterprises over time, but at different speeds and times depending upon the sector of production. Factory production came first in silk reeling, but the scale of most silk reeling factories remained relatively small throughout the period. Few employed more than a hundred workers. The introduction of Western cotton spinning technology brought with it potential economies of scale, and from the turn of the century many cotton spinning firms employed over 1,000 workers. The scale of operation grew over time, particularly as many cotton spinning firms extended their operations into weaving, resulting in integrated mills. By the interwar years some specialist weaving companies in the local production areas (sanchi) were also employing hundreds of workers in a single mill. The cotton industry also witnessed a growing concentration of capital, with production dominated by four very large companies. Ownership in silk reeling remained much more fragmented, although there were some large producers. This process of concentration had advanced least in weaving.14 In 1935 there were still nearly 377,000 workers employed in textile workshops with four or fewer workers, and the persistence of small operators is shown in Table 12.4. Table 12.4 Textile workers by size of factory/workshop, 1920–1940 Percentage of total workers in factories with > 100 workers

Percentage of total workers in factories with < 9 workers

1920

46.4

28.2

1925

53.5

25.0

1930

46.0

31.0

1935

44.4

30.0

1940

40.1

33.7

Source: Calculated from figures in M. Umemura et al. (eds), Rōdōryoku, vol. 2 of Chōki Keizai Tōkei (Tokyo, 1988), p. 250.

The spread of mechanized capitalist production in the textile industry was stimulated by three main factors during this period of Japan’s development. The first was the pressure of export demand. Japan was potentially well placed 14 Teijiro Uyeda, Small Industries of Japan (New York, 1938), pp. 20–111. For the sanchi, see Takeshi Abe, Nihon ni okeru Sanchi Men’orimonogyō no Tenkai (Development of Sanchi Cotton Weaving in Japan) (Tokyo, 1989). 315

The Ashgate Companion to the History of Textile Workers to export textiles from the mid-nineteenth century, but needed to work to gain a foothold in international markets, and to sustain it. The need for international competitiveness was a powerful stimulus to mechanization, concentration and scale, although very small producers could be integrated into the export trade, for example through trading companies. The second was the imperatives of Western technology. Particularly in industries dependent on imported technology, such as cotton spinning and wool manufacture, technology brought with it Western patterns of organization and investment, even though these might be modified to suit the Japanese environment. Finally, the context of these organizational changes in production was a broader move to ‘modernize’ Japan, to build a strong industrial economy that could command international respectability and strength. Both the copying of Western organizational and technological forms and the development of indigenous forms of production were important in this process. Some textile employers and employees also looked to the West for models of collaboration and coordination, but cartel-type activity between employers was limited in the early decades after 1850. The change of government with the Meiji Restoration of 1868 brought formal abolition of the old guild privileges, and the demise of old regulations controlling master and servant relationships initiated a period of what many textile employers regarded as ‘confusion’ in employment relations.15 New types of employer associations appeared. These included industry associations, which brought together small employers, and in textiles were mostly of local importance; national employers’ associations, the most famous of which was the Cotton Spinners’ Association (Bōren); and regional employers’ associations, notably the Suwa Silk League. These associations had diverse aims, including coordinated restraint of production and other cartel activities, and enhanced control over labour. Coordinated employer attempts to control labour were most conspicuous in the years up to and immediately following the First World War.16 The first labour unions in Japan appeared on a very small scale in the late 1890s, but involvement by textile workers remained limited. Labour union activity reemerged in Japan towards the end of the First World War. Total numbers involved were relatively small, but organization among textile workers was conspicuous. Formal attempts to legalize union activity in the 1920s were unsuccessful, and union influence and numbers waned in the 1930s in the context of political suppression. Throughout this period there rarely existed any coherent form of collective bargaining even within a factory or firm, let alone across an industry or sector of an industry. Where unions were involved there was a tendency for ad hoc dispute resolution within a single mill rather than any broader industry-wide negotiation involving more than one firm or employer. Overall, the extent of union

15 Akiko Chimoto, ‘Shokkō Mondai Taisaku kara mita Meijiki Koyō Kankei’ (Meiji Employment Relations seen through Strategies to deal with the ‘Worker Problem’), Shakai Kagaku 35 (1985): 126–177. 16 Janet Hunter, Women and the Labour Market in Industrialising Japan: The Textile Industry before the Pacific War (London, 2003), ch. 8. 316

Japan organization in textile firms was very limited, and the powers of those unions that did exist weak.17

Textile workers18 Textile workers were initially recruited in the locality of the mill or workshop through means such as billboards, personal contact or word of mouth. These methods continued to be used by small local employers throughout the pre-Second World War years. With the expansion of production and growth of factories, local sources of labour proved increasingly inadequate, compelling employers to recruit further afield. The majority of textile companies became dependent on recruiting agents, who would scour particular regions on behalf of one or more factories. Employment exchanges were established from the 1920s, but the textile industry was slow to make use of them. The dependence on recruiting agents was in part the result of labour market imperfections, and the rapid increase in demand for labour, but it also resulted from the nature of the labour force. Three attributes in particular characterized a majority of textile workers in the new mills, and in turn shaped the institutions of recruitment and production, the pattern of labour protest and workers’ image and identity. A majority of textile workers were young, female and of rural origin. Although in the interwar period some cotton factories had high concentrations of Korean immigrant workers, the majority of textile workers throughout these years were Japanese. They came from throughout the Japanese islands, but recruitment was focused in particular on specific areas such as the island of Kyūshū in the south-west and the Japan Sea coastal and inland central areas. Extensive internal migration led to cotton workers from the furthest reaches of the archipelago being concentrated in the urban areas around Osaka, Nagoya and Tokyo, and silk reelers moving into the mountainous central areas from neighbouring prefectures. Patterns of migration were shaped by a range of factors, including the activities of individual recruiters and localized transport developments. Factory textile workers were invariably characterized by employers as coming from ‘poor’ families, and there is no doubt that many workers did indeed come from low income families and villages. However, there were also textile workers from relatively well-off backgrounds, and in some cases textile employment was considered an education as much as an income-earning opportunity.19 Workers came from various occupational 17 Yūko Suzuki, Josei to Rōdō Kumiai (Women and Labour Unions) (Tokyo, 1990). 18 Most of the data in the following section is taken from Hunter, Women and the Labour Market. See also Toshiaki Chokki, ‘Labour Management in the Cotton Spinning Industry’, in Keiichirō Nakagawa (ed.), Labour and Management (Tokyo, 1979), pp. 145–167; Gail Lee Bernstein, ‘Women in the Silk Reeling Industry in Nineteenth Century Japan’, in Gail L. Bernstein and Haruhiro Fukui (eds), Japan and the World (New York, 1988), pp. 54–77. 19 Wakizō Hosoi, Jokō Aishi (Pitiful History of Female Workers) (1925, repr. Tokyo, 1954), pp. 52–53. 317

The Ashgate Companion to the History of Textile Workers backgrounds, but by far the largest proportion came from families whose main occupation was agriculture. Even where workers’ families were not engaged in agriculture they were likely to reside in localities designated as ‘rural’. Many early cotton factories in the 1870s–1880s employed both men and women, but increasingly the textile labour force both within factories and outside them became predominantly female. Some indicative figures are shown in Tables 12.5 and 12.6. Table 12.5 Employment in private textile factories, 1896–1920 All workers

All workers

1896

347,000

1910

652,000

1900

360,000

1915

832,000

1905

474,000

1920

1,193,000

Note: Figures rounded to nearest thousand Source: M. Umemura et al. (eds), Rōdōryoku, p. 243.

Table 12.6 Textile workers by branch (% female workers in brackets), 1920–1940 1920

1930

1940

438,900 (83)

480,700 (86)

234,000 (84)

51,200 (58)

30,400 (54)

47,200 (67)

Spinning

297,100 (69)

204,100 (68)

235,500 (73)

Weaving

701,800 (72)

424,300 (69)

480,500 (73)

51,700 (51)

59,200 (40)

68,600 (54)

Dyeing/finishing

300,000 (44)

202,300 (12)

211,200 (20)

Sewing, tailoring

213,200 (57)

232,900 (46)

384,400 (54)

Silk reeling Thread manufacture

Knitwear/braid

Note: Includes all categories of worker, numbers to nearest hundred, percentages rounded Source: Janet Hunter, ‘Gendering the Labor Market: Evidence from the Textile Industry of Interwar Japan’, in Barbara Molony and Kathleen S. Uno (eds), Gendering Modern Japanese History (Boston, MA, 2005), pp. 359–392, at p. 365 (calculated from 1920, 1930, 1940 censuses).

318

Japan The growing importance of female workers in both factory and small-scale textile production was accompanied by increasing gender segmentation of occupation within the textile industry. More and more the minority of male workers were assigned to jobs that required either skill or physical strength. In cotton production, for example, men were allocated to preliminary processes such as unbundling raw cotton, and finishing processes such as fulling and dyeing, whereas women took the core spinning and weaving tasks. Women continued to dominate the relatively skilled task of silk reeling but in this sector of production, as elsewhere in textiles, it was men who were engineers, clerks and managers. As the proportion of female workers increased, the number of tasks open to them diminished.20 A number of factors contributed to textiles becoming ‘women’s employment’. One was the existence of a tradition of female textile work. Another was the rapid expansion in the demand for labour, which pushed employers into looking further afield for labour, and hence towards female members of farm families considered more dispensable to agricultural production. Above all, though, female workers were regarded as more docile and cheaper than their male counterparts. These latter characteristics were confirmed by the employment of younger women, because daughters were more easily released from the farm family economy than were wives. The new factories followed an earlier tradition of sending some daughters away for employment at the age of 11 or 12, and the age profiles of textile workers in the 1890s show a considerable proportion of the workforce was under the age of 14. Some male weavers were also very young, with both male and female weavers indentured at an early age. Over time the very youngest workers were excluded from the textile workforce by legislation and the spread of compulsory education. By the interwar years it was illegal to employ workers under 15 who had not completed compulsory education (although breaches were not uncommon). By this time, moreover, a distinct divergence had appeared in the age profiles of women and men in textiles. Although there were some women employed in textile production who were older and married, the vast majority were between the ages of 15 and 25. The age profile of male textile workers showed a much wider age span, with less concentration on the late teens–early twenties, much more in line with the general age profile of the population. This pattern inevitably meant that the majority of female workers were unmarried. Textile work had come to be associated with a life cycle view of women’s work, in which daughters only worked for a limited period prior to marriage. Such short-term employment held down career advancement opportunities and limited wage progression. Working conditions in nineteenth-century textile factories were poor, although how far they were inferior to conditions in equivalent factories in other early industrializing countries is hard to assess. Working hours were extremely long, ranging from 12-hour shifts in the cotton spinning mills (including at night) through to shifts of 16 hours or even more in silk reeling or weaving in the peak 20 Janet Hunter, ‘Gendering the Labour Market: Evidence from the Textile Industry of Interwar Japan’, in Barbara Molony and Kathleen S. Uno (eds), Gendering Modern Japanese History (Boston, 2005), pp. 359–392. 319

The Ashgate Companion to the History of Textile Workers season. Health and safety provision was inadequate, and the extent of illness and injury very considerable. By the first decade of the twentieth century textile factory conditions were considered a national disgrace. Over time there was substantial improvement in larger mills, especially in the cotton industry, and by the 1930s some cotton firms had become models for international textile labour management. Some silk mills followed suit, but in many working conditions remained poor, and small workshops outside any legislative control often fell seriously short of what was deemed acceptable. It would be a mistake to tar all small producers with the same brush. The family firm could offer advantages as well as disadvantages. It would also be an error to regard the best ‘paternalistic’ textile firms of the interwar years as representative of the industry as a whole.21 Formal written contracts did not become widespread in the textile industry until the early twentieth century. Even then many contracts were couched in very general terms, with only limited reference to pay levels or workers’ rights. Because even older female workers were legal minors according to the provisions of the Civil Code, contracts were into the interwar years often signed by fathers or other male legal guardians. The involvement of a worker’s family in the employment relationship, including in the payment of wages, was a strong characteristic of much female employment in the textile industry. While male textile workers, who might be indentured or training as apprentices, might also be legal minors, the majority of male workers enjoyed a more autonomous relationship with the employer. It is hard to establish a consistent picture of wage levels and wage payment across the industry during the industrialization period. Female workers were in general paid less than male workers, but there was enormous diversity within the industry, and some were able to earn sufficient wages to permit independence from the family. From around 1900 most female workers were on piece-rates, whereas men tended to be paid on daily or monthly rates. Some common wage institutions emerged across the different sectors despite their diversity. These included payment according to a principle of ‘relative efficiency’, the setting of a cap on the total wage bill, and the withholding of workers’ wages for direct payment to guardians. Many families also benefited from wage advances at the time of contract. Particularly in the interwar decades Japanese textile labour was criticized as ‘cheap labour’, giving the country’s cotton industry an ‘unfair’ advantage in international markets. The extent of the ‘cheapness’ of Japanese labour remains much debated, but it is clear that the real cost of textile labour, taking into account non-wage labour cost and productivity levels, was much higher than suggested by wage levels alone.22 Most female textile workers were housed in mill dormitories, or in small workshops lived with the owners on a quasi-family basis. Some older married workers lived out, and small local producers were more dependent on commuters. Some male workers at larger companies also lived in company housing or 21 Hiroshi Hazama, ‘Labour Management in Japan’, Japanese Yearbook on Business History 2 (1985): 32–53; Shuichi Harada, Labour Conditions in Japan (1928, repr. New York, 1979). 22 Hunter, Women and the Labour Market, ch. 6; John E. Orchard, Japan’s Economic Position: The Progress of Industrialization (New York, 1930), chs 19 and 20. 320

Japan dormitories. The integration of residence with the production process made long working hours more possible and offered greater possibilities for labour control. Early dormitories were often crowded and unsanitary, compounding health problems, but there was considerable improvement over time, and large mills in particular sought to promote ‘paternalistic’ policies for the welfare of workers, providing educational and recreational facilities. However, even by the late 1930s there remained many smaller mills and workshops in which living conditions were of a low standard. Poor conditions, low wages and the emphasis on a life cycle view of women’s work came together to promote high turnover among female textile workers in particular. While there was movement from mill to mill, many worked in the industry only for a short time prior to marriage or other employment. Female textile workers were perceived as short-term, transient workers earning ‘supplementary’ wages for the family economy, and not worth investing in as workers. The public image of the female textile worker (jokō) was, however, enormously important, and from the turn of the century they were identified not just as the key to international textile success, but as objects of pity and victims of the growth of industrial capitalism.23 This image was taken up by campaigning radicals and social reformers, and left a deep imprint on interpretations of Japanese economic development. The ‘passive’ identity of textile workers suggested by their youth and gender is incorrect. Even before 1914 there were sporadic unorganized disputes in the industry, undertaken with mixed success. Forms of protest included the withdrawal of labour, articulation of complaints and absconding. Protests became more organized after the First World War, with some union involvement. Rates of turnover and absconding declined, but there were a number of major strikes characterized by withdrawal of labour. Many of the participant textile workers were male, but there was also female involvement, and in some disputes women workers took the lead. Textile workers were therefore hardly passive, but a very small proportion of workers were involved in organized protest.24

23 Symbolized by the persistent use of the phrase jokō aishi, literally translated as ‘pitiful history of factory girls’, from the title of a 1925 work on the industry. See also the approaches in Yasue Aoki Kidd, Women Workers in the Japanese Cotton Mills, 1880–1920 (Ithaca, 1978), pp. 4ff; Masanori Nakamura, Rōdōsha to Nōmin [Workers and Farmers] (Tokyo, 1976); E. Patricia Tsurumi, Factory Girls: Women in the Thread Mills of Meiji Japan (Princeton, 1990). 24 Suzuki, Josei to Rōdō Kumiai; Yūko Suzuki, Jokō to Rōdō Sōgi (Female Workers and Labour Disputes) (Tokyo, 1989); Eiji Matsumoto, Seishi Rōdō Sōgi no Kenkyū (Study of a Silk Workers Dispute) (Tokyo, 1991); Barbara Molony, ‘Activism among Women in the Taishō Cotton Textile Industry’, in Gail L. Bernstein (ed.), Recreating Japanese Women (Berkeley, 1991), pp. 217–238; Hunter, Women and the Labour Market, ch. 8. 321

The Ashgate Companion to the History of Textile Workers

1945–2000 The importance of the textile industry in Japan’s national economy prior to the Second World War is indisputable, but its post-war role in Japan’s economic development is more difficult to assess. Although by the mid-1970s in decline, textile production remained an important export industry during the early post-war decades, despite overall contraction, experiencing another ‘mini heyday’ up until 1975. During the last three decades, however, Japan has lost its competitive advantage in labourintensive areas of production such as natural textile manufacturing. The Japanese cotton industry was on its knees by 1945. Two-thirds of pre-war cotton spindleage was scrapped by wartime administrators, and bombing and destruction of urban areas caused a further loss of 20 per cent of spinning and 14 per cent of weaving capacity.25 After 1945 textile companies were faced with raw material shortages, unreliable power supplies, hyperinflation and a workforce whose members had returned to their rural homes. During subsequent decades, the Japanese textile industry saw great changes in both its internal structure and the economic environment within which it operated. Under the Occupation (1945–1952), the textile industry was soon designated as a key sector to lead Japan’s economic recovery. For the USA, Japanese textile companies could provide a market for surplus raw cotton. The first initiatives within the industry were the improvement and installation of new machinery, and the amelioration of working conditions for labour. By 1951 pre-war productivity levels had been regained; thereafter, labour productivity in the industry began to increase beyond pre-war levels. The Korean War (1950–1953) gave a kick-start to the textile industry through demand from UN forces. Restrictions on production were lifted and exports increased. The end of the war brought a sluggish domestic market and contributed to a brief recession in 1954/55. While the Korean War had brought economic revitalization to the industry and led to a large increase in production and workers, the first post-war ‘recommended curtailment’ of operations was imposed in March 1952 in order to solve ‘overproduction’.26 This policy (‘recommendation’) was utilized by the Ministry of International Trade and Industry (MITI) at various intervals throughout the 1950s and 1960s even as the industry benefited from a prosperous economic environment and witnessed increased production, capacity and labour, as shown in Table 12.7. Japan quickly re-established itself as a producer and exporter of cotton textiles during the 1950s (see Table 12.8), and by the 1970s was facing international pressure to agree to voluntary export restraints.

25 A.D. Walton ‘The Japanese Cotton Textile Industry, 1945–1955’, Journal of the Textile Institute 49 (12) (December, 1958): 696–705. 26 This section is based on Nenjū no Bōsekigyō, JSA, Annual (various years). 322

Japan Table 12.7 Textile mills and workers employed, 1948–1984 (excluding apparel) Establishments

Workers

Establishments

Workers

1948

13,732

633,545

1970

112,754

1,264,228

1950

26,997

811,847

1975

114,111

995,669

1955

32,404

964,011

1980

39,741

691,018

1960

38,773

1,163,253

1984

36,269

626,567

1965

100,157

1,326,872

Source: Japan Statistical Association, Historical Statistics of Japan, vol. 2 (Tokyo, 1988), p. 283.

Table 12.8 Value of textile exports 1945–1980 (million ¥) Cotton yarn

Raw silk

1945

24

–

1950

6,323

1955

Cotton textiles

Woollen textiles

Silk textiles

Synthetic yarn and fibres

12,969

642

4,810

–

14,091

74,101

1,569

7,981

–

8,756

18,005

82,757

10,003

5,662

15,588

1960

18,861

18,162

126,507

19,628

18,779

97,833

1965

7,353

4,867

108,944

31,266

13,049

328,179

1970

5,314

464

67,541

27,190

5,279

510,729

1975

6,676

1

77,107

13,060

4,432

654,681

1980

11,730

–

110,037

11,252

10,360

625,530

Source: Japan Statistical Association, Historical Statistics, vol. 2 (Tokyo, 1988), p. 39.

The textile industry followed national economic cycles during these first growth decades, with temporary inflationary booms and recessions within an overall climate of rapid economic growth. During the 1950s and early 1960s demand for textile products increased, especially for natural textiles, but the industry also faced increased domestic competition for labour from other Japanese manufacturing industries, particularly electronics. In 1971 a US–Japan textiles dispute led to further controls on exports of natural textiles to the US market, while the depressed conditions following the 1973 oil shock resulted in the curtailment of production and labour recruitment. A depression cartel was organized in the industry from the first half of 1975, although the effects of the economic shocks were relatively quickly absorbed; the cartel was 323

The Ashgate Companion to the History of Textile Workers abolished later in the year, and demand for production and labour revived. However, just a few years later, conditions had again worsened, and in 1978 the textile industry petitioned successfully to be designated as a ‘depressed industry’. This was a significant period of transition in the structural importance of the textile industry within the domestic economy and within a changing global market for textile products. These decades were characterized by a changing emphasis in consumer demand from quantity to quality; steady decline in the prices of natural fibre products; and a shift within textile production away from natural textiles to synthetic textiles. Cotton and wool retained a key position in production during the early years but silk declined significantly. The production of synthetic non-cellulose fibres in Japan increased tenfold between 1960 and 1974, undermining the position of the natural fibres processing industries. Japan’s mainstay natural textile industry had lost much of its production share and competitive advantage by the early 1960s. By 1980 the production of synthetic and chemical fibres represented some 68 per cent of all yarn production by volume, with cotton 25 per cent, wool 6 per cent and silk less than 1 per cent. The increasing competitiveness of natural textile industries in other Asian economies also reduced demand for Japanese exports, as did the relative decline of the textile industry’s importance within the national economy because of the rapid growth in the chemical and heavy industries. Textile goods as a proportion of Japan’s total exports by value declined from 37.3 per cent in 1955 to 13.5 per cent in 1965, and to 5.3 per cent in 1975.27 The production of textile fibres remained an important industry. Japan still accounted for around 25 per cent of global cotton yarn production in the mid1970s and was the world’s second largest producer of synthetic yarn behind the United States.28 However, the Japanese natural textile industry was in decline from as early as the mid-1960s, but particularly from the mid-1970s, suffering a loss of global competitiveness as comparatively high domestic plant and labour costs led to weakened regional (Asian) demand for Japanese products. The industry was forced to restructure in an effort to stave off this decline and regain competitiveness within the changing global market. Unlike their Western counterparts, many of Japan’s leading textile producers have survived the restructuring of their ‘declining’ industry. The Japan Spinners’ Association (JSA)29 was powerful and well organized during these post-war decades and cooperated with MITI in restructuring the industry and legislating for disposal and modernization of textile equipment. While these activities alleviated some of the hardships of decline for Japan’s largest textile companies, they could not prevent the overall decline in natural textile manufacturing. During the 1970s and 1980s, therefore, the large-scale textile companies increasingly diversified their 27 T. Nakamura, The Postwar Japanese Economy (Tokyo, 1995), p. 66 (Table 3.3.). 28 K.C. Park, An Economic Analysis of Trends in the Japanese Textile Industry (Ph.D. thesis, Catholic University of America, Washington DC, 1976), pp. 20ff. See also Dennis McNamara, Textiles and Industrial Transition in Japan (Ithaca, 1995). 29 The pre-war Cotton Spinners’ Association (Bōren). The post-war Japanese title was Nihon Bōseki Kyōkai. 324

Japan portfolios, specializing in synthetic textile production, increasing investment in overseas (particularly South East Asian) natural textile production and shifting into non-textile business. Survival was more difficult for smaller family-run firms, and these declined from the 1970s. After the 1990 collapse of the bubble economy, textiles, like other areas of manufacturing production in Japan, suffered from domestic recession and increased global competition from emerging/developing markets, as indicated by the figures in Table 12.9. From the late 1980s Japan increasingly lost out to China in particular as a base for manufacturing and production. In a 1998 report on the textile industry, the Ministry of Economy, Trade and Industry (METI) acknowledged that problems facing the industry included a lack of adaptability to liberalized markets, low profitability, inefficient supply systems and a lack of cost and consumer competitiveness particularly with respect to other Asian countries.30 There needed to be a global strategy for textile production acknowledging Japan’s position as a technically advanced country. Japan also remained until 2004 restricted in its trade with the USA and Western Europe under the Multi-Fibre Arrangement (MFA). Both METI and textile leaders expressed optimism that the textile industry could continue to play an important role in Japan’s economic future if these problems were overcome, proposing a new direction for the Japanese industry as a leader of a distinct Japanese and Asian fashion and consumer culture, as an industry at the frontier of textile technology, and as a promoter of ecological textiles. Table 12.9 Production and trade in textiles, 1980–2000 (thousand tons) Total production

Imports

Exports

1980

2,050

278

601

1985

1,983

466

631

1990

1,822

817

462

1995

1,382

1,369

394

2000

1,089

1,692

439

Note: Throughout this period chemical fibres accounted for over 80 per cent of total production. Source: Statistics Bureau/Statistical Research and Training Institute, Ministry of Public Management, Home Affairs, Posts and Telecommunications, Japan Statistical Yearbook, 2003 (Tokyo, 2002), p. 318. 30 Ministry of Economy, Trade and Industry (METI), Summary of the Report on the ‘Vision of Japan’s Textile Industry and Ideal Policies for the Industry’ (December 1998). 325

The Ashgate Companion to the History of Textile Workers Firms and production The structure of Japan’s natural textile industry was characterized during this period by a large number of competing firms and small average size of enterprises (see Table 12.7). Many small-scale (often family-owned or family-run) enterprises, particularly in weaving, still existed alongside the large-scale multi-mill (multiprocess) firms. In this respect textiles remained a very labour-intensive industry, at least up until the mid to late 1970s, and conditions varied across the scale and range of enterprises. The synthetic textile industry comprised a few large-scale firms requiring large capital investment and complex raw materials and equipment. The central umbrella organization in the industry was (and still is) the JSA. Despite its name, the JSA dealt with all sectors of the textile industry, not just spinning. During this period, the JSA classified affiliated companies into three main groups. The largest companies (ōtekaisha) were known as the ‘Big Ten’ cotton spinning companies. The Big Ten (‘Big Nine’ from 1966) originated in the pre-war period, and tended to engage in mass production as well as operating divisions concentrating on different processes, usually including spinning, weaving, dyeing and finishing. With some adjustments, they have continued as predominant textile firms. Behind them were the shinbō (‘new spinning companies’) and the shinshinbō (‘new-new spinning companies’). Although some of these companies were established prior to or during the Second World War, they were identified as ‘newcomers’ or ‘new-newcomers’ based on the timing of their membership of the JSA. They were seen as ‘younger’ by comparison to the large-scale Big Ten companies, and were predominantly smaller in scale, but in the context of JSA affiliation and plans to unify the industry during this period, they essentially acquired many of the features of the Big Ten companies. Several ‘new’ companies became as prominent as the Big Ten, at least in terms of scale of production. Despite fierce competition, textile firms often shared information on labour policies, as well as on technology, raw materials and markets. The JSA acted as an industry (employer) association, seeking to promote cooperation (for example joint wage negotiations) among members, and also transmitting information between textile companies and government. Firms also came together for joint discussions under the aegis of the national textile union federation Zensen. While small-scale enterprises made up about 80 per cent of the total number of textile enterprises in the 1950s, it was large-scale enterprises that dominated production.31 JSA figures show that the Big Ten cotton companies together produced about 70 per cent of all cotton goods exports in the mid-1950s. Many small enterprises specialized in weaving and dyeing, and often had indirect subcontracting links to the big spinners via trading companies. In spinning, members of the JSA represented around 80 per cent of total natural and multi-fibre spinning capacity, while the remaining 20 per cent was found in small-scale enterprises in sanchi (‘production 31 Helen, Macnaughtan, Women, Work and the Japanese Economic Miracle: The Case of the Cotton Textile Industry, 1945–1975 (London, 2005), pp. 9–15. 326

Japan areas’). Weaving firms, often located in the sanchi, were less likely to be members, as small-scale and family establishments were in general more likely to fall outside the JSA umbrella. However, the power of the JSA as an industry association for medium and large-scale enterprises meant that patterns and conditions established in the larger enterprises during these years also filtered down to small firms outside the JSA. The role of the union federation Zensen in improving and unifying conditions across the industry was also significant. Zensen was one of the largest and most powerful industry-wide unions during the high growth years, and was closely aligned with the Liberal Democratic Party (LDP) which continuously held political power until the mid-1990s. At the end of the Pacific War, Japan’s textile industry faced several challenges: modernizing equipment to save labour, improving operational speed to increase productivity, responding to foreign competition and pressure to increase the quality of output. From c.1950, the textile industry addressed these challenges by looking towards further automation. In the 1960s, however, a move to ‘large packaging’ machinery took place. From 1960 the Continuous Automated Spinning (CAS) system implemented continuous automated cotton mixing, carding and drawing processes.32 In the late 1960s high-speed drawing frames and open-end spinning frames were developed and introduced into the industry. One effect of CAS technology was an unprecedented reduction in labour requirements. Further labour-saving innovations in spinning mills included automatic can transport devices for drawing frames, and automatic doffing devices for use on spinning frames. A five-year Textile Industry Reorganization Programme (1967–1972) called for the scrapping of outdated machinery and replacement with modernized equipment. Weaving technology did not change as quickly. This was partly due to the nature of the weaving process itself. It also reflected the structure of the weaving industry, with its many small to medium-scale enterprises. Many enterprises continued to use the system of group running of automatic looms introduced in the late 1920s– early 1930s. Advances in the 1960s and 1970s centred on the domestic development of the water jet loom and the air jet loom, and an all-out move to this technology took place in the industry from 1979. Heavy investment in ‘state-of-the-art’ textile machinery resulted in a much more capital-intensive industry by the late 1980s– early 1990s. With the implementation of modern technology and machinery, labour productivity growth in the industry increased significantly. Spindleage increased, and more equipment led to greater scale of operation and more larger-scale enterprises. Output per spindle also rose, showing a 25 per cent increase between 1950 and 1975.33 Increased technology and productivity meant lower labour requirements, but did not automatically make the textile industry capital-intensive or solve the labour shortage that had emerged, and which will be discussed 32 Japan Spinners’ Association, Zoku Sengo Bōseki Shi (Post-war History of the Spinning Industry Continued) (Osaka, 1979), p. 200. 33 Ibid., pp. 204–205. 327

The Ashgate Companion to the History of Textile Workers below. Through the decade from 1965 the industry faced a struggle to fill quotas for registered full-time workers (traditionally young females). JSA documentation shows that in 1963 the industry recognized that labour supply was likely to continue to decline, and that it needed to rationalize and become less labour intensive. It was this labour shortage of young female workers that was in part pushing these technological improvements.34 Labour productivity across the textile industry was relatively low compared with Japan’s other major industrial sectors, and the weaving sector remained particularly labour intensive. However, employer attempts at labour rationalization were resisted by strong labour unions during the 1960s and 1970s. Therefore the modernization of equipment, improvements in technology and decline in labour requirements of the high growth rate period never really solved the problem of labour supply. Labour ‘insufficiency’ and recruitment problems persisted.

Organization of production A high proportion of factories in the industry continued with a shift-system of production, typically a two-shift (per day) system. The ‘early’ shift usually worked from 05:00 to 13:45, and the ‘late’ shift from 13:45 to 22:30 (that is eight hours of work with a 45-minute break). This two-shift pattern was particularly predominant in the large-scale cotton companies, but as more part-time female workers (often ‘older’ and married) entered employment, the proportion and variety of ‘day-only’ shifts increased. Twenty-four hour operation increased in the industry from the mid-1960s, particularly after 1969, but only for male workers. Female workers were banned from night work. From April 1947, under the Labor Standards Law, hours were officially limited to eight hours per day (48 hours per week). The shift system in the industry was adjusted accordingly to two shifts of eight hours each. Following further negotiations from 1957 between the textiles union Zensen and the cotton companies, hours were further reduced to 7.45 hours per day/shift – that is, 46.5 hours per week. This push for a reduction in working hours was of symbolic importance for the industry, as the issue was closely identified with debates over the exploitation of young female workers by the industry in the pre-war period. Reductions in working hours had previously been carried out in response to government legislation, or voluntarily by management, usually in the context of production curtailment. The 1957 claim for shorter hours was the first one to be pushed through under union pressure, and the industry viewed this as an ‘epoch-making’ event in its labour history.35 34 Japan Spinners’ Association. ‘Present Conditions & Problems of the Cotton Industry in Japan: Labour Situation’, Nihon Bōseki Geppo (Monthly Report of the Japan Spinners’ Association) 223 (July 1965): 76–78. 35 Japan Spinners’ Association, ‘News & Current Problems on the Japanese Cotton Industry’, Nihon Bōseki Geppo (Monthly Report of the Japan Spinners’ Association) 137 (May 1958): 4–5. 328

Japan From the mid-1950s up until the mid-1970s the majority of textile companies maintained a six-day working week. Thereafter a five-day working week system was implemented. Textile unions, given new legal rights under the Occupation, were also very active in improving working conditions, particularly during the early ‘reconstruction’ years. The most serious labour dispute in the industry during these years took place at the Ōmi Kenshi Spinning Co. in 1954.36 The dispute began on 2 June 1954 and was finally settled on 16 September, after 106 days and several unsuccessful attempts at mediation, including by the Ministry of Labour. The dispute came to be seen, both within Japan and abroad, as a ‘human rights strike’ against the company’s working conditions and labour management. The Ōmi Kenshi incident showed that some past labour practices persisted, and was influential in subsequent efforts to coordinate labour conditions across the entire textiles industry. The dispute also showed the growing strength of unions. Reorganized after 1945, the national textile union Zensen had 67,000 members in 1946, but 600,000 by 1973.37 A large-scale wage-strike by textile workers in 1955 heralded the emergence of annual shuntō (‘spring labour offensive’) unified wage-rise negotiations and bargaining between unionized workers and management, a continuing feature of post-war Japan. From the 1960s onwards, the main labour–management disputes in textiles were confined largely to wage issues, and for the most part reached amicable annual settlements. After the ‘turbulence’ of the 1940s and 1950s, subsequent decades saw relatively ‘harmonious’ labour–management relations.

The workers At its postwar peak in the mid-1960s the textile industry employed over 1.3 million workers (not including large numbers employed in the separate apparel industry). However, by 2000 worker numbers had declined to some 230,000 workers, reflecting not only the industry’s overall decline but also the move to capital-intensive synthetic textile production. The composition of the textile factory workforce continued throughout the post-war years to show a high concentration of female labour, particularly in spinning and weaving. As in the pre-war years, until the 1980s the core workforce consisted of young female workers residing in dormitories and working only short periods of time for the industry. However, analysis of the age, marital status and employee status of female workers points to a growth in the number of older, married, non-permanent female workers. This key change in the composition 36 Japan Spinners’ Association, Labour Situation in the Japanese Cotton Spinning Industry – with reference to the Ōmi Kenshi Strike, Osaka, Research Division, Japan Spinners’ Association (September 1954). 37 Zensendōmei (National Federation of Textile Workers’ Union), Danjokyō ni Shigoto to Katei ga Ryōritsu dekiru Kankyō Seibi e mukete (Towards Maintenance of a Male–Female Work Environment where Work and Home are Compatible) (Okayama, Zensen Chūō Kyōiku Sentā (Zensen Central Education Centre), 1997), pp. 12–13. 329

The Ashgate Companion to the History of Textile Workers of female labour within the industry (and nationally in Japan) took place from the 1960s.38 This was because during the high growth years legal, educational and socioeconomic factors combined to create a rapidly changing labour market environment in Japan. The traditional pool of young female labour (school-leavers) was declining, while a unified national recruitment system had been established with tight controls over the recruitment of young workers.39 Textiles was in relative decline, facing competition for labour from other manufacturing industries (particularly electronics). Despite this changed and increasingly regulated environment, the textile industry continued primarily to target young female labour. However, although it effectively channelled a significant proportion of the declining pool of available junior high school graduates into textile employment, it became increasingly difficult to secure sufficient numbers of such workers as a core workforce. The response was to implement special systems for the recruitment of older female workers (primarily employed as a ‘supplementary’ or ‘complementary’ workforce) from the mid to late 1960s, diversifying the composition of female labour within the industry. Both employer attitudes and strong social perceptions regarding domestic roles for older women contributed to their employment as ‘non-regular’ workers. The use of less systematized and temporary recruitment programmes to employ such workers highlights the significant economic contribution of older women during this period in female-dominated industries like textiles, a contribution that was often ‘hidden’ or not fully recognized. The substantial use of seasonal labour meant that some rural women had a ‘triple’ burden during these years, with supplementary factory (‘economic’) work added to the burden of domestic duties and agricultural work. Textile employers were at pains during the high growth years to extend and develop labour management systems vis-à-vis their female labour force.40 They not only had to combat the lingering pre-war jokō aishi image, but also had to respond to post-war socio-economic changes, not least to shifts in the supply and characteristics of the female workforce. Dormitories and education remained the two central strategies for managing young female workers, but there was substantial readjustment, diversification and development in the scope of such provision. Dormitory standards were subject to legislation from the late 1940s, but textile employers were also keen to revamp dormitories to overcome the old image. Advances from the 1950s often exceeded legal requirements, and dormitories became an important welfare benefit for young female workers, allowing them to assign a significant portion of their wages to savings and non-essential items of expenditure. Dormitories remained a focus of labour management strategies, constituting the key institution through which young workers could be integrated into the working environment. 38 Macnaughtan, Women, Work and the Japanese Economic Miracle, pp. 15–23. 39 This section is based on Macnaughtan, Women, Work and the Japanese Economic Miracle, ch. 3. 40 See Macnaughtan, Women, Work and the Japanese Economic Miracle, ch. 4. 330

Japan The education system developed by textile firms for young female workers after 1960 was particularly impressive given that most were only expected to work for a few years (unlike male workers who were increasingly expected to remain with the company for their lifetime). Education for female textile operatives moved away from traditional notions of ‘bridal training’ towards meeting the standards laid down by the national education system.41 Senior high schools were built by employers in factory grounds to recruit a core labour force of young female junior high school graduates annually in the spring, predominantly from rural areas where families were still less likely to be able to send their daughters on to further education. These factory schools were increasingly registered with the Ministry of Education as ‘non-regular’ senior high schools and improved the quality of their teaching staff and education curriculum. From the mid-1970s the industry also branched into tertiary education, offering factory girls the opportunity to attend junior colleges or to train as nurses, teachers, book-keepers or beauticians. This investment in education for young female textile operatives was a recruitment ‘pull’ into the industry, an attempt to ‘lock’ workers into several years of continuous employment while they graduated through company-sponsored educational programmes. This significant investment in dormitories, education and welfare increased labour cost, and combined with a rapid rise in real wages during the high growth years removed the labour cost competitiveness the industry had enjoyed in the first half of the century. Japan’s textile industry therefore made substantial efforts to improve employment conditions during the high growth decades. While improvements were prompted by the implementation of national legislative reform, including regulation of work hours, recruitment, standards for the workplace and dormitories, and specific protection for women workers, the industry also did much to provide an improved environment for textile workers over and beyond that required by law. Particularly notable was the provision of a full non-working environment and opportunities for educational and personal development.

Conclusions Textile production was widespread in pre-industrial Japan, and this development laid much of the basis for the industrialization process. However, the evolution of the Japanese textile industry in the pre-war period generated structural rigidities and a degree of path dependence. For textile employers through much of the twentieth century the cost of change was often deemed to be greater than adhering to established practice. However, criticism of pre-war conditions also encouraged the industry to seek new ways forward and to endeavour to dispel its ‘exploitative’ image. While demonstrating considerable continuities, therefore, the Japanese textile industry has also seen many changes, evolving from quasi proto-industry 41 See Macnaughtan, Women, Work and the Japanese Economic Miracle, ch. 5. 331

The Ashgate Companion to the History of Textile Workers to the leader of the country’s modern industrialization and economic growth, then to decline in the post-war decades in the context of the challenges of a competitive global market.

332

13 Mexican textile workers: from conquest to globalization Jeffrey Bortz

During the conquest of Mexico, Hernan Cortes wrote to Charles V that ‘the [Indigenous] women of rank wear skirts of very thin cotton’. Prior to the arrival of Spaniards, the indigenous peoples of Mesoamerica indeed wore cotton clothing and Aztec warriors often used cotton armour, though it protected little against Iberian steel. Pre-Hispanic Mexico lacked sheep and therefore wool. Disdaining the Indians, the conquering Spaniards brought cotton and wool from Europe. They initially imported their cloth from Europe rather than obtaining it from native producers. Throughout the long colonial period (1519–1821), the upper classes continued to import cloth and clothes. However, the growth of a Spanish society in the New World quickly led to a local market for cotton and wool textiles and the development of a textile industry. The high cost of shipping in the sixteenth and seventeenth centuries served as a natural protection for locally made, low-cost and sometimes low-quality products. While Indian production of cotton continued within native communities, Spanish tastes demanded a different kind of product that local artisans were only too eager to supply. Three forms of production characterized the industry: household production, which the Spaniards called telares sueltos; small non-household artisan production, trapiches or obradores; and larger workshops, obrajes, with both forced and free labour. The telares sueltos (literally, ‘scattered looms’) could be found among families of the poor Spaniards, castas, and Africans who produced for the Spanishspeaking world. Over time, in both wool and cotton, comerciantes came to control the trade in raw materials, often subordinating household producers. Adding to this subordination, many weaving families found themselves dependent on the credit provided by these merchants. The households themselves often did not differentiate between private and work spaces, so that the looms simply formed part of the family’s useful belongings. Indigenous households continued to produce  

Hernan Cortes, ‘The First Letter’, Letters From Mexico (Yale, 1986), p. 30. Manuel Miño Grijalva, Obrajes y Tejedores de Nueva Espana 1700–1810 (Mexico, 1998), p. 27.

Map 13.1 Mexico

Mexican textile workers for the Indian world. Throughout the seventeenth and eighteenth centuries, some indigenous communities expanded their output to satisfy not only community needs but also local markets. Over time, however, non-indigenous production grew in proportion to indigenous production. For most household producers, production of cloth was a seasonal and part-time activity. Many worked in other activities, often agriculture. The evidence suggests an important gender/racial distinction. Within the indigenous communities, women predominated in weaving. Among non-Indians, men predominated although women participated. In the non-Indian households, the production of cloth occupied various family members; reports of the period suggest that it required six persons to card, spin, weave and carry out the other tasks. The ‘head of the family with many children, or a husband and his wife with household relatives’ jointly produced, one weaving, the other spinning, with others contributing to the project. Whether Indian or not, household producers were invariably poor. Contemporary reports described them as ‘the most unfortunate class’, ‘of little substance’, ‘a few poor people’ or ‘very poor Indians’. Perhaps less poor were those who ran small workshops, trapiches or obradores, often not part of the household. Many of these artisans belonged to their respective guilds, cotton or wool, and often employed a few assistants, some of whom were resident workers. At the bottom end, some owners were as poor as those who worked the telares sueltos, while at the top, some were small businessmen who not only employed assistants but lent money and material to household producers. Salvucci mentions one owner who employed ‘a creole and his castiza wife, an Indian couple, another Indian, and perhaps eight or ten labourers who lived in or near the obrador’. It is difficult to know the number of workshops in New Spain but a report for Michoacan in 1800 indicated 250 in Valladolid with 350 operarios, 15 more in San Juan Zitacuaro, 10 in Taximoaroa with 33 operarios, and 2 in Angamacutiro with 19 operarios. This suggests the ubiquity of such workshops throughout the colony. Individual obrajes were much larger than trapiches but as a group employed less people and generated less product than household and artisan production. Mexico City, Puebla/Tlaxcala and the Bajio, Mexico’s largest markets, not surprisingly concentrated the largest number of obrajes. Local society treated the owners as artisans and labelled the workers operarios or sirvientes. Early eighteenth-century estimates suggest 49 obrajes and 2,200 workers in Mexico City, and 32 obrajes with 1,440 workers in Puebla. There were obrajes for both wool and cotton (as well as  Richard J. Salvucci, Textiles y Capitalismo en Mexico (Mexico, 1992), p. 13.  Miño Grijalva, Obrajes y Tejedores, p. 165.  Salvucci, Textiles y Capitalismo, p. 13.  Ibid., p. 16.  Rafael Carrillo Azpeitia, Ensayo sobre la historia del movimiento obrero mexicano, 1823–1912 (Mexico, 1981) p. 104.  Salvucci, Textiles y Capitalismo, p. 88.  David J. McCreery, The Sweat of Their Brow: A History of Work in Latin America (Armonk, 335

The Ashgate Companion to the History of Textile Workers silk) and there is some evidence that Hernan Cortes himself may have established one of the first wool obrajes among his properties in Cuernavaca.10 Table 13.1 Textile workshops and looms in New Spain, 1759 City

Workshops

Looms

Mexico City

15

195

Queretaro

24

153

San Miguel el Grande

4

65

Salvatierra

3

26

Valladolid

2

20

Puebla

12

75

Cholula

16

39

Tlaxcala

8

23

84

596

Total

Source: Manuel Miño Grijalva, Obrajes y Tejedores de Nueve Espana, 1700–1810 (Mexico, 1998), p. 33.

The larger obrajes carried out a wider range of textile activities – carding, spinning, weaving, dyeing and finishing – than the smaller workshops. There is little evidence, however, that they were more modern or more productive, and in fact some studies indicate a decline in obraje production in the late eighteenth century.11 In the wool obrajes the technology ‘had been imported whole cloth, so to speak, from late medieval Europe and experienced little innovation before the end of the eighteenth century’.12 Most workers did not like the obrajes, whether cotton or wool.13 Work often began at 4 a.m., always under the direction of supervisors. Initially, Indian and unfree labour predominated. During his visit to Mexico in 1803, von Humboldt visited some obrajes, noting ‘the unhealthiness of the situation and the bad treatment

2000), p. 68. 10 Carrillo Azpeitia, Ensayo, p. 60. 11 See Salvucci, Textiles y Capitalismo, pp. 39–44, and Miño Grijalva, Obrajes y Tejedores, p. 75. 12 McCreery, Sweat of Their Brow, p. 69. 13 Salvucci, Textiles y Capitalismo, p. 97. 336

Mexican textile workers to which the workmen are exposed … all appear half naked, covered with rags, meagre and deformed. Every workshop resembles a dark prison.’14 By the seventeenth century, Blacks and castas had mostly replaced indigenous labour, though a mix of free (wage) and unfree labour continued. Over time the owners employed various forms of unfree labour: encomienda Indians, slaves, debt peonage and convict labour, the mix varying according to period and place. In Queretaro in 1800, perhaps 60 per cent of the workers were nominally free and 40 per cent unfree.15 The owners didn’t find convict labour particularly useful; their employment was due more to government pressures than the desires of owners. The obrajes varied in size, with some employing a hundred or more workers, of whom the majority were male. About 6–20 per cent were apprentices. Although law regulated the use of women in obrajes, it was permitted. Society and culture rather than technology determined male predominance. In Mexico, textiles, particularly cotton, became man’s work. Table 13.2 Official estimate of textile industry in New Spain, 1801 Obrajes

39

Looms of Spaniards and castas

4,440

Looms of Indians

3,369

Source: Juan Chavez Orozco, Monografia Economico Industrial de la Favbricacion de Hilados y Tejidos de Algodon (Mexico, 1933), p. 34.

In calculating the local market, Richard Salvucci has estimated that imports satisfied 17 per cent of local demand (the luxury market), subsistence for village use satisfied about 40 per cent of the demand, so that the monetary market for production by obrajes, obradores and telares sueltos constituted about 40 per cent of total needs in late colonial times.16 Textiles were an important but small part of an economy in which mining generated the greatest monetary wealth and agriculture employed the most people. Only a small fraction of the population worked in textiles, and most of them worked at home or in small artisan shops. In the cities, however, textile artisans often comprised a significant percentage of the total artisan population. In Mexico City in 1788, more than half the artisans (51.2 per cent) worked in textiles, broadly defined.17 14 Alexander von Humboldt, Political Essay on the Kingdom of New Spain (Norman, 1972), p. 189. 15 Salvucci, Textiles y Capitalismo, p. 115. 16 Ibid., p. 19. 17 Sonia Perez Toledo, Los Hijos del Trabajo (Mexico, 1996), p. 138. 337

The Ashgate Companion to the History of Textile Workers Skilled artisans organized and regulated their work and commerce through guilds, or gremios, inherited from the Spanish world. At the end of the eighteenth century, there was a guild for the weavers of Mexico City and another for the tailors, among the many artisan guilds that played an important role in city life and governance. The guilds brought together owners, master workers and apprentices, where they mediated rather than fomented class conflict. Within the workshops, however, some differentiation existed between maestros, oficiales and aprendices. The owners of workshops called themselves maestros and directed the work of the others in the shop. A formal system existed in which apprentices served with masters for three to five years, then took an examination to become officials. After another period, at least five years, they took another exam, paid a serious tax, then joined the ranks of masters. It is less clear, however, how rigorously this European system functioned in the New World.18 Formal distinctions often meant little when the workshop was tiny or employed only family members, but during the nineteenth century the maestro became more strongly identified as an owner in a capitalist sense, while oficiales and aprendices increasingly understood their role as wage workers.19 Operarios, the colonial term for obraje workers, became the category of choice to describe textile workers in nineteenth-century factories.

Industrialization, 1821–1910 Between 1810 and 1821, Mexico fought its revolution for independence. The prolonged violence, lack of a stable state and England’s cheap exports from its newly industrialized textile industry combined to hammer Mexican textiles. With the opportunity to purchase cheaper and better quality imported cloth, some Mexicans diminished their purchases of the locally produced material. Textile artisans comprised half the artisan population of Mexico City in 1788 but only a third in 1842.20 Meanwhile, the apparent failure of the new state during the first half century of independence – numerous invasions, loss of national territory, extreme political instability – led some Mexicans to look towards European industrialization to solve Mexico’s problems. Lucas Alamán, a minister in the government of General Anastasio Bustamante (1830–1832), founded the Banco de Avío, a development bank, in 1830. Esteban de Antuñano, who had purchased the hacienda of Santo Domingo on the Atoyac River in Puebla city, used the bank’s financing to transform the existing wheat mill into a cotton mill. He imported the machinery from England and opened Mexico’s first factory and cotton textile mill, La Constancia Mexicana, on 7 January 1835.21 18 Jorge Basurto, El Proletariado Industrial en Mexico (1850–1930) (Mexico, 1975), p. 57. 19 Carlos Illades, Hacia la Republica del Trabajo (Mexico, 1996), pp. 42–43. 20 Perez Toledo, Los Hijos del Trabajo, p. 138. 21 Jesus Rivero Quijano, La Revolucion Industrial y La Industria Textil en Mexico (Mexico, 1990), p. 103. 338

Mexican textile workers The mill spun crude cotton which it sold to artisans who transformed it into cloth. Antuñano then built La Economia, also on the Atoyac River.22 Both mills used water power to drive the machines. In 1836 Alamán entered into a business agreement with the Legrand brothers to build a textile factory, Cocolapam, just across the state boundary from Puebla state in Orizaba, Veracruz.23 ‘Cocolapam’ became the country’s largest mill in the first half of the century. Meanwhile, Gumersindo Saviñón transformed ‘El Mayorazgo’ in Puebla from flour to cotton production.24 These early industrialists followed similar paths, transforming flour into cotton mills, utilizing nearby rivers for power and importing English machines and technicians for the new factories. Antuñano believed that industrialization would lead Mexico out of underdevelopment, claiming that ‘The modern machines established in Mexico and worked by Mexicans … will develop our cotton agriculture to an incredible degree … will bring together in the country many millions of pesos, whose circulation will make prosper all the branches [of industry].’25 Cotton textiles formed Mexico’s first factory industry and remained the country’s most important factory industry throughout the nineteenth and into the twentieth century. Table 13.3 Growth of the Mexican cotton textile industry, 1843–1927 Factories

Spindles

Looms

Workers

1843

59

106,708

2,609

n.d.

1862

57

133,122

n.d.

n.d.

1898

118

468,574

13,944

21,960

1907

142

694,000

23,500

33,000

1917

99

573,072

20,489

22,187

1927

132

777,380

29,290

41,008

Sources: Fernando Pruneda and Miguel A. Quintana, ‘Estudio sobre la modernización de la industria nacional textil de algodón…’, Secretaría del Trabajo y Previsión Social, Mexico City, Abril 1943, p. 37, unpublished monograph in Archivo General de la Nación (AGN), Archivo Gonzalo N. Robles, Caja 76, Exp. 9; for 1907 figures, C. Reginald Enock, Mexico, It’s Ancient and Modern Civilisation History and Politcal Conditions Topography and Natural Resources Industries and General Development (London, 1909), p. 337. 22 Mariano Torres Bautista, Testamento del Administrador (Puebla, 1989), p. 8. 23 Rivero Quijano, La Revolucion Industrial, p. 111. 24 Fernando Pruneda and Miguel A. Quintana, ‘Estudio sobre la modernización de la industria nacional textil de algodón…’, Secretaría del Trabajo y Previsión Social, Mexico City, Abril 1943, p. 24, unpublished monograph in Archivo General de la Nación (AGN), Archivo Gonzalo N. Robles, Caja 76, Exp. 9. 25 Esteban de Antuñano, La Industria del Algodón en México (Puebla, 1833), p. 74. 339

The Ashgate Companion to the History of Textile Workers Although wealthy Mexicans continued to prefer the better quality of imported cloth and clothes, and despite the persistence of subsistence agriculture and isolated communities that lived mostly outside the market economy, many poor Mexicans bought the cheap cotton cloth, or manta, produced by the factories. The local cloth was less expensive than imports because of newly established government tariffs and the de facto protection of international transport costs. The creation of a cotton textile industry and a true factory system was Mexico’s industrial revolution, although one limited to a single industry and based on imported rather than locally developed technology. In England, locals invented, developed and refined the technology that substituted machines for people.26 On the cutting edge of manufacturing productivity, their machines produced goods with competitive prices in domestic and foreign markets. In Mexico, however, capitalists did not innovate. They instead imported their machinery, techniques and even the most skilled workers. The technology was modern for Mexico but not for Europe and was never truly a function of local conditions. This would remain a constant in the Mexican textile industry as technology was simply what foreigners had to offer. When federal Labour Inspector Miguel Casas visited the Rio Grande mill in Jalisco in 1913, he noted ‘As can be seen by the brand name of each machine, these are from England.’27 Although the early industry suffered from the country’s political instability, persistent lack of capital, poor transport systems and anaemic internal market, it grew steadily, from eight mills in 1836 to 62 in 1845.28 These ‘modern’ factories were much larger than the old workshops. Saviñón’s ‘El Mayorazgo’ had 2,400 spindles and Antuñano’s ‘La Constancia Mexicana’, 7,680.29 The factories required larger and more secure supplies of raw cotton than the old obrajes, contributing to the shift in the countryside from small ranchos to larger agricultural units. Within the factories, not only production, but also total factor productivity, labour productivity and capital–labour ratios increased significantly throughout the nineteenth century.30 By the end of the 1830s, the new cotton textile industry employed 8,741 workers, with 5,212 spinners and 2,074 weavers. The spinners earned 3.50 reales a week and the weavers 2.50, at a time when rural labourers earned only 2.00.31 Most mill hands were men, continuing the colonial pattern of textile employment. In La Magdalena in 1853, there were 190 male workers and 23 female. Tizapan reported 197 male and 15 female. Both mills employed few children; in Tizapan, only eight, of 15 years old or younger.32 26 R.A. Buchanan, The Power of the Machine (London, 1992), ch. 5. 27 Author’s translation, Miguel Casas, Informe, February 10, 1913, AGN, DT, Caja 6, Exp. 27. 28 Rivero Quijano, La Revolucion Industrial, p. 122. 29 Pruneda and Quintana, ‘Estudio sobre la modernización de la industria nacional’, pp. 24–25. 30 Armando Razo and Stephen Haber, ‘The Rate of Growth of Productivity in Mexico, 1850–1933’, Journal of Latin American Studies 30 (1998): 481–517, at p. 502. 31 Carrillo Azpeitia, Ensayo, p. 120. 32 Mario Trujillo Bolio, Operarios Fabriles en el Valle de Mexcio 1864–1884 (Mexico, 1997), 340

Mexican textile workers The new factories changed little the industry’s geographic patterns, with two exceptions. As before, the owners built mills near the country’s largest markets, Mexico City, Puebla and the Bajio. Now, however, the new factories depended on water power; the late nineteenth-century shift to electric turbines also required rivers. The Puebla city mills took advantage of the Rio Atoyac. Some textile industrialists found it productive to build large mills in a new textile area, Orizaba, in the Sierra of Veracruz, astride the road from Puebla to the port of Veracruz. The mills took advantage of the effluents from the nearby Sierra Madre Oriental, particularly Rio Blanco. Orizaba became one of the two most important mill towns in Mexico. The other, Atlixco, about 30 km south of Puebla, benefited from streams flowing down the nearby volcano, Popocatepetl. Finally, the industry became more concentrated, so that the scattered workshops, seen with the example of Michoacan, ceded to a concentration of factories in a few textile zones. Table 13.4 Regional percentage share of national production of cotton textiles, 1880–1910 State

Before 1880

In 1910

Puebla

23

32

Distrito Federal

14

12

Veracruz

9

21

Tlaxcala

4

6

50

29

Others

Source: Dawn Keremitsis, La Industria Textil Mexicana en el Siglo XIX (Mexico, 1973), p. 123.

The early industrialists often built mills in old haciendas. The new establishments acquired some of the characteristics of the haciendas: large, ornate buildings, workers’ houses on the premises, paternalism in running the enterprise and a chapel, or at the very least, a space for the Virgin of Guadalupe, an expression of the society’s and the workers’ profound Catholicism. In 1877, for example, the administrator of Cocolapam, Alberto G. Morphy (sic) ordered the construction of a ‘Capilla que de perpetuamente dedicada a la Stma Virgen de Guadalupe’.33 Celebration of the local saint’s day became an important part of workers’ culture. Mechanized cotton textile mills created a new set of social relations of production in Mexico, owners who did not work and workers who did not own, neither group fitting comfortably in the old guilds. Although the early owners tried to maintain pp. 106–119. 33 To Yllmo Senor, noviembre 26 de 1877, Archivo del Catedral Sn Miguel Arcangel, Orizaba. 341

The Ashgate Companion to the History of Textile Workers the paternalistic labour relations systems of the old haciendas and workshops, the growing scale of factories inevitably brought more impersonal and hierarchical relations of work. The owners, however, continued to believe in their right to run the mills as they wished. Relations of domination became even more complex when the larger mills began to construct housing, creating large mill towns in Atlixco and Orizaba, and smaller ones near large factories in other areas. A radical paper complained in 1878 that the owners of the three mills near Queretaro, including ‘El Hercules’, ‘are those that dictate the laws, make them obeyed, punish the transgressors, judge delinquents’.34 Inevitably, mill hands began to organize the mills, though slowly and unevenly, in part because trade unions lacked legal recognition and collective contracts were unknown. Influenced by European socialist ideas, in 1865 textile workers organized their first Mexico City union, the Sociedad Mutua del Ramo de Hilados y Tejidos del Valle de Mexico.35 Weavers in two local mills, San Ildefonso and La Colmena, then carried out the first textile strike in June 1865, against wage reductions and for shorter hours of work.36 During the 1860s and 1870s, labour organization spread among textile workers, who suffered from 15-hour workdays, arbitrary fines and capricious supervisors.37 Some of these organizations attempted walkouts or strikes. Known as societies, they also fought the abuse of women textile workers by male supervisors, defending traditional female virtue with traditional male honour in a working environment that challenged if not undermined traditional society.38 In 1871–1872, influenced by the recent establishment of the International Workingman’s Association in London, a group of activists founded the Gran Circulo de Obreros de Mexico (GCOM). The organization quickly established branches in textile factories in the country’s textile zone, the Mexico City–Puebla–Orizaba corridor.39 Half the GCOM branches were in textile factories but often they did not represent a majority of workers. Between 1821 and 1876 Mexico experienced a succession of short-lived and failed governments, military coups, foreign invasions and civil wars. In 1876, General Porfirio Díaz, the Liberal hero of the wars of the Reform (1857–1867) took power in a successful military uprising, then surprised the country by imposing the longest period of political stability in Mexico’s independent history. Buoyed by a stable government and the late nineteenth-century boom in international commerce, the Mexican economy experienced its first sustained economic growth since independence. Between 1893 and 1907 the Mexican economy ‘grew

34 El Hijo del Trabajo, marzo 5, 1878, cited in Basurto, El Proletariado Industrial en México, p. 50. 35 Trujillo Bolio, Operarios Fabriles, p. 172. 36 Carrillo Azpeitia, Ensayo, p. 186. 37 Trujillo Bolio, Operarios Fabriles, p. 175. 38 Ibid., p. 180. 39 Juan Felipe Leal, Del Mutualismo al Sindicalismo en Mexcio: 1843–1910 (Mexico, 1991), pp. 24–26. 342

Mexican textile workers more rapidly than that of the United States’.40 Investments in railroads and the consequent decline in transport costs benefited both the national economy and the textile industry, which needed to ship cotton from long distances and send its products throughout the large country. The railroad lowered the cost of shipping one ton of cotton from Mexico City to Queretaro from 61 pesos in 1877 to 3 pesos in 1910.41 Textiles grew steadily from 1876 to 1890, then even more strongly during the 1890s as companies, factories and machines became larger and more modern. From 86 mills, 8,132 looms, 234,386 spindles and 10,871 workers in 1877, capacity jumped to 141 mills, 18,553 looms, 600,707 spindles and 28,192 workers in 1900.42 Foreign owners, particularly French and Spanish, played a significant role in the late nineteenth-century expansion of the industry. French immigrants developed some of the country’s largest mills, San Antonio Abad in Mexico City, Metepec in Atlixco and Rio Blanco and Santa Rosa in Orizaba. The large Metepec plant near Atlixco opened in 1902 with 36,852 spindles and 1,570 looms.43 By 1911 the Frenchowned textile giant CIDOSA became Mexico’s eighteenth-largest corporation and largest industrial corporation, whose size was exceeded only by a few banking, railroad and mining firms. CIDOSA owned the Rio Blanco, Cerritos, San Lorenzo and Cocolapam mills in Orizaba, and by 1912 controlled 13.5 per cent of the national market.44 Spanish investors controlled much of the Puebla textile industry. As in the colonial period, many of the larger textile firms grew in association with commercial enterprises, which invested in them in order to ensure a steady supply of product to their large department stores in Mexico City.45 By 1900, 44 per cent of Mexico’s hydroelectric power was generated by plants installed in textile factories, driving automatic looms and high-speed spindles.46 Metepec, Santa Rosa, Río Blanco and the other large mills employed the new turbine power to operate the modern, automatic machinery, requiring larger capital investments. This led to an increase in production, productivity and factory size as the industry experienced horizontal and vertical integration, increasingly capitalintensive production and ownership consolidation.47 New companies with larger and more modern factories came to dominate the industry. From the individually owned private companies of the mid-nineteenth century, there was a turn to limited liability joint stock companies, with some giant companies publicly traded on the new Mexico City exchange. The principal firms – the Compania Industrial de Orizaba (CIDOSA), the Compania Industrial Veracruzana (CIVSA), the Compania 40 John H. Coatsworth, Growth Against Development: The Economic Impact of Railroads in Porfirian Mexico (DeKalb, 1981), p. 4. 41 Carmen Ramos Escandon, ‘Working Class Formation and the Mexican Textile Industry: 1880–1912’ (Ph.D. diss., State University of New York at Stony Brook, 1981), p. 45. 42 Mexico, Instituto Nacional de Estadisticas, Geografia e Informatica, Estadisticas Historicas de Mexico (Mexico, 1994), p. 616. 43 Dawn Keremitsis, La Industria Textil Mexicana en el Siglo XIX (Mexico, 1973), p. 117. 44 Stephen Haber, Industry and Underdevelopment (Stanford, 1989), p. 56. 45 Rivero Quijano, La Revolucion Industrial, p. 344. 46 Ramos Escandon, ‘Working Class Formation’, p. 47. 47 Haber, Industry and Underdevelopment, pp. 190–193. 343

The Ashgate Companion to the History of Textile Workers Industrial de San Antonio Abad (CISAASA), and the Compania Industrial de Atlixco (CIASA) – ‘were mammoth enterprises, even by US standards’.48 Outside of mining, cotton textile manufacturing was Mexico’s largest industry with onequarter of the total value of manufactured goods.49 In 1909 the British engineer C. Reginold Enock noted that Mexican ‘cotton mills are among the foremost in the world, and their large capacity and splendidly-built factories are a source of surprise to the European or American traveller’.50 The growth of the industry brought workers, labour organization and labour strife to the forefront of Mexican society. In 1800, there was no true factory proletariat in Mexico. The obrajes were declining. In 1900, the 141 textile factories employed almost 30,000 workers, many of them in modern factories with more than 1,000 mill hands each. Rio Blanco employed 2,000 and Metepec in 1908, 1,330.51 Many of the workers came from rural backgrounds, and frequently left their jobs; labour turnover was a great problem for the mills.52 The labour force was overwhelmingly male; women worked in some factories and in some jobs but were almost always in unskilled positions and paid less than men, usually significantly less. It was more common to find women workers in associated industries and in the smaller mills, particularly those producing cotton goods and clothing, the boneterías that employed seamstresses in numerous small workshops. Furthermore, working outside the home identified maleness in Mexico, whereas home defined women, though the reality for working women often defied the social stereotype. Although there were mills scattered throughout the Republic, most textile workers lived in four places along a single corridor that ran from Mexico City through Atlixco and Puebla to Orizaba. This allowed for great communication among workers. For them, if life in the factory was better than the often oppressive conditions in the countryside, it was still a sharply stratified existence in which owners and bosses, many foreign and white, looked down on skilled and unskilled workers alike, mostly Mexican and dark. When Metepec mill hands founded a union in 1907, they described themselves as ‘we humble workers’.53 These ‘humble workers’ in cotton textiles, the country’s first and largest industrial proletariat, not surprisingly led the way in creating labour organizations and trade unions. Juan Felipe Leal has argued that most large mills had a mutualist society, which in times of conflict or strikes could function as a union. Textile workers carried out at least 30 strikes between 1881 and 1895.54 In 1884 the Puebla textile organizations formed a regional labour federation, the Federacion Obrera. 48 Razo and Haber, ‘Rate of Growth of Productivity’, p. 512. 49 Rodney Anderson, Outcasts in Their Own Land (De Kalb, 1976), p. 100. 50 Enock, Mexico, p. 337. 51 Moises Gonzalez Navarro, Las Huelgas Textiles en el Porfiriato (Puebla, 1970), p. 226; Leticia Gamboa Ojeda, La urdimbre y la trama, Historia social de los obreros textiles de Atlixco, 1899–1924 (México, 2001), p. 66. 52 Rosalina Estrada Urroz, Del telar a la cadena de montaje (Puebla, 1997), p. 63. 53 Leandro Aguilar y Sanchez et al., ‘A los Obreros...’, noviembre 29 de 1907, Archivo Municipal de Atlixco (AMA), Presidencia, Gobernacion, 1907, Caja 164. 54 Felipe Leal, Mutualismo al Sindicalismo, pp. 60–61. 344

Mexican textile workers During the latter half of the Porfiriato, the strikes became larger, many of them sparked by owners’ attempts to lower wages. Frequently political authorities intervened, sometimes to mediate, sometimes to repress. In 1898 mill hands in two State of Mexico mills (La Colmena and El Barron) struck when the owners lowered wages, as did mill hands in El Mayorazgo in Puebla in 1900 and in La Hormiga in Mexico City in 1901. In the State of Mexico strike, the governor mediated but the workers lost. With the El Mayorazgo strike, mill hands in the other Puebla City mills joined the strike, but the owners defeated the workers because the large supplies of stocks in the warehouses outlasted the hungry families. Only in La Hormiga did the workers win, because the local political authority sided with them.55 Increasing mill size and more distant ownership became explosive elements when the Porfirian economy turned downward in the last years of the dictatorship. The most famous Porfirian strike of all was that at Rio Blanco in 1906. In April a small group of mill hands founded the Gran Circulo de Obreros Libres. When labour agitation spread to neighbouring Puebla, textile owners there founded the Central Industrial Mexicano (CIM). The next month the CIM issued regulations for textile mills in the Puebla area ‘which were intended to halt increasing labour agitation’.56 The recently organized workers responded with a December strike in Puebla and Tlaxcala, to which the owners answered with a massive nationwide lockout on 22 December. The Diaz government intervened to negotiate a settlement, albeit one that favoured the owners rather than the workers. In Tlaxcala the governor sent the strike leaders to prison, effectively ending the movement there.57 Throughout the country defeated textile workers sullenly returned to work, except in Orizaba. At Rio Blanco angry strikers confronted scabs entering the mill. Porfirio Diaz ended the strike with a labour decree on 4 January 1907.58 On 7 January, troops intervened, ‘killing scores in the largest single massacre of non-indigenous people in the history of the regime’.59 Although the literature has treated the 1906/1907 conflict as a symptom of late Porfirian discontent, it was in fact a fight about work rules, increasingly inevitable in an industry of large factories. The owners’ rules included the 14-hour day and six-day week, heavy fines for flawed work and a prohibition against allowing guests in the workers’ homes on mill property.60 With the slaughter in January 1907, the owners succeeded in imposing their will, for the time being. 55 Ibid., p. 96. 56 David W. Walker, ‘Porfirian Labor Politics: Working Class Organization in Mexico City and Porfirio Diaz, 1876–1902’, The Americas 37 (3) (1981): 257–287, at p. 282. 57 Blanca Estela Santibanez Tijerina, ‘Las Huelgas Textiles Tlaxcaltecas Durante el Porfiriato’, Universidad Autonoma de Puebla, 1996, p. 22. 58 A copy of the laudo is in AGN, Gobernación-legagos, Caja 817, Exp. 1, Correspondencia Particular del Ministro de Gobernación, n.d. 59 John M. Hart, Anarchism and the Mexican Working Class, 1860–1931 (Austin, 1987), p. 97. 60 Karl B. Koth, ‘Not a Mutiny but a Revolution: The Rio Blanco Labour Dispute, 1906–1907’, Canadian Journal of Latin American and Caribbean Studies 18 (35) (1993): 39–65, at p. 49. 345

The Ashgate Companion to the History of Textile Workers Thus at the end of the Porfiriato, cotton textile workers in the largest mills confronted serious problems of hierarchy, authority, low and sometimes declining wages, and poor working conditions, with no formal institutionalized labour relations system with which to seek solutions. Mexico had a factory system and a growing industrial proletariat but no body of labour legislation. Law recognized neither unions nor collective contracts. Labour organizations were barely and only sometimes tolerated. Government regulation consisted of the personal interest of local political bosses, state governors or the President himself. Owners, with their allies in government, ruled the factory.

Textile workers and Mexican revolution, 1910–1927 In November 1910, a wealthy landowner from the north, Francisco Madero, challenged President Porfirio Diaz and initiated the Mexican revolution.61 Following important military defeats, Diaz fled Mexico in May 1911, leaving the government to the interim presidency of Francisco Leon de la Barra, who turned it over to Madero in November 1911. Under Leon de la Barra, union activists took advantage of the unstable political situation to radicalize existing mutualist organizations in the textile mills while organizing those without unions. A concerned de la Barra proposed a federal labour office. Justifying the interim president’s concern, on 26 December 1911, just six weeks after Madero took office, cotton textile workers launched ‘the first labour action of great importance in the Republic’, a general strike that shut down the industry.62 In perhaps the most significant strike in Mexican history, mill hands closed the most important textile centres.63 On 20 January 1912, in order to get the workers back on the job, textile owners agreed to a 10 per cent pay increase, a reduction in the workday to 10 hours, and to organize a convention to solve the industry’s labour woes. The agreement marked a permanent change in labour relations in Mexican history. For the first time in the history of the Republic, workers in a key industry organized unions, carried out a successful nationwide general strike, forced the owners to cede significantly better wages and hours of work and did not suffer the usual repression, loss of jobs and imprisonment or exile of strike leaders. This was the moment that rank and file workers experienced the Mexican revolution as their

61 There is a vast historiography of the revolution but the modern reader would begin with Alan Knight, The Mexican Revolution (Lincoln, 1990) and Friedrich Katz, The Life and Times of Pancho Villa (Stanford, 1998). 62 El Imparcial (Mexico City), 21 January 1912, p. 1. 63 A strike by oil workers in the 1930s led to the nationalization of the industry, but the 1911 textile strike launched a workers’ revolution that radically transformed the Mexican labour regime. 346

Mexican textile workers revolution. They assaulted authority, stood together in their organizations and won. This had never happened in Mexican history, a lesson that nobody forgot.64 The promised textile convention met during July 1912 and ratified the workers’ gains while also drafting an industry-wide set of work rules and wage scale, both enforced by a government tax on textile mills. The two documents represented a de facto industry-wide collective contract, the first in a country that still lacked collective contracts. The July convention was also the first tripartite meeting (government, owners and workers) in Mexican history.65 The country’s new Labour Department claimed that the Convention resulted in ‘greater harmony between owners and workers’, concluding that such agreements should be spread to other industries.66 Over the next decades, the Mexican government would follow the cotton textile model – tripartite conventions and industry-wide collective contracts – to attempt labour peace in the most important industries. In February 1913, Victoriano Huerta, the head of the federal army, assassinated Madero and assumed the presidency. Venustiano Carranza, another wealthy landowner from the north, opposed Huerta and organized a resistance under the Constitutionalist banner, with the help of his military commander Alvaro Obregón. The more radical Pancho Villa and Emiliano Zapata, revolutionaries who had fought Diaz, organized their own military resistance to Huerta while also pursuing radical land reform. Mexico’s second revolution was more violent than the first until in July 1914 Huerta fled Mexico. The victorious revolutionaries soon fell to fighting among themselves, the Constitutionalists, Carranza and Obregón, against Villa and Zapata, further spreading revolutionary violence. By 1916, however, the Constitutionalists had the upper hand, allowing them to write a new Constitution, promulgated in February 1917. During these difficult years of revolution the textile industry suffered greatly. Nonetheless, worker militancy did not wane. Textile workers – cotton and wool – organized strikes during the 1912 convention, just after it concluded and then in the waning months of 1912. In July 1912, a small group of anarchists founded the Casa del Obrero Mundial, with the participation of radical textile unionists, among others.67 Casa organizers travelled to the textile regions arguing that ‘to get what they desire, the workers … must use guns and not laws’.68 Women workers, though a small percentage of the labour force, had participated in the struggles against authority since the Rio Blanco strike, and now engaged in a new militancy about their own grievances, many about male supervisors. The women in ‘El Hercules’ 64 Jeffrey Bortz, ‘Without Any More Law than Their Own Caprice: Cotton Textile Workers and the Challenge to Factory Authority during the Mexican Revolution’, International Review of Social History 42 (1997): 253–288. 65 Jeffrey Bortz, ‘The Genesis of Mexico’s Modern Labor Regime: The 1937–1939 Cotton Textile Convention’, The Americas 52 (1) (1995): 43–69. 66 marzo 29 de 1913, AGN, DT, Caja 47, expo. 7. 67 Rosendo Salazar, Las Pugnas de la Gleba (Mexico, 1972), p. 37. See also Hart, Anarchism and the Mexican Working Class, p. 114. 68 Enrique H. Hinojosa et al., 3/15/15, AGN, DT, Caja 104, Exp. 11. 347

The Ashgate Companion to the History of Textile Workers complained that supervisor Octavio Regalado ‘is so abusive with us … treating us badly, saying things that as women we cannot mention’.69 The assault on traditional authority did not wane under Huerta, as mill hands carried out numerous strikes and other disruptive actions. Many mill hands carried knives to work under their shirts, and on more than one occasion stabbed or killed a supervisor in a heated dispute. Others joined revolutionary bands. Still others joined the numerous and violent strikes and walkouts that shook the mills during this period.70 The labour actions continued strongly in the year and a half following the collapse of the Huerta government, despite the great violence and economic dislocation. Mill hands wanted owners to ‘know that the days of tyranny have passed’.71 Owners, factory directors and supervisors came to share the opinion of the Director of the Mirafuentes mill, that ‘to give in to the desires of the workers to fire [department heads] is to invite disorder and disaster, because without discipline and supporting them outside of any logic or proof, would make it impossible for us to work’.72 It was an admission that the workers’ revolution, the constant walkouts and strikes against supervisors threatened to destroy traditional authority in the mills. There was not a single large mill that did not suffer a strike, wildcat walkout, murder of a supervisor, burning of looms or some other militant labour action during the workers’ revolution within the revolution. The 1911 strike had forced government and owners to seek an institutional solution to conflict in the cotton mills. The continuing labour violence of 1912–1916 pressured revolutionary governments to follow the same path. Between 1914 and 1916, Constitutionalist military commanders slowly took control of various regions of the country. In the textile zones, they realized that they couldn’t win without solving the ongoing labour revolt and responded by issuing military labour decrees. The decrees constituted de facto labour law in a country that lacked such. Among the earliest and most radical decrees were those in Veracruz and Puebla, textile strongholds. Military commanders like Pablo Gonzalez, General en Jefe del Cuerpo de ejercito del Nordeste, and Cándido Aguilar, Gobernador y Comandante Militar, issued edicts that ratified what textile workers had won through their strikes, work actions and labour militancy: minimum wages, overtime pay, the legal recognition of unions and the right to strike, the right of unions to sign collective contracts, and better working conditions.

69 Ma Trinidad Resendiz, Ma Rosario Yanez, Ma Feliza Lemus and Ma Lina Marend to Sr. Antonio Ramos Pedrueza, agosto 3 de 1912, AGN, DT, Caja 8, Exp. 16. 70 See for example, the Mirafuentes strike, 1913, Departamento del Trabajo, Cuestionario para la estadistica de diferencias y huelgas, December 1913, AGN, DT, Caja 1, Exp. 4. 71 Martin Perez, Julian Suarez, Luis Viveros to Antonio Valero, Dir. del Depto. del Trabajo, October 8, 1914, AGN, DT, Caja 87, Exp. 2. 72 Joseph Taylor to Director del Departamento del Trabajo, May 12, 1915, AGN, DT, Caja 97, Exp. 14. 348

Mexican textile workers Table 13.5 Gender composition of cotton textile industry labour force, 1927 Region

State

Centre

Active factories

Men

Women

Children

Total

102

20,284

2,795

1,970

25,049

DF

24

4,622

1,471

381

6,474

Mexico

7

1,489

167

88

1,744

Puebla

53

9,929

365

1,269

11,563

Tlaxcala

7

1,579

…

129

1,708

North

18

2,980

1,355

217

4,552

Gulf

11

6,533

286

469

7,288

11

6,533

286

469

7,288

Pacific

13

2,315

1,819

215

4,349

Mexico

144

32,112

6,255

2,871

41,238

Veracruz

Source: Secretaría de Industria, Comercio y Trabajo, Monografía sobre el Estado Actual de la Industria en México (Mexico, 1929), p. 39.

When the Constitutionalists gathered in late 1916 to draft a new Federal Constitution, many of the military commanders and authors of labour decrees served as delegates to the Constitutional Convention. Their experience of the textile revolt and the labour decrees made Article 123, Mexico’s first labour code, inevitable. Article 123 of the 1917 Constitution formalized fundamental labour rights: the right to unions, to strikes, to the eight-hour day, to minimum wages and to vastly improved and mandated benefits. Because Mexican constitutional law requires implementing statutes, during the next decade the states amended their own constitutions to include the necessary labour provisions, with great latitude to expand on the federal protections for labour. The two states with largest textile industry – Veracruz (1918) and Puebla (1921) – were among the first to implement labour codes and their codes were among the country’s most progressive. The two codes recognized unions, collective contracts and the right to strike, virtually implementing the union shop. They also required owners to register work rules with state governments and placed such severe restrictions on firing that it became impossible for large firms to fire, or even discipline workers, without consent of the union. The Puebla Code made unions, rather than firms, responsible for hiring.73 73 On the military decrees and the state labour codes, see Jeffrey Bortz, ‘The Legal and Contractual Limits to Private Property Rights in Mexican Industry during the Revolution’, in Jeffrey Bortz and Stephen Haber, The Mexican Economy, 1870–1930: 349

The Ashgate Companion to the History of Textile Workers Thus the workers’ revolution in the textile industry, beginning with the general strike of 1911 and continuing with the ongoing violence in the mills, radically changed the labour regime not only in textiles, but in Mexico’s work world. In 1910 Mexico had no labour law, no legally recognized unions and no collective contracts. Owners frequently fired unionists and regularly defeated attempts to organize the mills. By 1921, workers in the textile and other industries enjoyed powerful unions that controlled hiring and firing, displacing the power of the owners.74 They also had the right to reduced hours of work, overtime pay, numerous holidays and a vast array of education, housing, vacation, maternal and medical benefits. The institutionalization of the labour relations system between 1917 and 1922 was accompanied by a rapid formalization of textile and other unions, as well as continued conflict in the mills, albeit increasingly characterized by intra-union squabbles. In May 1918, former members of the Casa del Obrero Mundial and other unionists founded the Confederacion Regional Obrera Mexicana (CROM), with a leader of the Mexico City electricians union, Luis N. Morones, as its head. Later that year Morones agreed to a secret pact of mutual support with Alvaro Obregón. When Obregón assassinated Carranza in 1920 and assumed the presidency of Mexico, he gave full government backing to Morones and the CROM, while Morones gave his political support to Obregón. From 1920 to 1928, Presidents Alvaro Obregón (1920–1924) and Plutarco Elías Calles (1924–1928), and CROM leader Luis N. Morones dominated Mexican politics and the labour movement. In February 1921, dissident groups from the CROM, surviving sections from the Casa del Obrero Mundial, members of the nascent Mexican Communist Party and some anarchist groups founded the Confederacion General de Trabajadores (CGT).75 About half the founding organizations claimed to be ‘revolutionary unionists’, by which they meant anarcho-syndicalists. During these years, the newly powerful textile unions led general strikes in Orizaba, Puebla, Tlaxcala and Mexico City. The successful 1919 Orizaba walkout implemented the union shop and significant wage increases. In 1924 the Orizaba workers struck again, backed by radical Governor Adalberto Tejeda. The strike brought a further round of wage increases, the establishment of formal job categories and classifications, and more extensive rules governing shop-floor behaviour. Meanwhile, the Veracruz state government forced industrialists to provide medical care for workers, imposing strict fines for non-compliance.76 The government-allied CROM found a bitter enemy in the more radical CGT, leading to violent labour wars in the textile towns in the early 1920s. Union leaders fiercely defended their positions, which now benefited from their control over jobs Essays on the Economic History of Institutions, Revolution, and Growth (Stanford, 2002), pp. 255–288. 74 Jeffrey Bortz, ‘Authority Re-Seated: Control Struggles in the Textile Industry during the Mexican Revolution’, Labor History 44 (2) (2003): 171–188. 75 Juan Felipe Leal, Agrupaciones y Burocracias Sindicales en Mexico 1906/38 (Mexico, 1985), p. 97. 76 Delegación Patronal de Oriz­aba to Villalobos, 8/24/38, 4, AGN, DAT, Caja 12, Exp. 3, 350

Mexican textile workers in the mills, their management of mandatory union dues and their great power in local politics. They organized armed labour gangs to physically attack rivals. The violence became a daily occurrence in Atlixco and Orizaba, particularly when the owners tried to establish Catholic unions against both the CROM and the CGT. There were many nights like that of Christmas eve, 1923, when four Atlixco workers from one group assaulted a unionist from the other band, beating his head with gun butts, breaking into his house to steal his money, then kicking his pregnant wife in the stomach until she suffered a miscarriage.77 Atlixco was representative of the violence and also the confusion among unionists in the textile industry. Baraquiel Martinez organized the local CROM affiliate in Atlixco, then abandoned it to form a CGT branch. The CGT then launched a general strike in 1921 ‘to protest … reductions in shifts and the number of machines assigned to workers, as well as to intervene in the current [inter-union] conflicts in Metepec and El Leon’.78 The success of this general strike was the high point of the CGT in Atlixco. For the next three years, the better-financed CROM labour gangs mostly displaced the CGT, which remained dominant only in the small El Volcan and El Carmen mills.79 By 1925 the CROM ruled Atlixco, the unions, the mills and the town itself. Table 13.6 Mexican cotton textile industry, average daily wage for men Pesos per day 1919

1.70

1924

2.18

1930

3.02

1932

3.58

Source: Marjorie Ruth Clark, Organized Labor in Mexico (Chapel Hill, 1934), p. 192.

The labour wars brought a new kind of labour leader to the textile industry, people like Martin Torres in Orizaba and later, Antonio J. Hernandez in Atlixco. Torres rose through the ranks as union gunman, local labour leader, head of the regional CROM, mayor of Orizaba and finally the region’s strongman, or cacique. He controlled the unions that controlled the jobs in the region’s buoyant textile industry and he could just as easily find employment for friends as assassinate 77 Presidente Municipal Provisional to Secretario General de Gobierno, March 29, 1924, AMA, Presidencia, 1924, Caja H-21, num. 39. 78 Ernesto Cabrera (Srio Gral), et al. to Presidente del H. Ayuntamiento, November 30, 1921, AMA, Presidencia, 1921, Caja 11. 79 Leticia Gamboa Ojeda, ‘La CROM en Puebla y el Movimiento Obrero Textil en los Anos 20’, in Memorias del Encuentro sobre Historia del Movimiento Obrero (Puebla, 1984), p. 49. 351

The Ashgate Companion to the History of Textile Workers his enemies. He controlled the mill town for a number of years. Benito G. Leon exercised similar power in Atlixco, as did Antonio J. Hernandez from the Second World War until his death in 1985. They were powerful and feared leaders and the mill owners had to treat them with the same deference and respect that workers once treated the owners. In 1924, for example, the CROM mayor of Atlixco, Benito Flores, arrested the Metepec administrator, Federico Fantini, and made him sweep the city streets, much to the delight of rank and file unionists.80 Years later in Atlixco, Hernandez, according to one worker, ‘created an action group to kill anybody who didn’t serve him’.81 This did not prevent the Church hierarchy in the town from labelling Hernandez ‘a man of noble ideals and Christian sentiments’.82 When Calles became president, he made CROM leader Morones the Minister of Industry, Labour and Commerce. Morones organized a second convention of the textile industry in 1925, which completed the drafting of a contract in 1927. The 1927 contract was, unlike the 1912 documents, a legal contract backed by federal and state labour law. Its 256 pages included 116 basic articles and many more job-specific items, as well as rules for an industry-wide system of comisiones mixtas for dispute resolution.83 Although it ratified the numerous gains workers and unions had made during the revolution, it further strengthened the control of the labour bureaucracy over rank and file through the clausula de exclusion, with which union leaders could expel dissidents from the union and therefore from their jobs. Furthermore, its explicit and detailed work obligations for each job category necessarily linked the contract to specific technologies and machines, subjecting technological progress to negotiations between unions and owners. Textile owners did not like the pro-labour bias of the contract and sought judicial relief, but the government prevailed, supported by a 1929 presidential ruling that made the contract obligatory for Mexico’s textile industry.84 Factory owners and directors of mills soon discovered that they were right to oppose the contract because they could not fully control the mills. They neither hired nor fired nor had the power to discipline workers. When, for example, the Metepec mill tried to fire a group of employees for stealing in July 1927, it found that it couldn’t remove them but instead sank into an endless morass of litigation impossible to win.85 Meanwhile, problems in the Mexican economy led to a period 80 ‘Los Años Que Fueron Nuestros’, in Los Días Eran Nuestros (Puebla, 1988), p. 171. 81 ‘De cómo unos hurtadores hiceron….’, in Los Días Eran Nuestros (Puebla, 1988), p. 241. 82 Magdaleno Rosales, ‘El Movimiento Obrero en Atlixco’, unpublished manuscript, p. 70. 83 As in 1912 but unlike later contracts, the 1927 contract included cotton, wool and the other branches of the textile industry. Later contracts would separate them. Secretaría de Industria, Comercio y Trabajo, Convención Colectiva de Trabajo y Tarifas Minimas de Aplicación en la República, para las Fábricas de Hilados y Tejidos de Algodón, Estampados, Lana, Bonetería, Yute, y Trabajos Similares (México, 1927), p. 256. 84 Comision Nacional Obrera de la Industria Textil, Los Contratos Ley de la Industria Textil (Mexico, 1979), p. 13. 85 Jeffrey Bortz, ‘The Revolution, the Labor Regime, and Conditions of Work in the Cotton Textile Industry in Mexico, 1910–1927’, Journal of Latin American Studies, 32 (3) (2000): 352

Mexican textile workers of crisis; between 1925 and 1932 textile sales declined from 104,395,152 pesos to 84,161,245, and the number of workers fell from 42,359 to 34,626.86 A similar decline took place in the much smaller wool industry when the number of workers dropped from 5,029 to 3,775.87

The mature labour regime, the good years, 1927–1946 Textile unions won their revolution, even if rank and file workers found themselves subordinated to a government-allied labour bureaucracy. Article 123 and the 1927 contract provided unprecedented job stability and benefits to textile workers. For many, the early post-revolutionary years represented a time of extraordinary prosperity for textile workers. In an industry of mostly skilled workers with pride in their work and a strong sense of job hierarchy, a few job categories stood at the top, a large number of skilled occupations occupied the middle, with only a few, unskilled occupations at the bottom. Unions normally gave preference in hiring to family members, so that many families had more than one person working in the mill. Typical was the case of Elias Avila, born in 1920. He entered ‘El Triunfo’ in 1935, where his father, three brothers and eventually his son obtained employment.88 Many, like José Solís Moreno, could claim that ‘I carried workers’ blood, I belong to the third generation of workers.’ Like his father before him, Solis was a spinner in El Carmen (Atlixco).89 For most Mexicans, family was primary, so that the victory of unions in favouring and protecting family members demonstrated to mill hands that they had indeed won the revolution. In any case, textile work often ran in the family. As Francisco Reyes Lopez wrote in 1932, ‘having been raised in this, I began to work at the age of 12, in 1908’.90 Like many of the children, he slowly worked his way up in the mill, becoming a weaver a decade later. Textile contracts specified four types of positions, permanent, definitive, temporary and eventuales. The latter lacked the benefits of the former, so that workers sought positions in the mill first, then permanency afterwards. By the 1950s, textile contracts stipulated that 5 per cent of the workers must be entry-level apprentices, chosen by the union with preference for children of workers in the same factory and in the same occupation.91 Both cotton and wool continued to favour men over women. In 1935 the cotton industry employed 35,247 men and 4,599 women while the wool industry employed 4,468 men and only 937 women.92 671–703, at p. 689. 86 Moises T. de la Pena, La Industria Textil en Mexico (Mexico, 1934), p. 125. 87 Ibid., p. 289. 88 Estrada Urroz, Del telar a la cadena de montaje, p. 71. 89 ‘Tienen Fama los del Carmen’, in Los Días Eran Nuestros (Puebla, 1988), p. 13. 90 Francisco Reyes López to Martin Torres, octubre 14, 1932, AMO, Caja 917, exp. 62. 91 Estrada Urroz, Del telar a la cadena de montaje, p. 71. 92 Mexico, Resumen General Del Censo Industrial de 1935 (Mexico, n.d.), pp. 64, 104. 353

The Ashgate Companion to the History of Textile Workers The Orizaba region, with the largest and most modern mills, stood at the centre of the industry. Orizaba itself was the most populated of four adjacent textile muncipios – Orizaba (pop. 42,952), Camerino Z. Mendoza (pop. 10,200), Nogales (pop. 7,215) and Rio Blanco (pop. 10,000) – in which the five local mills were located. A 1932 local report claimed that ‘this region is the most important textile centre of our country and perhaps of America because of its great production of cotton spinning and weaving, which, because of its quality, manufacture, and fine finish, make it competitive with foreign production in other textile centres’.93 The Orizaba Camara de Trabajo easily controlled local governments, local politics and hiring and firing in the mills. Similarly, in Atlixco, the unions ruled the town and working-class culture permeated daily life. Morones and his friends in the labour bureaucracy earned great incomes from mandatory union dues, government contracts and concessions, and their political influence. Marjorie Ruth Clark noted that ‘When a government department was placed in control of a CROM man, he made it known to every employee in that department that “voluntary” contributions to the support of the Mexican Labour Party would be welcome and would be of material assistance in keeping the employee in the job.’94 Most owners understood that to settle a strike, cash would have to change hands. As a contemporary wrote, ‘The head of the Mexican labour movement, Luis N. Morones has become a man of wealth. He owns many properties including a textile factory – though not in his own name. He lives lavishly. He sports not less than a half dozen automobiles. His parrandas staged every weekend in the suburb of Tlalpam are notorious for their orgiastic extravagance.’95 Outgoing president Calles took two decisions that threatened to diminish the CROM. He decided to institutionalize the new ruling elite, founding the Partido Nacional Revolucionario (PNR). The move angered Morones, who held ambitions for his own political wing of the CROM, the Partido Laborista Mexicano (PLM). Secondly, he supported the presidential candidacy of Emilio Portes Gil, an opponent of the CROM, in 1928. Losing support with the country’s ruling group, the CROM began a slow decline as many regional labour leaders sought to create their own governmental alliances. Within the CROM, Vicente Lombardo Toledano and Fidel Velázquez led a dissident movement. In 1931, the federal government replaced the state labour codes with the Ley Federal de Trabajo. The new law allowed for industry-wide contract law (contratoley) in which the federal government guaranteed contract enforcement. The 1927 textile contract became contract law but was restricted to the cotton industry. Between 1935 and 1941, tripartite conventions wrote separate contract laws for the other branches of the textile industry, including wool in 1937. All six textile contracts gave unions rather than owners control over hiring, and each imposed the

93 Leonardo Altamirano to Presidente de la Junta Central de Conciliacion y Arbitraje, septiembre 30 de 1932, Archivo Municipal de Orizaba (AMO), Caja 913, exp. 41, p. 11. 94 Clark, Organized Labor in Mexico, p. 105. 95 Ernest Gruening, Mexico and its Heritage (New York, 1928), p. 390. 354

Mexican textile workers clausula de exclusion, strengthening the labour bureaucracy.96 Ongoing productivity problems in cotton forced the government to hold industry-wide conventions and contract modifications in 1934, 1936 and 1937.97 In 1932 Lombardo abandoned the CROM, founding a new organization (CGOCM) in 1933. The growing distance between the government and Morones, and the success of dissidents like Lombardo and the Communist Party, led to increasing dispersion of organized labour in the early 1930s. In 1934 the PNR made Lazaro Cárdenas President of Mexico. Cárdenas sought a political base among campesinos and urban workers, leading him to support efforts to unify the now fragmented labour movement. In 1935 former president Calles and Cárdenas suffered a public split, lending urgency to Cárdenas’ attempts to gain labour’s backing. In 1935 Lombardo organized a meeting of the CGOCM, the Confederación Sindical Unitaria de México (CSUM) and powerful unions in railroads and electrical power, forming the pro-Cárdenas Comité Nacional de Defensa Proletaria (CNDP). He then utilized the CNDP to organize the Confederación de Trabajadores de México (CTM) in February 1936. Lombardo became the CTM’s first Secretary General in 1936 and remained so until 1941. His organization included the unions and federations formally affiliated to the CGOCM (some of which supported Lombardo, others, Velázquez), the PCM-affiliated CSUM and powerful national unions in railroads, electrical power, telephones, mining and other activities. Only the CROM and CGT unions stayed out. The old CROM dissident Fidel Velázquez was the CTM’s first Secretary of Organization, 1936 to 1941, creating state and local federations personally loyal to him. In March 1938 Cárdenas reorganized the PNR into the Partido de la Revolución Mexicana (PRM), a mass party of four sectors, labour, popular, agrarian and the military. Despite the withdrawal of the Miners Union in June 1936 and PCM opposition to labour joining the official party, the CTM claimed almost a million workers in 1938 and was the foundation of the PRM labour sector. The PRM later became the Partido Revolucionario Institucional (PRI) and controlled the Mexican presidency until 2000. The CTM continued to lead the PRI’s labour sector. Throughout the 1930s, textile unions remained powerful in the mills and in mill towns but increasingly dependent on their affiliations to national confederations and their ties to federal government and the ruling party. By 1940, the country’s 425 textile unions affiliated to eight separate labour centrals, reflecting conflict among various labour tendencies.98 The 1940s was a difficult period for urban 96 Comision Nacional Obrera de la Industria Textil, Los Contratos Ley de la Industria Textil, p. 21. 97 Ana Maria Hernandez, La Mujer Mexicana en la Industria Textil (Mexico, 1940), p. 123. 98 The centrals included the Federacion Nacional Obrera Textil del Ramo de Lana; the Federacion Nacional Textil, CROM, led by Eucario Leon and Longinos Hernandez; the Federacion Obrera del Ramo Textil, CGT; the Federacion Nacional de Trabajadores de la Industria Textil, CTM, led by Wolstano Pineda and Sabino Cuellar; the Federacion Nacional Textil, CROM, led by Martin Torres; the Federacion Nacional Textil, CNT; the 355

The Ashgate Companion to the History of Textile Workers labour in Mexico. Wartime inflation brought the sharpest decline in real wages in Mexican history. The Communist Party and the increasingly anti-communist CTM supported a pact with the government to suppress strikes and wage increases. In 1946 the strongly pro-US and pro-business Miguel Alemán (1946–1952) became Mexico’s president, transforming the official party into the Partido Revolucionario Institucional (PRI). The CTM continued its control of the party’s labour sector. CTM leadership, headed by Fidel Velazquez, tightened its control over the affiliated unions while becoming an increasingly docile tool of an increasingly conservative government.

The mature labour regime, the bad years, 1946–1976 After the Second World War, three important developments hammered Mexican textile workers: lagging productivity, the rise of a competitive artificial fibre industry and mostly anti-labour PRI governments. In the 1930s many contemporaries had noted the industry’s increasingly backward technology. Marjorie Ruth Clark wrote that ‘Methods and machinery are far from modern. The workers are inefficient and their output is very small and very costly as compared with that of textile workers in such countries … Fearful of unemployment, the workers have consistently and effectively opposed the installation of modern machinery and methods.’99 She added that Minister of Industry and former labour leader Luis N. Morones believed that the modernization of the industry would lead to unemployment and supported worker and union opposition to increasing productivity. She noted the case of an American company in Guadalajara that installed machines so that one woman could manage three of them. The union intervened, forcing the company to hire one man and one assistant for each machine, negating the benefits of capital investment.100 By the mid-1940s, Mexico’s textile industrialists agreed that their main problem was low productivity. They argued that the industry’s contract laws hindered competition with the United States. For example, in 1947, Mexican weavers managed 2–4 looms, while American weavers attended 20–100. A Mexican atador de mano attended 1,000–1,200 threads an hour, while in the USA a machine tied 10,000–15,000 threads an hour.101 Cotton textiles remained an important but no longer a leading industry in Mexico. Utilization of raw cotton dropped between 1927 and 1932, grew to 69,740 Sindicato Nacional de Trabajadores Textiles y Conexos; and the Sindicato de Empleados y Obreros de la Industria de Boneteria del Distrito Federal, CGT. Hernandez, La Mujer Mexicana en la Industria Textil, pp. 99–100. 99 Clark, Organized Labor in Mexico, p. 183. 100 Ibid., p. 184. 101 ‘Informe sobre los trabajos realizados…’, in Memoria General, Segunda Convecnion Mexicana de Empresarios Textiles (Mexico, 1948), p. 137. 356

Mexican textile workers tons in 1946, then fell to 62,441 tons in 1948, and to 54,563 in 1953. A similar decline took place in wool.102 Employment in cotton dropped from 55,811 workers in 1946 to 32,640 a decade later. In wool there was a decline from 6,083 in 1955 to 5,351 in 1957.103 Table 13.7 Employment in Mexican cotton textile industry, 1943–1950 Workers

Workers

1943

50,815

1951

46,532

1944

52,457

1952

44,632

1945

54,000

1953

44,161

1946

55,811

1954

39,556

1947

52,139

1955

41,225

1948

50,824

1956

32,640

1949

48,291

1957

32,847

1950

48,489

Source: Javier Barajas Manzano, Aspectos de la Industria Textile de Algodón en Mexico (Mexico, 1959), p. 26.

During the 1940s and 1950s, real wages in Mexican industry fell significantly.104 According to Barajas, real manufacturing wages dropped 21 per cent between 1940 and 1948 and real wages in cotton textiles fell 33 per cent. In 1954 real manufacturing wages were 11 per cent below 1939 levels, while in cotton they were almost 40 per cent below.105 Barajas blamed the wage decline on obsolete equipment in the mills, but textile union collaboration with government wage policy also played a role.106 Union leaders in textiles refused to break with the government’s antilabour wage policy, preventing a rank and file challenge to falling wages. As a consequence, textile workers did not recover their real wage levels of the 1940s until the 1970s.107 The third factor that influenced traditional textiles was its replacement by the artificial fibre industry, an outgrowth of Second World War developments in 102 Mexico, STyPS, Informaciones Estadisticas sobre algunas ramas de la industria textil (Mexico, 1955), p. 11. 103 Javier Barajas Manzano, Aspectos de la Industria Textile de Algodón en Mexico (Mexico, 1959), p. 26. 104 On the decline in wages, see Jeffrey Bortz, Los salarios industriales en la Ciudad de México, 1939–1975 (Mexico, 1988). 105 Barajas Manzano, Aspectos de la Industria Textile, p. 31. 106 Ibid., p. 32. 107 Estrada Urroz, Del telar a la cadena de montaje, p. 277. 357

The Ashgate Companion to the History of Textile Workers chemicals. In Mexico the producers of artificial fibres were mostly foreign chemical companies, without ties to Mexico’s traditional textile industries. The government first registered significant production of artificial fibre in 1960, 3,198 tons, which grew to 204,900 tons in 1982, a year in which production of cotton fibre was only 183,375 tons.108 Until the late 1950s, many Mexicans considered cotton textiles to represent what it had always represented, Mexico’s largest and most important factory industry. In the 1960s, however, foreign investment in new industries like artificial fibres, automobiles and chemicals displaced the increasingly obsolete natural fibre industry. Puebla, which had been a colonial textile centre, the site of the first textile factory and a predominantly textile city from the 1840s to the 1950s, saw the arrival of four automobile plants in the 1960s.109 In 1965 the traditional textile industry accounted for 52 per cent of Puebla’s industrial production, but only 19 per cent a decade later.110 The changes in the industry affected the labour force, where less and less of the workers came from textile families, and fewer textile families sent their children to work in the mills. Family traditions of skilled textile work slowly ended. Textile work was highly skilled in the nineteenth century and represented the possibility of a better life for many Mexicans. Despite the rigors of factory life, cotton textile workers, through their revolution in the revolution in the 1910s, created an industry after the revolution that provided housing, education, medical care and permanent employment for mill hands, and a decent life for their families. In the mill towns, textile unions and their leaders controlled local politics. As productivity and wages lagged in the second half of the twentieth century, many mill hands found better opportunities in other sectors. Textile work lost its glow.

Globalization, 1976– In 1976 the Mexican economy collapsed as soaring inflation, devaluation and economic contraction pounded workers and their families. Through the 1980s and 1990s, short-term recoveries were followed by ever more severe economic contractions as well as sharp devaluations in 1982 and 1994. Mexican governments in the 1980s, 1990s and into the new century increasingly welcomed globalization and its discontents. A quarter century of acute economic problems led the country to abandon the PRI in 2000, after more than 70 years in power. Mexico initiated tighter integration with the US economy through the maquiladora programme in 1965. The programme permitted foreign companies to import machines and materials without tariffs, assemble in Mexico and export the product without restriction. Beginning as a (US–Mexico) border industrialization 108 INEGI, Estadisticas Historicas de Mexico, T. II (Mexico, 1999), pp. 519–520. 109 Estrada Urroz, Del telar a la cadena de montaje, p. 34. 110 Ibid., p. 40. 358

Mexican textile workers programme, it later spread to the interior. By 1995, there were reports of 3,000 maquiladoras with almost 750,000 workers. That year, maquiladoras provided 20 per cent of all Mexican manufacturing jobs.111 Many textile firms took advantage of the maquiladora provisions to establish plants in Mexico. Maquiladoras have been generally non-union, with lower wages and less benefits than most Mexican manufactories. Studies estimate that about 60 per cent of the maquiladora labour force is female, earning minimum wages. There have been a number of labour conflicts in maquiladoras as women workers fought to unionize their firms, but the foreign companies found strong allies in Mexico’s government and the traditional labour confederations, and mostly defeated the challenges. In the 1980s, very low paid seamstresses formed the combative Seamstress Union, the Sindicato de Costureras 19 de Septiembre. The women fought for unionization and higher wages in the maquiladora sector but the government, the traditional and male-dominated union leaderships, and the companies eventually defeated them, often when the firms declared bankruptcy, then re-opened without a union or a contract.112 On 1 January 1994, a free trade agreement between Canada, Mexico and the United States – the North American Free Trade Agreement (NAFTA) – became operative. The agreement reduced the tariffs on Mexican textile exports to the USA from 9.1 per cent in 1993 to 1.9 per cent in 1999, which not only increased Mexican textile exports to the United States, but led many US manufacturers to relocate their production facilities to Mexico to take advantage of significantly lower labour costs. Furthermore, the implementation of NAFTA had, by 2001, repealed the restrictions on domestic maquiladora sales. At the beginning of the twenty-first century, the largest natural fibre firms in Mexico were not the old CIDOSA or CIVSA mills of the late nineteenth and early twentieth century, but rather foreign and often US firms, like Burlington Industries, long active in textile production in the United States. With modern (and imported) machines, they generally hired low wage, non-union women. The CTM and other Mexican unions allied to the PRI worked with the technocratic governments of late twentieth-century Mexico to undermine textile wages and the law-contracts that had once served to protect textile workers in domestic industry. The defeat of a half-hearted textile strike in 1992 signalled the death of the old labour regime when a series of rulings undermined the prevailing contract laws.113 Many workers remained combative but could not overcome the resistance of the labour bureaucracy.

111 Migration News, 3 (2) (1996). 112 Ana Alicia Solis de Alba, El Movimiento Sindical Pintado de Magenta (Mexico, 2002), pp. 224–236. 113 Solis de Alba, Movimiento Sindical, p. 191. 359

The Ashgate Companion to the History of Textile Workers Table 13.8 Mexican textile production, 1985–2000 Cotton (tons)

Wool (tons)

Cotton (tons)

Wool (tons)

1985

210,780

3,128

1995

192,720

1,375

1990

177,100

2,341

1996

253,000

1,350

1991

180,840

1,471

1997

189,100

1,300

1992

30,140

1,456

1998

218,900

900

1993

24,200

1,504

1999

132,000

920

1994

100,540

1,440

2000

73,040

920

1995

192,720

1,375

Source: INEGI, Anuario Estadistico de los Estados Unidos Mexicanos, 2002 (Mexico, 2003), p. 388.

Despite labour’s defeat, Mexico did not receive nearly as much foreign investment in textiles as it might have wished. Instead, many firms moved to China and other Asian countries. Although Mexican wages in 2000 were about 16 per cent of US wages, Chinese wages were about 12 per cent of Mexican wage levels. With natural fibres an old industry with standardized technologies, companies competed on labour costs. Mexico became less attractive than other alternatives. Globalization thus did not favour Mexico. Twenty-first-century textiles were the most global of industries, leaving textile workers at the mercy of foreign firms that could leave if unions or higher labour costs appeared. Without advanced technologies, a strong financial system or the lowest wage costs, it was not clear that Mexico had a role to play in the international restructuring of the industry. Textile workers, once at the centre of Mexican politics and economic development, were relegated to jobs without a future. From an industry of mostly skilled, unionized, highly paid male workers, cotton textiles had become a symbol of early twenty-first-century globalization: unskilled, non-unionized, low-paid women workers. In the sixteenth century, making cotton and wool textiles represented a skill. Many colonial Mexican families earned a living or supplemented their income producing textiles. In the nineteenth century, industrialization brought new opportunities but also problems inherent in large-scale, hierarchical and often exploitative factory production. At the beginning of the twentieth century, cotton textile workers carried out a workers’ revolution within the broader Mexican revolution, ultimately transforming the modern Mexican labour regime – in law, Constitution and practice. Article 123 of Mexico’s revolutionary Constitution stands as a testament to that revolution. Powerful unions soon controlled work in the mills and life in the mill towns. The decline in the industry after the Second World War made textile work less attractive, and by the 1960s and 1970s many workers in 360

Mexican textile workers the old mill towns felt that ‘it was shameful to be seen as a worker’.114 With the globalization of the industry at the beginning of the twenty-first century, foreign industrialists mostly hired relatively unskilled young women in order to avoid unionization, decent wages and the good times of Mexico’s earlier textile boom. Textile workers had once again become ‘the most unfortunate class’.

114 ‘Habia una vez un hombre…’, in Los Días Eran Nuestros (Puebla, 1988), p. 181. 361

This page has been left blank intentionally

14 The Netherlands Elise van Nederveen Meerkerk, Lex Heerma van Voss and Els Hiemstra-Kuperus

Until 1568, the northern Netherlands had belonged to the Habsburg Empire. A combination of economic crisis, a power struggle between the Dutch gentry and the Spanish king over his policy and taxes, and religious differences led to the revolt of several provinces in the north against the Spanish Crown. This revolt resulted in a war of independence that would last for 80 years, until the Peace of Westphalia was signed in 1648. In 1588, seven provinces proclaimed the Dutch Republic of United Provinces (hereafter ‘the Republic’). The Republic was governed by the Estates General, in which all provinces were represented. Despite the ongoing warfare, relative peace existed in the economically most important provinces of the Republic on the west coast, Holland and Zealand. In the southern Netherlands, the war against the Spanish troops continued, resulting in a decline of trade and industry. Thousands of Protestant migrants fled from the southern Netherlands to the safer cities in the north. This migration provided the northern Netherlands with a new labour force. Furthermore, many of the refugees were rich merchants from cities like Antwerp and Ghent, who brought capital and knowledge, giving an economic stimulus to the Dutch cities in which they settled. The unstable international situation, with England and France at constant war, continued for almost a century. Under these circumstances, the Dutch Republic became the centre of the world economy from the 1580s until at least 1650. The Dutch leading role in shipping and trade favoured the development of a large staple market in the city of Amsterdam, where goods from all over the world were traded.  See I. Schöffer et al. (eds), De Lage Landen van 1500 tot 1780 (Amsterdam, 1991), p. 140.  A standard work on southern Netherlands migration to the Republic is J.C.G.A. Briels, De Zuid-Nederlandse immigratie, 1572–1630 (Haarlem, 1978). See Oscar Gelderblom, Zuid-Nederlandse kooplieden en de opkomst van de Amsterdamse stapelmarkt (1578–1630) (Hilversum, 2000), pp. 15–28 for a good overview of the discussions on the role of Flemish merchants in the Republic’s economy.  Jonathan Israel, Dutch Primacy in World Trade, 1585–1740 (Oxford, 1989); Clé Lesger,

Map 14.1 Netherlands

The Netherlands Because of the Republic’s primacy in world trade, historians have mainly focused on this aspect of the early modern Dutch economy. Lately it is recognized, however, that industry has also been crucial for the economic development of this region. The export-oriented textile industry was especially large in certain parts of the Republic in the seventeenth and eighteenth centuries. Much of our knowledge of the textile industry in this period is based on the vast work on the wool industry of the largest textile-producing city, Leiden, written by N.W. Posthumus in the first half of the twentieth century. Other research is also mainly regionally oriented. Here, we will try to extend this knowledge, by looking at several textile-producing regions within the northern Netherlands at the same time, and by analysing major shifts and interdependencies between them.

The Dutch textile industry: general developments 1650–1810 The Dutch economy was actually on its way down after 1650, although several sectors continued to flourish for quite some time. At the beginning of the seventeenth century woollen cloth weaving was carried out in most parts of the Republic, although not always as large-scale, export-oriented production. However, the undisputed centre of woollen textiles was Leiden, a city in the province of Holland. Most other cities in Holland could not compete with Leiden, and in the first half of the century their production of woollens declined or vanished. In cities like Amsterdam and Rotterdam, textile entrepreneurs specialized in the finishing and trading of cloth instead. The woollen industry in Leiden reached its peak in 1664. From then onwards, the city gradually lost its position as the main clothproducing centre of Europe. International competition increased, and within the Republic itself new production areas rose as well. This happened especially in the more peripheral agrarian regions, where labour costs were relatively low. For the spinning and weaving of wool, a village in the south of the Republic, Tilburg, became increasingly important. Especially in the course of the eighteenth century, more and more villages in the countryside of North-Brabant developed a cloth industry. Nevertheless, Leiden remained the most significant centre of woollens Handel in Amsterdam ten tijde van de Opstand. Kooplieden, commerciële expansie en verandering in de ruimtelijke economie van de Nederlanden ca. 1550–ca. 1630 (Hilversum, 2001).  Jan de Vries and A. van der Woude, Nederland 1500–1815. De eerste ronde van moderne economische groei (Amsterdam, 1995), p. 322.  N.W. Posthumus, De geschiedenis van de Leidsche lakenindustrie, vols II and III (’s-Gravenhage, 1939).  Leo Noordegraaf, ‘The New Draperies in the Northern Netherlands 1500–1800’, in N.B. Harte (ed.), The New Draperies in the Low Countries and England, 1300–1800 (Oxford, 1997), pp. 173–195, at p. 183.  National Archives The Hague, Archives Department of Internal Affaires, 1796–1813, inv. no. 783. 365

The Ashgate Companion to the History of Textile Workers until the end of the pre-industrial period. The decreasing number of weavers’ guilds illustrates the decline of woollen weaving in the cities of Holland (see Table 14.1). Of course, these figures only offer us a rough qualitative indication, but they nevertheless mark the general shift of woollen production out of the cities. Table 14.1 Wool weavers’ guilds in the Dutch Republic, 1600–1800 Total Province of Holland

Percentage of total

1600

25

7

28.0

1650

23

6

26.1

1700

18

3

16.7

1750

12

2

16.7

1795

12

1

8.3

Note that the total number of guilds increased in the seventeenth and eighteenth century. The decline of the number of weavers’ guilds can thus not be explained by an overall decrease in corporative organization. Source: Database Guilds IISH.

After wool, linen was the second most important textile product in the seventeenth and eighteenth centuries. Haarlem had been the leading linenproducing city, but it declined rapidly after 1650. From the 1650s onwards, the production of linen in the Brabant and Twente countryside (in the south and east of the Republic) increased dramatically. The Dutch production of cotton fabrics was nevertheless still underdeveloped in this period. Cotton yarn was used in some sectors as a raw material for mixed fabrics such as fustians or bombazines, but it was not produced in the Netherlands on a large scale until the end of the eighteenth century. Only after the mechanization of the Dutch textile industry in the course of the nineteenth century did the production of cotton yarns and fabrics become

 See, for Haarlem, F. Mulder, ‘De Haarlemse textielnijverheid in de periode 1575–1800’, in H. Rombouts (ed.), Haarlem ging op wollen zolen: opkomst, bloei en ondergang van de textielnijverheid aan het Spaarne (Schoorl, 1995), pp. 55–109, and for Twente (the main production area of linen after 1650), Cor Trompetter, Agriculture, Proto-Industry and Mennonite Entrepreneurship: A History of the Textile Industries in Twente, 1600–1815 (Amsterdam, 1997). 366

The Netherlands really important. Therefore, we will mainly focus on the woollen industry for the early modern period.

Size of production and workforce It is hard to measure the size of cloth production and the workforce, because there are no national statistics for production and labour in this period. No coherent series of total output can be reconstructed with the scarce information that is available. Production figures for the early modern period were probably best documented in Leiden. Posthumus has produced a series for woollen production in this city (see Figure 14.1).

                  Figure 14.1 Production of pieces of woollen cloth in Leiden, 1650–1810. Source: Posthumus, De geschiedenis, pp. 129, 1099.

 See Jaap Boot, ‘Handspinnen van katoen en handkatoenspinnerijen (deel 1)’, Textielhistorische bijdragen 24 (1984): 52–80, at p. 58 and Jaap Boot, ‘Handspinnen van katoen en handkatoenspinnerijen (deel 2)’, Textielhistorische bijdragen 26 (1986): 38–80, at pp. 43f. 367

The Ashgate Companion to the History of Textile Workers Tilburg, the second most important cloth-producing area in the Republic after 1650, showed a different development. Textile production was very much restricted outside the cities. Furthermore, industrial products from the province of North-Brabant, where Tilburg was situated, were heavily taxed. Nevertheless, entrepreneurs from cities like Rotterdam and Amsterdam tried to stimulate woollen production in the area of Tilburg. In most cities of Holland, the finishing and trading of cloth had become very important and the influential merchants were treated with leniency by the central government when they sent their wool to be spun and woven in Tilburg, where labour costs were much lower than in Holland, without being taxed. In 1687, in spite of protests by textile-producing cities in Holland such as Leiden and Haarlem, the merchants even managed to reach an official agreement, the so-called ‘Concession of Tilburg’. From then on, they were definitely free to put out their production in the countryside, albeit in this village alone.10 Unfortunately, there are not many production figures for Tilburg, but it appears that cloth production rose by more than 125 per cent from 1650 to 1810.11 Regarding the numbers of spinners and weavers we have some – though not much – more information. The data are heterogeneous and hard to compare, but the textile industry was certainly an important employer. In its heyday, it must have provided thousands of people – men, women and children – with work. Posthumus estimates that around 1650, more than 36,000 inhabitants of Leiden depended on the cloth industry.12 Unfortunately, solid numbers are not known for the whole of the Republic. Although Leiden was the main production centre, there was also weaving activity in many other cities and in the countryside – there were still 23 wool weavers’ guilds in the Republic in 1650. Therefore, at least several thousands more people must have been working in the industry at this time. Surely, this number declined drastically during the last part of the seventeenth and the eighteenth centuries. Table 14.2 shows, for instance, that in cities like Gouda and Utrecht, where a reasonable cloth production existed before 1650, there were very few weavers left in 1671. In Leiden, the number of weavers dropped as well, from over 2,500 in 1650, to not even 1,500 around 1750 and less than 900 in 1810. Nevertheless, still more than 13,000 people in the Netherlands were spinning or weaving in the woollen industry in 1810.13

10 Gerard van Gurp, Brabantse stoffen op de wereldmarkt. Proto-industrialisering in de Meierij van ’s-Hertogenbosch 1620–1820 (Tilburg, 2004), pp. 47–53. 11 Cf. Gerard van Gurp, ‘De Tilburgse lakenhandel met Holland en Brabant in de zeventiende en achttiende eeuw’, Textielhistorische bijdragen 38 (1998): 50–73, at pp. 54–55 and Michael Jansen, De industriële ontwikkeling in Nederland 1800–1850 (Amsterdam, 1999), p. 312. 12 The people working in the linen, knitting and other textile industries are subtracted from Posthumus’ figures. The number of 36,000 people amounted to approximately 60 per cent of Leiden’s total population, Posthumus, De geschiedenis, pp. 882, 937. 13 Jansen, De industriële ontwikkeling, p. 133. 368

The Netherlands Table 14.2 Weavers and spinners in the Dutch Republic, 1650–1810 Wool Wool weavers spinners Leiden

c.1650

2,535

Leiden

1661

3,365

Tilburga

1665

139

Leiden

1667

3,400

Gouda

1671

14

Utrecht

1671

35

Leidena

1749

1,446

1,587

Tilburg

1810

592

2,466

Leiden

1810

880

3,289

Twente

1810

North-Brabant

1810

Netherlands total a

c.1810

Cotton weavers

Cotton spinners

205

1,032

2,781

9,658

170

179

539

1,670

Only heads of households

Sources: J.E.J. Geselschap, Gouda zeven eeuwen stad. Hoofdstukken uit de geschiedenis van Gouda (Gouda, 1972), p. 146; Jansen, Industriële ontwikkeling, pp. 133, 306, 316, 318; J. Klinkenberg, ‘Meesters, kooplieden, fabrikeurs. Vroegkapitalisme in de Utrechtse textielnijverheid in de late middeleeuwen en de vroegmoderne tijd’ (unpublished doctoral thesis, Amsterdam University, 1991), p. 13; Nederlands Historisch Data Archief, Database kohier Leiden 1749; Posthumus, De geschiedenis, pp. 936–938; Regionaal Historisch Centrum Tilburg (RHCT) Dorpsbestuur, inv. no. 380–I; RHCT, Volkstellingen, inv. nos 1275–1277.

369

The Ashgate Companion to the History of Textile Workers Cotton, however, was on the rise. Large-scale mechanization of the industry would only take place after 1810, but since the late eighteenth century some entrepreneurs in the peripheral regions of Twente and North-Brabant had introduced new machinery like the spinning jenny. This innovation accelerated after 1810, and resulted in a rapid increase of cotton spinners and weavers. With this development, important conditions for the future were created. Twente and North-Brabant would remain the core of the Dutch textile industry until the second half of the twentieth century.

Organization of production and institutions In most Dutch cities, industries and trades were regulated through guilds, which had been established since the thirteenth century. Guilds were corporative organizations, operating with local government support to protect the economic and social interests of the members of their specific trade. They were mostly founded in cities, although there were some guilds in the countryside.14 Consequently, woollen weaving in the cities was also mostly organized within a guild. In general, citizenship was required to be admitted to a guild. To prove his skill and obtain his mastery in order to participate in the weavers’ guild, the weaver had to produce a ‘masterpiece’. It took several years of training with another guild master before a weaver could take this test. Once a master, the guild offered many advantages, like the right to weave within the city, with more or less fixed prices, guarding against uneven competition, and material support in case of sickness or death for him and his family. These advantages of course had their price: a yearly contribution was required, and if a weaver violated the guild’s regulations, he was (sometimes heavily) fined.15 Moreover, the guild system excluded many social groups. Non-citizens and unskilled workers could not join the club. Most guilds were also reluctant to admit women. In some guilds, widows could take on their deceased husband’s profession as a master, but married and unmarried women were usually excluded. This monopolistic policy is supposed to have driven women out of the weaving craft in the late Middle Ages. Spinning was, according to most sources, a predominantly female occupation. It was associated with irregular, low-skilled work and a virtually non-existent professional identity. Exactly this gendered division of labour would explain why spinning was hardly ever organized in a guild or otherwise.16 14 See Piet Lourens and J. Lucassen, ‘De oprichting en ontwikkeling van ambachtsgilden in Nederland (13de–19de eeuw)’, in Catharina Lis and Hugo Soly (eds), Werelden van verschil. Ambachtsgilden in de Lage Landen (Brussels, 1997), pp. 43–77. 15 Ibid., pp. 53–54; Bibi Panhuysen, Maatwerk: kleermakers, naaisters, oudkleerkopers en de gilden (1500–1800) (Amsterdam, 2000), p. 23. 16 Merry Wiesner, ‘Spinsters and Seamstresses: Women in Cloth and Clothing Production’, in M.W. Ferguson, M. Quilligan and N.J. Vickers (eds), Rewriting the Renaissance: The Discourses of Sexual Difference on Early Modern Europe (Chicago, 1986), pp. 191–205, at 370

The Netherlands Despite the importance of the guild system, the two major cloth producers of the seventeenth and eighteenth centuries did not have a guild for most of the period. In Leiden, the drapers’ guild had been abolished in 1561. However, cloth production was also regulated in a corporative way: in neringen, established by the city council around 1585. The various woollen fabrics, like serge, cloth and baize, were organized within their own nering. Contrary to guilds, non-citizens could also take on woollen production. Another difference was that the workers of various stages of the production process were subjected to the rules of their specific nering. One or two superintendents and two to five governors made sure that all the workers observed the regulations and they also settled differences between entrepreneurs and their employees. Each nering had its own central building, the ‘hall’, where the workers brought their pieces of cloth to be examined, measured and sometimes even sold.17 Until 1630, most Leiden drapers were independent entrepreneurs. The draper led the production process, and he employed several weavers, spinners and other workers. More and more, however, poor drapers came to lean on cloth finishers and merchants for their subsistence. From 1630 to 1670, most drapers lost much of their independence. They started working as ‘middlemen’ between a new group of capital-owning entrepreneurs, the reders, and the actual workers. The spinners and weavers worked at home, or they were concentrated in small workshops. In the eighteenth century a new type of entrepreneur, the fabriqueur, appeared on the scene. This fabriqueur employed many workers who produced for him at home or in larger manufactories. It is likely that he was not a real merchant, but rather a producer of cloth. The processes of producing and selling were increasingly separated during the eighteenth century.18 It is unsure when the Tilburg woollen weavers’ guild was established, but it existed in 1767.19 It probably had little influence.20 Cloth production was organized in various ways. Firstly, there were independent drapers, who bought their own materials and sold their cloth themselves. Secondly, there were merchants from out of Tilburg, who brought raw materials to the village to be spun and woven there on a commission basis. In the second half of the seventeenth century, however, many independent weavers could no longer compete with the production of wealthier entrepreneurs. More and more, they started to weave for these ‘capitalist’ drapers and merchants in return for wages. The ‘putting-out’ system that thus came into being made most Tilburg weavers dependent on wage labour.21 pp. 191–193. 17 Posthumus, De geschiedenis, pp. 338–339, 438f. 18 Ibid., pp. 502–504; Boudien de Vries, ‘De Leidse textielnijverheid in de zeventiende en achttiende eeuw’, in J.K.S. Moes and B.M.A. de Vries (eds), Stof uit het Leidse verleden. Zeven eeuwen textielnijverheid (Utrecht, 1991), pp. 77–89, at pp. 86–87. 19 H.F.J.M. van den Eerenbeemt, De opkomst van Tilburg als industriestad. Anderhalve eeuw economische en sociale ontwikkeling (Nijmegen, 1959), p. 163. 20 Van Gurp, Brabantse stoffen, p. 87. 21 Gerard van Gurp, ‘Proto-industrialisatie in Tilburg en Geldrop’, Textielhistorische bijdragen 39 (1999): 13–50, at pp. 17–18. 371

The Ashgate Companion to the History of Textile Workers Apart from its interference with the guilds and the neringen, local government hardly meddled with textile production. Most local government policy was aimed at improving the impoverished cloth industries.22 Even conflicts were usually not taken to court, but resolved within the corporate bodies. Most conflicts were between individual workers and drapers or merchants, and the quarrel was almost always about wages. Organized protests by a great number of workers against ‘capitalist’ employers were remarkably rare or at least undocumented in this period.23 Central government policy regarding textiles usually concerned trade and textile production in the countryside. Already in the sixteenth century many cities in Holland had complained to the Estates General about the competition they experienced, and in the seventeenth century complaints finally resulted in regulations in Den Bosch and some cities of Twente.24 In the long run, as we have seen above, nothing could be done to stop textile production moving to the more peripheral and more agrarian parts of the Dutch Republic.

Division of labour We have already noted the interregional division of labour in the Dutch woollen industry. This division concerns the merchant cities of Holland, where the dyeing and finishing of cloth were carried out, and the area of Tilburg, where much of the spinning and weaving took place. However, this was an evolutionary process rather than a revolutionary one. Leiden, although in decline since its heyday in the mid-1600s, remained a very important weaving area within the Republic. In 1749, still approximately 1,450 heads of households were registered as weavers, a number that Tilburg would never reach during the pre-industrial period. Still, in broad lines, major geographical shifts had started in the second half of the seventeenth century. Tilburg and other rural areas of Brabant became an increasingly important cloth-producing region during the eighteenth century.25 The decline of woollen weaving in the cities is illustrated by the rapid decrease in the number of guilds between 1650 and 1800 (see Map 14.2).

22 We have found examples in the municipal archives of Den Bosch, Leiden, Utrecht and Zwolle. 23 Regional Archives Leiden, Archive Halls, inv. nos 216–221. 24 Herman Kaptein, De Hollandse textielnijverheid 1350–1600: conjunctuur en continuïteit (Hilversum, 1998) pp. 83–84; Trompetter, Agriculture, p. 48. 25 National Archives The Hague, Archives Department of Internal Affaires, 1796–1813, inv. no. 783. 372

Map 14.2 Woollen weavers’ guilds in 1650 (left) and 1780 (right)

The Ashgate Companion to the History of Textile Workers Another division of labour was that between the sexes. The general view is that in the Middle Ages men had taken over weaving from women, who were left with the lower skilled and poorly rewarded spinning. Women could easily combine spinning with running a household and raising children, and it was therefore supposed to be one of the few jobs suitable for them. There are no reasons to doubt that it was men who did the weaving in the early modern Dutch Republic. In most areas in the seventeenth and eighteenth centuries, the few women we found weaving were widows, and doing mostly linen weaving, not woollen. On the other hand, it appears that not only women spun in the Republic. In all areas of research, an important – though mostly not predominant – percentage of spinners were men (see Table 14.3). Table 14.3 Division of labour in spinning according to sex Male spinners

%

Female spinners

%

1665

Tilburg

40

34

76

66

1712

Zwolle

11

13

71

87

1742

Zwolle

14

8

171

92

1749

Leiden

772

49

813

51

1775

Den Bosch

59

39

91

61

1810

Tilburg

297

18

1,364

82

1810

Tilburg (incl. children)

626

25

1,841

75

Note: only heads of households, except Tilburg 1810. Sources: various tax and population registrations (see Table 14.2).

How can these figures be explained, since it is usually supposed that only women spun because they represent much cheaper labour? Firstly, there might be a methodological issue to explain the high percentages of male spinners. In most official population and tax records, only heads of households were registered, which automatically leads to the under-registration of (spinning) married women and children. Although this argument might reduce the relative importance of male spinners, it does not explain why, in many cases, men were spinning in the first place. This leads to a second explanation, connected with the fact that most men who were spinning were wool spinners. In linen and cotton spinning, we almost only find women. Obviously, the nature of the industry mattered. Since woollens were much more expensive than linens and mixed fabrics, the status (and perhaps the payment) of a spinner in this branch was probably higher than elsewhere. 374

The Netherlands Thirdly, another factor could be the degree of organization of production. In Leiden’s archival sources, we found many subcontractors called ‘master spinners’, who were paid by a draper for subcontracting other people (men and women) to spin. The master spinners were never female. Another example of shifts in the sexual division of labour is wool-combing, a ‘female’ occupation in the sixteenth century,26 which was taken over by men in the seventeenth. Around 1700, they even established a wool-combers’ guild, which was only accessible to male wool-combers. Only occasionally, a master’s widow was allowed to join the guild. Similarly, the only known wool spinners’ guild was established in the city of Utrecht around 1700.27 Only men were taken on as apprentices. Once admitted to the guild, they were called ‘master spinners’. Possibly the diminishing of the woollen industry, which led to a smaller demand for weavers, caused men to take on ‘inferior’ and ‘female’ textile jobs. By adopting the title of ‘master’ or erecting a guild, they were attempting to institutionalize the job and give it a higher standing. This new status attached to spinning probably made it a more acceptable profession for men.

The workers: ethnic background, age, labour and living circumstances, and work culture The Dutch Republic is known as a typical immigrant country, with an outspoken religious tolerance. Migrants from France and the southern Netherlands moved to the cities of Leiden and Haarlem to find work in the woollen and linen industries. Amsterdam, in particular, attracted a lot of Portuguese, Spanish and East European Jews, who had often fled their countries because of persecution. Also, labour migration from Germany was very common.28 Occupations such as combing, spinning and weaving were of course easily accessible because of their low capital intensity and skill requirements. In Leiden, this was perhaps even more the case, since one did not have to be a citizen to work in the textile industry. But were spinning and weaving typical immigrant jobs because they were low status? Earlier research has pointed out that immigrants did not necessarily have inferior or low-paid jobs compared to native-born citizens of Holland. It is probable, however, that they were the first to suffer from the depression in the second half of the seventeenth century, as appears from the decline of the percentage of immigrants in the silk industry in Amsterdam.29 Among wool-combers (not the most rewarding occupation) in

26 Els Kloek, Wie hij zij, man of wijf. Vrouwengeschiedenis en de vroegmoderne tijd: Drie Leidse studies (Hilversum, 1990), pp. 63–65. 27 Utrecht Archives, Guild archives, inv. no. 114. 28 Jan Lucassen and Rinus Penninx, Nieuwkomers, nakomelingen, Nederlanders. Immigranten in Nederland 1550–1993 (Amsterdam, 1999), pp. 34–35, 52. 29 Jan Luiten van Zanden, Arbeid tijdens het handelskapitalisme. Opkomst en neergang van moderne economische groei (Bergen, 1991), pp. 60, 73. 375

The Ashgate Companion to the History of Textile Workers Leiden, many of the names were French around 1700.30 Possibly immigrants were the first to take on lower paid work like combing and spinning, but this does not mean that spinning and weaving were typical immigrant jobs. The age of Dutch spinners and weavers ranged from very young to very old. This was the case throughout the entire period. At the beginning of the seventeenth century, orphans in Leiden were contracted out to learn a skill at a master’s house. Most masters worked in the textile industry, teaching the orphans spinning or weaving. Boys, who usually started to learn spinning and spooling when they were little, often received training as weavers or shearers from the age of 14. A contract was drawn up in which it was agreed that the boys would learn the trade of weaving and would stay with their master for several years. This very much resembled the formal training weavers’ apprentices received in other cities, where woollen weavers’ guilds were present. The education of orphan girls, on the other hand, mostly did not develop likewise. Some girls were found spinning with the same boss from their early teens until they were over 20 years old.31 Obviously, women were usually not trained for any occupation except spinning. In the poor relief records of Zwolle for around 1700, we see women and men of all ages spinning, although adult men were usually underrepresented.32 In 1810, the total population registration of Tilburg shows the same pattern (see Figure 14.2). Living and working circumstances were harsh. Weaving and especially spinning are usually associated with low rates of pay and poverty. They worked long hours in the same physical position in dark, badly ventilated rooms, for which most of them only earned a few stuivers per day. This view is confirmed by Posthumus for Leiden, and Prak for Den Bosch, who have found out that the majority of weavers and spinners belonged to the lowest ranks of the income classes of these two cities.33 In the city of Zwolle, the poor relief records show that of all 3,524 records over the period 1650–1700, 539 men and women (more than 15 per cent of all records) were listed as ‘spinner’. Their earnings were strikingly low, from ‘almost nothing’ to a maximum of 30 stuivers for a father and a mother spinning together, with which they could hardly pay the rent and feed all members of their family. 34 We see many spinners living in basements, renting small rooms from richer citizens or living in special houses for the poor, offered to them by the city council. On the other hand, there must have been a lot of spinners who did manage to get by without municipal poor relief in the form of money or goods. When we compare data on poor relief in Den Bosch with population registers, we find that over half of the people who were

30 Regional Archives Leiden, Guild Archives, inv. nos 1374/3 and 1374/4. 31 Regional Archives Leiden, Archive Halls, inv. nos 3844, 3845, 3847. 32 Based on Historical Centre Overijssel (HCO), City Council, inv. nos 309–312. 33 Posthumus, De geschiedenis, pp. 187, 572–573, 971–973; Maarten Prak, ‘Arme und reiche Handwerker in ’s-Hertogenbosch am Ende des 18. Jahrhunderts’, in K.H. Kaufhold and W. Reininghaus (eds), Stadt und Handwerk in Mittelalter und Früher Neuzeit (Keulen/ Weimar/Wenen, 2000), pp. 241–267, at pp. 249–251. 34 HCO, IA025, inv. nos 309–312. 376

The Netherlands registered as spinners did not receive any help from the city.35 Of course, this leaves out the fact that they could receive informal support from family, neighbours or friends, which is unfortunately difficult to track in archival sources.

    PDOHVSLQQHUV



IHPDOHVSLQQHUV

       RYHU      DJHJURXS

  

PDOHZHDYHUV



IHPDOHZHDYHUV

       RYHU      DJHJURXS Figure 14.2 Spinners and weavers in Tilburg 1810, according to age and sex Finally, it is hard to tell much about work culture and professional identity. Spinners and weavers often worked alone, or with members of their family, and it is difficult to find sources that directly report on issues of work culture and identity. In most cities, weavers were organized in a guild, which formed an institutional 35 City Archive Den Bosch, Old Archive, inv. nos 1433–1443 compared with inv. no. 1452. 377

The Ashgate Companion to the History of Textile Workers means to stand up for their rights and their welfare within the city. Furthermore, through the guilds they formed networks of solidarity (for example burial and sickness funds) and social life. The status and power of weavers was low however, for as we have seen on a national level, the more important cloth shearers, dyers and textile merchants usually got the better deal when it came to protecting their economic interests. With spinners, who were hardly ever organized, it is even more difficult to define what particular work culture existed and what status and identity was attached to the profession. Sometimes, female spinners gathered together to spin in the same place. In these ‘spinning bees’, women spun, sang songs and told each other tales, and young girls were even allowed to meet their possible future husbands.36 As noted above, there are not many indications of massively organized labour protests or upheavals in the pre-industrial period. We will now turn to the industrial period, to discover whether, and under what conditions, this all changed.

1810–1950 The second period is marked by the introduction and growth of industrialized production in the textile industry. During the first half of the nineteenth century the centre of Dutch textile production moved further from the old urban setting to the new centres on the sandy soils of the east and the south. Between 1812 and 1849 the number of textile workers increased from 35,000 to 45,000. This growth was concentrated in the newer centres. The Twente cotton industry was late to mechanize. The production of cotton goods was limited by the poor quality of the cotton yarn produced. The production consisted primarily of mixed linen and cotton fustians (bombazijn) for inland consumption.37 To re-conquer the Indonesian markets from the English after the end of the Napoleonic Wars, the Dutch government stimulated the export of Dutch cotton goods to the colony in the period 1824–1873, through an agreement with the semi-public Nederlandsche Handel Maatschappij (NHM, Netherlands Trade Company).38 At the beginning of the nineteenth century hand-driven spinning machines with 100 spindles each had been introduced in Twente, which by that date were already old-fashioned elsewhere. Given the low wages they paid, the Twente

36 Gerard Rooijakkers, Rituele repertoires. Volkscultuur in oostelijk Noord-Brabant 1559–1853 (Nijmegen, 1994), pp. 307–308. 37 Jansen, De industriële ontwikkeling, pp. 113–150; Jan Luiten van Zanden and Arthur van Riel, Nederland 1780–1914. Staat, instituties en economische ontwikkeling (Amsterdam, 2000), p. 168. 38 Thomas Lindblad, ‘De handel in katoentjes op Nederlands-Indië, 1824–1939’, Textielhistorische Bijdragen 33 (1993): 89–103. 378

The Netherlands manufacturers could still make profits when they employed outdated technology.39 This led to a change from spinning at home to production in sheds, which were forerunners of actual textile mills. Weaving still took place in the home, often in combination with other occupations. As the heavy hand shuttle was used, this was considered a man’s job. In 1833 the NHM founded a weaving school were English workers taught young weavers the use of the flying shuttle. This enabled the Twente entrepreneurs to produce calicoes for the Dutch East Indian market.40 The growth in production led to an increase in wages and employment in weaving in Twente in the 1830s and 1840s.41 As better quality yarn was still imported from England there was no comparable rise in employment in cotton spinning.42 The Brabant wool industry around Tilburg specialized in rather simple fabrics like baize and flannel.43 From about 1870 buckskins and other finer fabrics were added. The first mechanical spinning mill was established in 1809. The number of mechanized spinning mills increased quickly in the next decade. Pieter van Dooren in Tilburg introduced the first steam engine in any Dutch textile mill in 1827. The number of workers working in their home declined quickly: from 75 per cent in 1817 to 50 per cent in 1825. For the time being weaving was done by hand, and in the home. Until about 1820 two weavers operated one loom. When the flying shuttle was introduced around 1820, this was reduced to one weaver. It was only in 1860 that the first mechanical loom was introduced in the Tilburg wool industry. At first the weaving factories were small. Only after 1888 did the number of steamdriven looms significantly increase. By the end of the century mechanical weaving had become predominant, but as late as 1905 20 per cent of the weavers were still working at home. By then the phenomenon was disappearing quickly: in 1925 just 1 per cent of the weavers was working at home.44 Spinning, which had been a female job in the cottage industry, changed into a male occupation when it was mechanized and centralized in spinning mills.45 Adult men operated the spinning machines, assisted by boys. Women and girls reeled the yarn, and boys wound it on wooden spools. Men continued to dominate weaving, but some women operated looms at home. Burling and darning was done by women and girls, fulling, raising and dry shearing by men. 39 Erik Fischer, Fabriqueurs en fabrikanten. Twente, Borne en de katoennijverheid 1800–1930 (Utrecht, 1983), pp. 61–90; J.C.M. Jansen and C. Trompetter, ‘Hoezo achterlijk? Een studie naar de situatie in de vroeg negentiende-eeuwse Twentse textielnijverheid (1800–1835)’, Textielhistorische Bijdragen 35 (1995): 101–119. 40 Joel Mokyr, Industrialization in the Low Countries, 1795–1850 (New Haven and London, 1976), pp. 99–103. 41 Ibid., pp. 111–113. 42 Van Zanden and Van Riel, Nederland 1780–1914, pp. 171–174. 43 Mokyr, Industrialization, pp. 114–121, P.J.M. van Gorp, Tilburg eens de wolstad van Nederland. Bloei en ondergang van de Tilburgse wollenstoffenindustrie (Tilburg, 1987), pp. 97–122. 44 Gertjan de Groot, Fabricage van verschillen. Mannenwerk, vrouwenwerk in de Nederlandse industrie (1850–1940) (Amsterdam, 2001), pp. 186–195. 45 Ibid., pp. 188–194. 379

The Ashgate Companion to the History of Textile Workers Table 14.4 Gender composition of the Dutch textile labour force, 1920 and 1930 1920 Male (%)

1930

Female (%)

Male (%)

Female (%)

Cotton Roving

152 (18)

834 (82)

Ring spinning

621 (52)

569 (48)

Self-actor spinning

982 (71)

398 (29)

2,687 (61)

1,719 (39)

3,964 (55)

3282 (45)

296 (10)

2,731 (90)

371 (10)

3,226 (90)

Total spinning mills Reeling

263 (98)

5 (2)

396 (99)

6 (1)

Weaving

10,340 (77)

3,015 (23)

13,824 (86)

2,261 (14)

Total weaving mills

12,266 (65)

6,711 (35)

16,065 (71)

6,559 (29)

1,755 (84)

339 (16)

1542 (81)

366 (19)

293 (44)

369 (56)

397 (54)

340 (46)

2,839 (96)

119 (4)

3,165 (96)

118 (4)

5 (1)

800 (99)

4 (1)

895 (99)

4,103 (73)

1,489 (27)

4,475 (75)

1,528 (25)

Starching

Wool Total spinning mills Reeling Weaving Burling and darning Total weaving mills

Source: Centraal Bureau voor de Statistiek (CBS), Beroepstellingen 1920, 1930.

By 1908, when almost all production had become concentrated in factories, work had become segregated not only along gender lines, but also according to marital status. Men and boys worked in the factories, as did girls and unmarried women. Women were mainly employed in burling and darning: unmarried women and girls in the factories, married women at home.46 To a lesser extent they also worked at reeling. On average, women earned about half as much as men.47 Work by children has been researched for Leiden. Here in the 1850s it was normal to start work in the textile factory at age 10, and 7 per cent of the workforce were even younger than that. Almost one-third of the workforce were below the age of 16. Around 1860 a public outcry against young children working in factories somewhat reduced the use of very young children, but 10- and 11-year-olds soon returned, only to be banned from the factories definitively by the Child Labour Act 1874. This limited factory work to children aged 12 and over.48 46 Cor van der Heijden, ‘Personeel in de Tilburgse wolnijverheid, 1920–1940’, Textielhistorische Bijdragen 26 (1986): 92–111. 47 De Groot, Fabricage, pp. 201–206. 48 Cor Smit, ‘Fabriekskinderen. Kinderarbeid in de Leidse textielindustrie in de 380

The Netherlands In the 1850s steam machines began to be introduced in both cotton spinning and weaving, albeit somewhat hesitatingly. The new steam-driven spinning mills were at first unable to compete with English yarn, but conquered the interior market for coarser yarns from hand-spinning. Hand-spinning had lost all importance by 1870. In 1852 G. and H. Salomonson founded the first steam-driven weaving mill in Almelo. An experienced handloom weaver could learn to operate a power-loom in 14 days. Hand-weaving lost out to steam-looms between 1860 and 1885. It persisted longer in the weaving of coloured fabrics.49 The growth in productivity in spinning has to be ascribed to the progressive replacement of mule-jennies by self-actors (which had become dominant by 1870) and continuous spinning machines. The number of workers per 1,000 spindles was more or less constant, at 12–14, in 1870–1900, but decreased to 11 in 1910. The percentage of youths employed decreased in the same period from 30 to 14 per cent. The growth in productivity in weaving, as measured in cotton yarn consumed per loom, was less impressive than that in spinning. However, we must take into account that by the end of the period the Twente cotton mills had switched to lighter fabrics, which means that the increase in productivity was larger than suggested by the table. Table 14.5

Cotton spinning and weaving in the Netherlands, 1861–1910 1861

1871

1881

1891

1910

55,800

229,600

212,200

264,700

487,000

13

25

42

39

44

3,500

10,900

13,500

19,400

33,000

Cotton consumed per loom (in kg)

n.a.

1,000

1,100

1,300

1,400

Cotton yarn consumed (x 1,000 kg)

7,900

16,100

17,600

24,200

46,000

Of which by steam-driven looms

n.a.

11,000

14,900

24,200

46,000

Of which by handlooms

n.a.

5,100

2,700

–

–

10,900

10,100

10,100

14,000

22,700

Spindles in steam mills Cotton consumed per spindle (in kg) Looms in steam mills

Workers in Twente textile mills

Sources: De Jonge, De industrialisatie in Nederland, pp. 104, 106; Erik Fischer, ‘De ontwikkeling van de Twentse katoenindustrie en de toename van de arbeidsproduktiviteit tussen 1800 en 1930’, Textielhistorische Bijdragen 22 (1981): 3–40.

negentiende eeuw’, Textielhistoriche Bijdragen 36 (1996): 61–96. 49 J.A. de Jonge, De industrialisatie in Nederland tussen 1850 en 1914 (Nijmegen, 1976), pp. 97–104. 381

The Ashgate Companion to the History of Textile Workers In cotton spinning the Twente factories followed British examples in gender segregation: women worked in the roving departments and men in the self-actor departments. Only when the crisis of the 1930s gave an extra impetus to cutting wage costs, were women employed as helpers in the self-actor departments. Women earned 50–70 per cent of male wages. The self-actors themselves continued to be operated by men. By then self-actor spinning had become an outdated technology. Ring frames, which were introduced from the 1880s, but on a larger scale from the 1900s, were operated by women. In wool spinning, ring frames were introduced in Tilburg from 1930. As the cotton spinning firms fired women when they married or became pregnant, the number of women declined sharply from their twenties. Around 1880 a large number of boys were employed. When young men were no longer required as helpers in the spinning department, this figure declined.50 In cotton weaving girls reeled the yarn, and men bleached and dyed. Boys and girls were responsible for winding. Women and men, later only women, wound the warp, and men starched it. In checked cotton, weaving was the domain of adult men. In the 1900s and 1910s Twente employers threatened several times to introduce female weavers, usually during wage conflicts. This would have robbed the men of a privileged male preserve. In mills that produced plain, unpatterned cotton, both men and women wove. In 1908 women in the Enschede cotton mills earned only 80 per cent of what the men earned. Piece wages – which were employed in the textile industry wherever possible – were uniform, but men worked more often on four looms, with an assistant, and women only on two.51 Increasing the number of looms operated by weavers was one of the main ways in which the employers tried to raise productivity in the weaving mills. Men often brought their children to assist them. Weaver Frans Weltevreden at the Van Heek factory in Enschede already operated six looms in this way, assisted by two of his offspring: ‘I am standing between my two children, a boy of 16 and a girl age 13’, he told a parliamentary committee in 1890. ‘The boy, who works energetically and quickly, operates two looms; I supervise all six looms and so we help each other.’52 Around the turn of the century it was quite usual for a weaver to operate four looms single-handedly. In the 1920s several Enschede cotton mills tried to introduce weaving on six looms without a helper. In 1929 the Jannink and Sons firm had weavers weaving on eight or even nine looms without assistance. By then weavers normally handled four looms at other Twente mills. This huge increase in productivity was made possible by the introduction of automatic or Northrop looms, which changed bobbins automatically. Stimulated by the crisis, in 1937 the number of looms handled by a single weaver had increased to 12 or 16, and even 24 in the case of fully automatic looms. The weavers, who had to jump from one loom to the other to repair broken threads, complained that their job had become ‘too nervous’.

50 De Groot, Fabricage, pp. 337–399. 51 Ibid., pp. 161–186. 52 Quoted ibid., p. 181. 382

The Netherlands In a similar way the number of spindles operated by one spinner had increased from about 20 in a spinning machine operated by hand in the early nineteenth century to 60 in the early steam-driven spinning machines around 1860, to 200 at the end of the 1910s. Production per spinner increased tenfold in the 90 years between 1830 and 1920, and production per weaver increased sixfold. Before the turn of the twentieth century this was mainly due to the introduction of more advanced machines. In the twentieth century piecemeal technological improvement continued, but the increase in the number of spindles and looms a worker operated became much more important.53 Table 14.6 Spinning and weaving in the Netherlands, 1921–1950 1921

1929

1939

1946/48

671,100

1,190,000

1,305,000

1,350,900

Looms

47,100

55,300

48,700

47,800

Workers (per 15 September)

31,800

42,100

38,600

44,450

252,100

257,800

264,000

287,100

4,600

5,000

5,800

6,000

10,300

11,800

15,400

13,900

Cotton Spindles

Wool Spindles Looms Workers (per 15 September)

Source: CBS, Statistiek van Voortbrenging en verbruik.

Workers and working conditions Textile entrepreneurs introduced the textile industry to the sandy soils of Brabant and Twente because there was a peasant labour surplus. In both regions the industry started as a putting-out system, with members of peasant families spinning and weaving when work on the family farm permitted it.54 In the second half of the nineteenth century communities like Enschede in Twente and Tilburg in Brabant changed from groups of agricultural hamlets to proverbial industrial towns. This growth entailed immigration, mostly from the nearby countryside. There was no 53 Fischer, ‘De ontwikkeling’. 54 Francois Hendrickx, ‘In Order not to Fall into Poverty’. Production and Reproduction in the Transition from Proto-Industry to Factory Industry in Borne and Wierden (The Netherlands), 1800–1900 (Amsterdam, 1997). 383

The Ashgate Companion to the History of Textile Workers large-scale long-distance migration, and immigration seems not to have unsettled the existing social fabric. Social theory predicts certain changes in life style and demographic behaviour as a result of proto-industrialization and industrialization: higher marriage frequencies, lower age at first marriage and higher fertility, an individualistic outlook and a breakdown of extended kin relations. So far, the outcome of in-depth research has run against these predictions, both in Twente and in Brabant.55 Many textile workers’ families remained tied to the soil. The reasons were explained by the 48-year-old weaver Frederikus Gervink from Borne: ‘I put a lot of effort in my little farm, so I will always have something to eat, even if I am fired from the factory.’ Gerrit te Flierhaar also saw spreading his risks as a good strategy. He and four of his six children worked in textile firms, but all in different ones. He thought it ‘unwise, if all work in the same factory, because if that factory stops, all are out of work at the same time’.56 As late as 1920 in a survey of Twente textile workers more than 80 per cent worked a plot of land and more than half kept animals for meat, eggs or milk.57 In the survey, more than 70 per cent of the workers declared they would prefer to do other work, like 34-year-old Hendrikus Brilman, a sterker (starch worker) from Almelo who desired other work because of ‘the free air and to flee the seclusion in the large prisons’. Half of the workers that specified an alternative mentioned work in agriculture. Indeed, succeeding one’s father in keeping a small farm was the main way to climb the social ladder for Twente textile workers. The high number of workers longing for alternative employment is quite understandable if we take into account labour relations.58 Most Twente textile workers entered the factory gate as soon as this was legally allowed. Ten years later, in their early twenties, as spinner or weaver, they had already reached the highest wage level attainable. Only a handful of workers could attain a better paid position as foremen. To start a business of their own was impossible. In Twente, the spinning capacity grew between 1860 and 1930 from 42,500 to over a million spindles, but the number of factories that housed these decreased from 10 to 9. The number of weaving factories did increase in this same period from 10 to 45. However, as the number of looms increased much faster, the average weaving mill grew from 200 to 1,000 power-looms. In the cottage industry a weaver had often 55 Angélique Janssens, Family and Social Change: The Household as a Process in an Industrializing Community (Cambridge, 1993). 56 Quoted in Francois Hendrickx, ‘Een wereld van verschil? Landbouwende wevers en wevende landbouwers in Twente en Brabant in de negentiende eeuw’, Textielhistorische Bijdragen 41 (2001): 7–26. 57 Lex Heerma van Voss, ‘Twente 1920: gelijktijdige generaties?’, Tijdschrift voor Sociale Geschiedenis 15 (1989): 1–23. 58 Lex Heerma van Voss, ‘Van CAO’s tot zonden; 330 Twentse textielarbeiders in 1920 over hun drijfveren ondervraagd’, Textielhistorische Bijdragen 27 (1987): 31–52. The description of working conditions follows to some extent Lex Heerma van Voss, ‘Dertien uur per dag aan het weefgetouw. Mechanisering, arbeid en sociaal protest in de Twentse textielindustrie’, Spiegel Historiael 38 (2003): 528–533. 384

The Netherlands owned his loom, but buying a power-loom was not within the means of weavers, let alone the dozens of looms necessary to start a competitive factory. There were niches for smaller firms in Twente, which occupied themselves with processing waste matter, finishing, bleaching or other steps in the production process. But none of these were usually founded by workers. As the textile firms grew in size, their owners became increasingly rich, powerful and important. In local political and social life these directors and their spouses set the tone. The social distance between ordinary workers and their all-powerful employers was immense. In the 1920 survey workers were asked if they ever spoke to their employers. Foremen occasionally had a conversation with a director, but ordinary workers hardly ever. Textile worker W. Konijnenbelt declared that he had spoken to his director ‘almost never’, and he specified that this had happened four times in the 22 years that he had worked for his employer. The inspector of finished cloth H. Lubbers belonged to the Protestant-Christian trade union, which believed that employer and worker should cooperate and that the worker should know his place. But he bitterly wrote that his employer only addressed him to correct him. Workers at home could decide how much time they spent spinning or weaving, but working days in the factory were long. In 1840 12- or 13-hour working days were normal. As work continued on Saturday, this meant working weeks of some 75 hours. Employers had invested enormous sums in their machinery. To earn back their investment, they wanted the machines to run as many hours as possible. For that reason they changed Catholic holidays into working days in Almelo in 1870, a move which the weavers protested against. In later decades working hours decreased slowly, until the workweek amounted to about 60 hours by 1910. The introduction of the 1911 Factory Act (Arbeidswet) meant that women and children could not work more than 10 hours daily and 58 hours weekly. In practice this meant that the workweek was also reduced for men working together with women or children. Eight years later the 1919 factory act brought an important reduction to 8½ hours per day and 48 hours per week. As the nineteenth century progressed, the argument was voiced that from a moral point of view it was not good if women and men or boys and girls worked in the same space. This would provoke immoral speech, if not immoral conduct. In the Twente textile firms this fear was not shared everywhere. The Enschede Chamber of Commerce wrote in 1893: ‘The morality of acts was not at risk in the midst of the crowd of people, and that of words is safeguarded by the noise on the shop floor.’ The noise of the machines was deafening. However, not only noise but also dust polluted the atmosphere in the factories. Every Saturday the looms had to be cleaned, before the workers could leave. To lessen the chance that threads would break, the atmosphere was also artificially made damp and hot. Small wonder that workers preferred work in the open air, for as one worker put it in 1920, in the factory ‘one is no human being but a machine’. The wages that workers received in compensation were considered low. The reason why the firms had been relocated to Twente and Brabant was precisely that workers there accepted lower wages. Wage costs were not very important as a share of total costs. In Enschede wages fell from 19.1 per cent of the value of the 385

The Ashgate Companion to the History of Textile Workers product to 12.3 per cent in 1920.59 In the 1920s this figure rose again. However, as they could not influence world market prices for raw materials or finished goods, and as competition was often fierce, employers tried to limit their wage costs as much as possible.

Protest and organization Trade unions were relatively late in establishing themselves in Twente and Brabant. But workers’ protest did not wait for permanent organizations. In 1827 Tilburg workers demonstrated and broke the windows of manufacturer Pieter van Dooren to protest the introduction of steam power.60 Figure 14.3 illustrates the number of strikes that took place in the Dutch textile industry.61 Until the closing decade of the nineteenth century they were isolated incidents.                              

Figure 14.3 Strikes and lockouts in the Dutch textile industry, 1800–2000 Socialism spread to Twente from 1880. In 1888 a strike broke out at Scholten’s weaving mill in Almelo. It was a spontaneous outburst of anger, as had happened before. The socialist activist Gerrit Bennink organized the strike, and the outcome was a partial success for the workers. This led to the establishment a year later 59 A. Blonk, Fabrieken en menschen (Dissertation Universiteit van Amsterdam, 1929), p. 225. See also Fischer, Fabriqueurs, pp. 237–238. 60 Van den Eerenbeemt, De opkomst van Tilburg. 61 Source: Database Stakingen in Nederland (Database Strikes in the Netherlands); http://www2.iisg.nl/databases/stakingen/index.php, consulted 15 September 2004. All conflicts in occupational group XV, textiles, were included, a total of 821. Thanks to Sjaak van der Velden and Gordan Cupac for help in handling the data. 386

The Netherlands of the Wevers- en Spinnersbond Vooruit (Weavers’ and Spinners Trade Union Forwards). A number of Catholic workers joined, and this stimulated chaplain Alfons Ariëns to found a Roman Catholic Union in 1891: the Rooms-Katholieke Fabrieksarbeidersbond (Roman Catholic Factory Workers’ Union). In 1896 an interconfessional trade union of Protestant and Roman Catholics, Unitas, followed. After a number of split-offs and regroupings, and a decisive intervention by the Dutch episcopate, which forced Catholic workers out of inter-confessional Unitas and into a separate Roman Catholic organization, a pattern resulted which would become normative for the Dutch trade union movement as a whole.62 Four organizations of textile workers resulted: 1. The Socialist Algemene Bond van Textielarbeiders De Eendracht (General Trade Union of Textile Workers ‘Unity’, 1904) 2. Roman Catholic St Lambertus (1907) 3. Unitas, which as a result of the departure of the Catholic workers became a de facto Protestant organization 4. The Syndicalist Landelijke Federatie van Textielarbeiders (National Federation of Textile Workers, 1904). Of these unions, Socialist Eendracht and Catholic St Lambertus organized the most workers. St Lambertus was especially strong in homogenous Catholic Brabant, Eendracht in religiously mixed Twente. As was generally the case, socialist trade unions found it very hard to organize workers in the Catholic South.63 Protestant Unitas was much smaller, and the Syndicalist LFT was by far the smallest of the four. Usually the LFT would put forward the most far-reaching demands, followed by Eendracht. St Lambertus and especially Unitas would formulate demands that were much closer to what the employers were offering. This of course made it often easy for the employers’ organizations to play one trade union against the other. Another result of the strike at Scholten’s was that in 1888 the Enschede employers founded the Fabrikanten vereniging ter voorkoming van werkstakingen (Manufacturers’ Association to prevent strikes). In the following years similar associations of employers were founded elsewhere in Twente. The Employers’ Association, better known as the Fabrikanten Vereniging Enschede (FVE), fought tough in the class struggle.64 Harsher than the methods of Dutch employers in other branches of industry was the tactic of locking out workers in the course 62 For an overview, see W.H.M.E. Alberts et al., Marges van de vakbeweging. Een analyse aan de hand van de Twenste katoenstakingen in 1923–1924 en 1931–1932 (Den Haag, 1982), pp. 15–19 and on Unitas especially Arno Bornebroek, De strijd voor Harmonie. De geschiedenis van de Industrie- en Voedingsbond CNV 1896–1996 (Amsterdam, 1996), pp. 93–103. 63 A.A.J. Thelen, ‘Het Nederlandsch R.K. Textielarbeiderssecretariaat en Unitas; van een amorfe veelheid naar een verdeelde eenheid’, Textielhistorische Bijdragen 30 (1990): 52–71. 64 F. van Waarden, Het geheim van Twente. Fabrikantenverenigingen in de oudste grootindustrie van Nederland ± 1800–1940 (Amersfoort/Leuven, 1987). 387

The Ashgate Companion to the History of Textile Workers of a strike. If a strike took place in one firm, in due time non-striking workers were locked out: first from the firm itself, then from other local textile firms and eventually all Twente textile workers were locked out. Trade unions would support their members financially during a strike. Due to the lockout, they were forced to support not only the strikers, but all their members. They also were handicapped in raising money among workers in support of the strike. In Twente, where the textile industry was the main employer, a general lockout meant hardship everywhere. The Employers Association used this weapon with success in 1890, 1902, 1906, 1907, 1910 and 1912. It is remarkable that this tough position of very dominant employers’ did not hamper trade union growth. Trade union density was 41 per cent in 1920 and 46 per cent in 1930 for the textile industry as a whole. In Twente the figures were even higher: 45 per cent in 1920 and 60 per cent in 1929.65 As Figure 14.3 makes clear, the trade unions were also not afraid to let it come to a strike or lockout. Strikes were numerous in 1903–1908, 1912–1917, 1919–1923, 1928–1931 and 1935–1936. The strike pattern in the textile industry was not markedly different from the pattern in the Dutch economy as a whole.66 In many of these years the unions were actually on the defensive. Good strategy would seem to suggest not striking when the conditions are unfavourable, but the militant stance of the employers often left the unions little choice. This can be illustrated by the lockout of 1923/24.67 In August 1923 the employers proposed lengthening the working week from 48 to 53 hours and lowering piecerates by 10 per cent. Sales were poor. The trade unions knew this, and realized their position was weak. But they also – correctly – interpreted the demand as an attack on the recently acquired eight-hour day, and decided to fight it out. They struck at one weaving mill, Van Heek in Enschede, on 29 October 1923. On 16 November the employers locked out 11,000 workers from 19 Enschede firms. On 24 December 20 firms in other towns, with roughly the same number of workers, followed. That the employers were locking out the workers, was good for the latter’s public relations. Even the Catholic Archbishop sent a gift in support of St Lambertus. The workers and their organizations were able to keep up the unequal fight for a long time, but had to give up in the late spring of 1924. It seemed as if they had lost the strike and suffered in vain. However, they had kept up their resistance until the business cycle had turned for the better early in 1924. Now the employers were in a hurry to resume their operations. They therefore made an important concession: the longer working hours would only be temporary. The workers had actually beaten off the attack on the eight-hour day. The price had been high, but the trade unions had done what 65 Overall trade union density rates were 31 per cent in 1920 and 28 per cent in 1930 for the total labour force and 45 per cent in 1920 and 38 per cent in 1930 for manufacturing, mining, utilities, construction and transport. 66 Sjaak van der Velden, Stakingen in Nederland. Arbeidersstrijd 1830–1995 (Amsterdam, 2000), p. 237. 67 Alberts et al., Marges; Lex Heerma van Voss, De doodsklok voor den goeden ouden tijd. De achturendag in de jaren twintig (Amsterdam, 1994), pp. 189–193. 388

The Netherlands their members expected of them. In 1931–1932 a quite comparable large conflict took place in Twente, when the employers tried to force higher productivity and lower wages on the workers.68 In August–October 1935 in Tilburg the employers also demanded lower wages and workers reacted with a strike.69 The two Twente textile conflicts are among the largest and hardest industrial conflicts ever fought in the Netherlands. They both reflected, and fuelled the tense labour relations in the industry. The enormous social distance between workers and employers, and the militant position of the employers in labour conflicts, the unhealthy and unpleasant working conditions, low wages, the permanent pressure to produce faster – all combined to make the textile workers militant, even if their organizations suffered regular defeats. It makes it understandable that workers who had spent their whole working life in a textile factory and were proud of their skills, nevertheless declared that they would rather do something else. As the Netherlands industrialized after 1945, and incomes grew, the opportunities to do something else increased.

1950–2000 The Dutch textile industry faced two serious problems in the post-war period: shortage of labour and a decline in demand. Fifteen years after the end of the Second World War, the Dutch textile industry fell into a crisis, which meant the beginning of the end. The significance of the textile industry was decimated: in 1950 the textile industry made up 20 per cent of industrial added value, in 2002 this had dropped to 2.3 per cent. Employment in the textile branch showed the same development: in 1950 textiles comprised 23 per cent of industrial employment, in 2002 only 2.6 per cent. The biggest drop in employment was in the 1970s. In the beginning of the 1980s it stabilized, followed by further decline.70 The first years after the war were turbulent for the textile industry in the Netherlands. The war had caused much damage, especially in Enschede where the spinning mills lost 20 per cent of their capacity.71 In the first years there was no lack in demand yet, and the industry was busy catching up, but there was a lack of everything else: materials, machines and machine parts, energy, foreign currency and labour.

68 B. Hesselink, Gerhard Jannink & Zonen te Enschede 1853–1938. Jaren van rationalisatie en verzet (Hengelo, 1983). 69 Van Gorp, Tilburg, pp. 182–184; T. Wagemakers, ‘De Tilburgse textielstaking van 1935; een beeldverhaal’, Textielhistorische Bijdragen 27 (1987): 94–105. 70 Feike Drost, ‘De neergang van de textielindustrie’, CBS Webmagazine (2004). 71 Francisca de Haan, Een eigen patroon. Geschiedenis van een joodse familie en haar bedrijven, ca. 1800–1664 (Amsterdam, 2002), p. 258. 389

The Ashgate Companion to the History of Textile Workers Shortage of labour Shortage of labour became a serious problem in the 1950s and 1960s. Especially young workers and trained technicians were in demand. The bad reputation of the textile industry was one of the main reasons. Young people no longer aspired to be textile workers; they preferred the metal industry or jobs in construction. The textile industry was notorious for its low salaries, paternalistic management and noisy, dusty and often hurried and tedious work. Because of the far-reaching division of labour and automation the work was reduced to routine operations and became less interesting and rewarding. This all had a negative effect on the social status of textile work. Still textile workers were no more or less satisfied with their jobs than other industrial workers. They took pride in their work and professionalism, but at the same time believed that other industrial work such as in the metal industry required more training and that metal workers therefore were more skilful workers. The vision the ‘outside world’ had of textile labour was mostly based on the real disadvantages of work in this industry.72 Although even after the Second World War workers were recruited with the help of family members – without introduction by a family member you could not even get into a company like J.J. Krantz in Leiden73 – family ties generally started to weaken within the companies. Textile workers wanted something better for their children. With increasing prosperity in the Netherlands textile families could afford to let their children go to school longer and pursue a career as school or nursery teachers.74 With the better economic position of workers it also was no longer necessary for women to work to maintain the household income. From the mid1960s more and more women left the factories. They often did continue working in home employment for a little extra income. With the young reluctant to work in the textile industry, the average age of the textile workers rose.75 Specialized textile education attracted fewer pupils and textile schools finally had to close their doors.76 In both the wool (Tilburg) and the cotton industry (Twente) this shortage of labour was a problem during the 1950s and 1960s. In Leiden the situation was somewhat different: here employment dropped by 35 per cent between 1950 and 1960. Textile companies in Tilburg and Leiden were located in the old city centres. When new city plans were developed they got in the way. In Tilburg new factories were therefore built in the industrial zones outside the city centre, whereas in Leiden the entrepreneurs chose to move out of the city altogether. In 1977 J.J. Krantz was 72 Th. J. IJzerman, Beroepsaanzien en arbeidsvoldoening met betrekking tot de arbeidsvoorziening in de Twents-Achterhoekse Textielindustrie (Leiden, 1959), pp. 110–111. 73 D.J. Noordam, J.K.S. Moes and J.Laurier, Door de wol geverfd. Herinneringen aan de Leidse textielindustrie in de twintigste eeuw (Zutphen, 1998), p. 44. 74 IJzerman, Beroepsaanzien, p. 110. 75 See McKinsey & Company, Mogelijkheden tot het verbeteren van de perspectieven van de Nederlandse textielindustrie (Amsterdam, 1981), table on pp. 43–44. 76 Van Gorp, Tilburg, p. 207. 390

The Netherlands the last textile company to leave. It moved to Vaals where it took over the Vaalse Textielfabriek, but just a year later it went bankrupt. The bankruptcy of the Leidse Wolspinnerij in 1986 marked the end of the textile industry in Leiden.77

The workers In this period women made up about a quarter of textile workers. There were both sex-specific jobs and a number of jobs done by men as well as women. The traditional division of labour was maintained in the weaving mills. Women worked in the preparatory and finishing jobs, but very few women wove. Roving machines, on the other hand, were operated almost exclusively by women, while men and women both worked as ring spinners.78 Because of the shortage of labour workers were able to make demands concerning salary and other conditions of employment. Female workers expressly came to the fore in the period 1945–1965.79 They sometimes took the initiative for strikes as in Leiden in April 1955, where 21 girls started a strike, demanding better pay – 16 men followed their example when asked to take over their work. Unfortunately the unions were opposed to this strike and all but two strikers were fired. In January 1955 in Nijverdal (Twente) there was a strike of mainly girls who had problems with an inspector. This strike turned out to be successful.80 There were several strikes in the 1950s, mostly for higher wages or better working conditions in all main textile regions. Collective resistance was usually concentrated in the Twente cotton areas, and relatively scarce in Tilburg. But in the 1970s, the period in which most companies closed down, there were several actions within individual companies. These strikes were mostly ‘wild cat’. The workers started the actions and (sometimes) the unions took over. Other forms of resistance were sabotage and theft. Mocking, scolding and grumbling are not exactly active forms of resistance but they did show the dissatisfaction of the workers with the way things were going in the textile companies.81 In general the degree of organization was very low in the wool industry in Tilburg, that is, among the blue-collar workers, because in comparison more foremen and managers were organized. The role of the trade unions was more 77 A. van Schelven, ‘Schering en inslag in historisch perspectief. Schets van een geschiedenis van de Nederlandse textiel- en confectie-industrie, 1850–1980’, in Textiel-, kleding-, leder-, schoen- e.a. lederwarenindustrie. Een geschiedenis en bronnenoverzicht (Amsterdam, 1993), pp. 13–69, at pp. 39–41. 78 For division of labour between men and women, see Erna Lohuis, Vrouwen in de Twentse Textiel 1945–1960 (Doctoraalscriptie Katholieke Universiteit Nijmegen, 1986), p. 19; Instituut voor ontwikkelingsvraagstukken (IVO), De ondergang van de Tilburgse Textielindustrie (Tilburg, 1981), vol. 6, pp. 147–151. 79 Lohuis, Vrouwen in de Twentse Textiel, p. 9. 80 Database Stakingen in Nederland. 81 Lou Keune, De textiel voorbij. Het wel en wee van Tilburgse oud-textielarbeiders in de jaren 1980–1990 (Tilburg, 1991), p. 24. 391

The Ashgate Companion to the History of Textile Workers important than was to be expected based on the number of members. The unions played a decisive part in the realization of conditions of employment, because they were the interlocutor and contracting party in negotiations on CAOs (Collective Labour Agreements).82 Labour being scarce employers had to do their best to commit the workers to the company. Labour conditions improved with the development of professional personnel management, workers participation, company social workers and the forming of staff associations that organized trips and other social events. Works canteens appeared as well as company stores where company products could be bought. The stores were very popular during the post-war shortage. There was also more care for personal hygiene. Wash places and locker rooms were extended and modernized. Despite the government policy of guided wages, wages in the textile industry rose to keep the workers from moving to other sectors. The textile industry rose from 22nd place in 1939 to third place in 1953 in the order of wages. This was as far as official wages were concerned. In reality most workers got something extra off the record.83 In 1964 the so-called ‘wage explosion’ took place and wages in the Dutch textile industry ranked among the highest in Europe. In addition, so-called guest workers were recruited in Italy, Spain and, some years later, Morocco to solve the shortage of labour. A group of textile entrepreneurs thought it necessary to improve the image of the textile industry in order to attract more workers. Already in 1947 they had started work on a foundation for this purpose, and in November 1950 the Stichting Textielvak was founded. Employers and employees were represented and some independent experts were also members of the foundation. In the beginning the foundation’s main task was recruitment. In the long run it tried to raise the standard of textile labour through education and training and worked to improve the image of the textile industry. It published a magazine that was distributed in the textile companies. Most textile workers and their families probably read it, but the middle management mostly produced the articles. In their final annual report of 1966 the foundation pleaded for good social policy during the wave of mergers, takeovers and staff reductions. The downfall of the industry was coming about fast. The Stichting Textielvak asked for attention for the long-term problems facing the industry, but most companies lost heart and were no longer able to respond; it ended its activities in 1968.84

82 Ibid., pp. 24–25. 83 J.L. de Jager, Draad van de toekomst. Opkomst, tegenslag en voorspoed van Koninklijke Nijverdal-Ten Cate 1945–1990 (Zutphen, 1991), pp. 57–58. 84 Giel van Hooff en Siebe Rossel, ‘Human en public relations in de Twents-Gelderse textielindustrie. De Stichting Textielvak tussen beeld en werkelijkheid, 1950–1968’, Textielhistorische Bijdragen 42 (2002): 59–88, at p. 84. 392

The Netherlands Market problems Shortage of labour was not the only problem the textile industry faced. Major international developments influenced the market for textiles. In 1949 Indonesia became an independent republic and its market was soon lost. The final orders were delivered in 1964–1965. In 1950 the Benelux customs union came into being, followed by the formation of the EEC in 1958. The consequences of these developments were felt in the first half of the 1960s.85 The production volume of the textile industry declined from the early 1960s. Not only does this distinguish the textile sector from other industrial branches in the Netherlands, but in comparison with other EEC countries, the reduction in production was relatively fast. There were two main causes for this: a reduction in demand and the international competitive position (high cost of labour, 10 per cent higher energy costs and expensive services).86 There were different reasons for reduction in demand. The cotton industry struggled with the loss of the colonial market. The wool industry faced a change in fashion to more casual attire like jeans. The demand for woollen blankets dropped because central heating became more common and the use of duvets came into fashion. Tilburg also lost the army as a customer.87 And then from the 1960s there was the emergence of large-scale production of synthetic fibres, a cheaper material for textile and clothing, forcing the lowering of prices of cotton and wool.88 International competition came mainly from Japan (cotton) and Prato in Italy (wool). Protectionist measures taken by the EC to control the influx of textiles from the low-wage countries (the Multifibre Agreement I 1974) only resulted in a growth of imports from other EC countries. West Germany, for instance, became one of the main competitors on the Dutch market, despite the comparable high labour costs.89 Already before the Second World War the cotton industry had begun market reorientation to end its dependence on exports to the Dutch Indies (Indonesia). The industry realized that more diversification was needed to supplement the one-sided staple commodities that had once been the basis of the cotton industry in Twente, but without giving up the traditional markets in Indonesia, South East Asia and Africa. Marshall Plan aid encouraged measures to strengthen the competitive position such as raising productivity, standardization and modernization of production, and the introduction of shift-work.90 The shift system was very unpopular with the workers, although in general they recognized its economic necessity. The 85 Dik Nas, Het Twentse Model. Honderdvijftig jaar vakbeweging in Enschede (Amsterdam, 1998), p. 140. 86 McKinsey, Mogelijkheden, p. 1. 87 Van Gorp, Tilburg, pp. 196–197. 88 W. Asbeek-Brusse, ‘De Internationale handel in textiel en kleding, 1950–1990’, Textielhistorische Bijdragen 34 (1994): 202–225, at p. 205. 89 Jan Luiten van Zanden and Richard Griffiths, Economische geschiedenis van Nederland in de 20e eeuw (Utrecht, 1989), p. 273. 90 C. Wolters, ‘De Twentse textielindustrie en de dekolonisatie. De neergang van de handel in “katoentjes” naar Indonesie’, Textielhistorische Bijdragen 30 (1990): 109–132, at p. 120. 393

The Ashgate Companion to the History of Textile Workers textile industry in Twente applied the shift system more than other branches in the Netherlands. In 1954 about half of all textile workers over 18 worked shifts.91 Moreover, manufacturer families (mainly in Twente) started to rationalize the management of their companies by appointing managing directors from outside the family and professionalizing their personnel departments. Factories were no longer as paternalistic as they used to be, but still the textile companies had a profound influence on the lives of the workers outside working hours. There were also a lot of technical developments, such as the introduction of Sultzer looms. In 1948 the industry organized itself with the foundation of the Vereniging Katoen-, Rayon- en Linnenindustrie (KRL), whereas labour relations had been settled in a Collective Labour Agreement in 1946.92 In the wool industry in Tilburg, however, no rigorous modernizations were implemented in the period 1950–1965.93 Many manufacturers in Twente tried to strengthen their position through mergers. One of the earliest and most successful was the merger between De Koninklijke Stoomweverij te Nijverdal (KSW) and H. ten Cate Hzn en Co. The companies had already started working together in 1952, resulting in the strong and successful Nijverdal-ten Cate in 1957. The initiators of this merger realized that the traditional textile industry in Twente was going to end and that in order to survive, they should cooperate, merge, raise productivity and develop innovative products and discover new markets. Many other mergers followed but were less successful and usually meant only a stay in operation. Most companies foreclosed or went bankrupt. In 1962 six companies from Twente and one from Brabant established the Koninklijke Nederlandse Textiel Unie (KNTU), hoping to jointly gather orders from all over the world and divide them. Branches that were not doing well were rejected; branches that were performing well had to work together. Personnel were cut drastically, but the management that consisted of members of the manufacturer families was spared. After a few years so many companies had joined that the Union comprised 14 companies and had 38 directors. In 1973 the KNTU was declared bankrupt and broken up.94 The wholesale dismissals in April 1967 at Van Heek & Co, once the leading company of the cotton industry, were a deep blow and people started to realize how serious the situation really was. In Tilburg the situation was just as bad. In the years 1951 to 1978, 64 of 83 companies in Tilburg closed.95 Workers that were dismissed tried to find jobs at one of the remaining companies only to soon face yet another closure. This happened more than once; the sad fate of an age-old industry is symbolized by the fact that a worker sometimes lost six jobs in a row. 91 IJzerman, Beroepsaanzien, p. 213. 92 Nas, Het Twentse Model, p. 141. 93 Van Gorp, Tilburg, p. 199. 94 See Van Hooff and Rossel, ‘Human en public relations’; Frans van Waarden, Bert de Vroom and Jan Laurier, Fabriekslevens. Persoonlijke geschiedenissen van arbeiders, fabrikanten, managers en andere betrokkenen uit de Twentse textielindustrie (Zutphen, 1987), p. 97. 95 Keune, De textiel voorbij, p. 28. 394

The Netherlands The century-old production of basic textile commodities had virtually vanished from the Netherlands. In order to survive, the remaining textile companies underwent drastic changes, such as a move in the direction of synthetic textiles or an orientation towards high-quality products (the famous ‘Ten Cate underwear’). What the future will bring, historians cannot tell, but it may well be that even these companies will not survive very long, as the recent financial crisis has caused severe problems for the few remaining Dutch textile producers.

395

This page has been left blank intentionally

15 Poland Piotr Franaszek

The economic and social history of Poland has always been inextricably linked to its political history. This dependence is particularly visible in the case of the textile industry. The period extending from the mid-seventeenth century to the second half of the eighteenth century is characterized by a gradual deterioration of the Polish textile industry, caused by numerous, ravaging wars, and by the fall of the central and provincial government in Poland. Due to the partition of Poland, which led to the abolition of Polish statehood by the end of the eighteenth century, and which were followed by new demarcation lines after the Congress of Vienna in 1815, the major centres of the textile industry, situated in Greater Poland (Wielkopolska) and incorporated into the Prussian state, disappeared. From the Middle Ages this had been the biggest traditional region for the activities of spinners and weavers, who sold their products in central Poland. However, the disappearance of those textile centres stimulated the fast growth of completely new concentrations of spinners and weavers on Polish lands, which was mainly due to the introduction of a political and economic border between Prussian Greater Poland and the Russian Kingdom of Poland. This new demarcation line caused the migration of the textile workers of Greater Poland and led to the creation of a textile centre in Łódź. The introduction of a customs border between the Kingdom of Poland and the Russian Empire in 1833 forced some Polish weavers to move their workshops from the Kingdom to the region of Białystok, situated within the Russian borders but outside of the Kingdom. At the same time in Galicia, that is, in the Polish territories incorporated into the Austrian Monarchy, the old textile production based on cottage industry vegetated. Because of the importance of the Łódź district for the Polish textile industry since the nineteenth century, Polish textile production was concentrated here both after the regaining of independence by Poland in 1918, and during the Communist regime after 1945, despite the damages caused during the First and Second World Wars. The abolition of the Communist dictatorship, on the other hand, radically changed the situation of Polish industry, including the textile industry. The 1990s witnessed its rapid decline.  Wiesław Puś, Rozwój przemysłu w Królestwie Polskim 1870–1914 (Łódź, 1997), p. 75.

Map 15.1 Poland

Poland

The years 1650–1815 Until the end of the eighteenth century industrial production in Poland, including textile production, remained in a guild system, which disintegrated gradually. This decline was particularly evident in big urban centres, where the rich patriciate had a policy that did not take into account the interests of craftsmen, who, in their turn, were also conflicted. Urban centres became depopulated. Under such circumstances, illegal crafts developed outside of guilds, mainly in rural areas. During this period textile production dealt chiefly with the processing of wool, flax and hemp, whereas the first workshops processing cotton appeared only at the end of the eighteenth century. The production of woollen products was concentrated in urban craft workshops, which operated within the guild system. In the second half of the seventeenth century elements of the cottage industry system, which organized textile production outside of urban areas, began to appear. The main area of woollen production was the woollen cloth centre of Greater Poland, and the region in the south of Poland on the border of Silesia and the Commonwealth also gained importance. The development of the woollen cloth industry in Greater Poland in the seventeenth century was boosted by the rise in demand for woollen cloth among the noblemen, although at the same time demand for this product on the part of the impoverished townspeople and peasants dropped. The woollen cloth industry developed thanks to the weavers arriving in the towns of Greater Poland from Silesia. The first seeds of capitalism appeared in the textile industry of this region, with merchants providing weavers with yarn and, in return, receiving cloth from them. The main centres for the woollen cloth industry in Greater Poland, and especially of the weaving industry, were the towns of Wschowa, Rawicz, Bojanów and Leszno. In Wschowa there were 180 weaving shops, in Rawicz 327 cloth-makers owned 200 looms for weaving wool, in Bojanów 175 weaving shops operated and in Leszno there were 138. Apart from Greater Poland, another important centre for the woollen industry was the textile centre of Bielsko and Biała, situated in the south of Poland, on the border with the Austrian Teschen Silesia. The woollen industry appeared there only in the seventeenth century. It was established by members of the Evangelical Church who fled from religious persecution in Silesia. They would buy yarn from merchants and sell them cloth which they produced. Throughout the seventeenth and eighteenth centuries the woollen industry of the region flourished. Thick woollen fabrics met local needs, and the surplus was exported further away. In the

Mirosław Krajewski, Historia gospodarcza Polski do 1989 roku. Zarys problematyki (Włocławek, 2000), p. 100.  Antoni Mączak, Sukiennictwo wielkopolskie XIV–XVII wiek (Warsaw, 1955), pp. 188, 295–296.  Karol Bajer, Przemysł włókienniczy na ziemiach polskich od początku XIX w. do 1939 r. Zarys ekonomiczno-historyczny (Łódź, 1958), p. 33. 

399

The Ashgate Companion to the History of Textile Workers eighteenth century the production of more delicate fabrics was launched thanks to the importation of woollen yarn from Hungary and Prussia. In contrast to the organization of the woollen cloth industry, the production of linen was the domain of cottage industry. Ownership of the tools and raw materials lay within the financial possibilities of a rural household. The domestic production was scattered all over the territory of Poland, but its biggest concentration was in Silesia and along the Sudetes and the Carpathians, with the town of Andrychów as its major centre. The workplace was the peasant’s house, and the labour force was the family. This is how a Central European centre of peasants’ textile industry was created, which was the first region among Polish lands to witness the processes of proto-industrialization. Beginning in the 1740s, merchants provided peasants with yarn, collected linen from them and directed it to subsequent stages of processing, that is, bleaching and mangling. The linen enjoyed great popularity not only in Poland; it was also exported to Russia, Western European states and even to Asia, Africa and America. The first manufactory in the Polish textile industry appeared in the second half of the eighteenth century. Attempts to establish manufactories were made by the rich townspeople of big centres such as Krakow, Warsaw and Poznan. In Poznan such an enterprise was owned by a banker named Klug. In Krakow a well-equipped centralized workshop owned by Frysztacki successfully manufactured woollen cloth. In 1786 the canon of the cathedral of Krakow, Father Wacław Sierakowski, set up a wool manufactory which hired orphans and the homeless. Manufactories producing woollen fabrics were also established by magnates. Famous enterprises of this kind included the manufactories of the Radziwiłł family in Nieśwież, Naliboki and Urzecze and those of the Potocki family in Machnówka and Niemirów. However, in contrast to townspeople’s manufactories, they manufactured goods chiefly for the families’ own needs, and not for the market. The last Polish king, Stanislaw Augustus Poniatowski, was also involved in establishing Polish wool manufactories. At his suggestion the Company of Woollen Workshops was created in 1766. Unfortunately, it went bankrupt after a few years due to bad management.10 By the end of the eighteenth century the first manufactories processing cotton were established in Poland. Five such plants were registered in 1788, compared to 25 wool, 10 silk and 9 flax and hemp-processing manufactories.11

 Ibid., pp. 189–191.  Mariusz Kulczykowski, Andrychowski ośrodek płócienniczy w XVIII i w XIX wieku (Wrocław etc., 1972), pp. 20–21.  Mariusz Kulczykowski, ‘Geneza i znaczenie historyczne manufaktury domowej w centralnej i wschodniej Europie’, Studia Historyczne 34 (1981): 169–182, at pp. 172–174.  Kulczykowski, Andrychowski ośrodek płócienniczy, pp. 13–14.  Bajer, Przemysł włókienniczy na ziemiach polskich, pp. 33–34. 10 Krajewski, Historia gospodarcza Polski, pp. 130, 139. 11 Bajer, Przemysł włókienniczy na ziemiach polskich, p. 35. 400

Poland

The years 1815–1918 Another division of Polish lands took place after the Vienna Congress. The new customs border introduced in 1822 between two parts of Poland – the Kingdom of Poland, joined to Russia, and the Great Duchy of Poznan (Greater Poland) which became part of Prussia – cut the textile production of Greater Poland off from its hitherto existing markets – the lands of central Poland. Moreover, the level of technological development of the weaving and spinning industries of Greater Poland was not advanced enough for them to compete successfully with textile products from other centres located in Prussia. Textile production in Greater Poland declined rapidly.12 Confronted with this situation, spinners and weavers migrated from the old centres of cloth industry to the Kingdom of Poland. Consequently, in the 1820s a large new centre for the woollen industry was created in the Kingdom of Poland and this is where the seeds of industrial cotton production were first planted. Until 1830 the Kingdom of Poland witnessed a boom in the production of woollen cloth in the early factory system. This period of spectacular growth was related to the opening of the unlimited Russian market to textile products, thanks to important customs facilities on the border between the Kingdom of Poland and the Russian Empire. Unfortunately, in 1829 production crashed. The crisis deepened due to customs restrictions introduced by the Russian authorities in 1833, after the failure of the November Uprising in 1831. The export of thick woollen cloth to Russia was limited dramatically. Between the years 1829 and 1832, the volume of production decreased by almost two-thirds, and for medium cloth by approximately 33 per cent.13 This caused another wave of migration of textile workers. They settled in Russia, right next to the border with the Kingdom of Poland. In such circumstances the textile centre of Białystok was created. The crisis in the textile industry had yet another positive consequence. It encouraged the dynamic development of the cotton industry, which was concentrated between the towns of Łęczyca, Sieradz and Tomaszów Mazowiecki, a region where the textile industry had mainly concentrated on woollen production. As foreign markets were lost and the domestic demand for woollen products decreased, the cotton industry became the key factor in the rapid development of the textile industry. This is when Łódź gained an important position, soon becoming one of the chief regions of textile production worldwide. In the early twentieth century in Łódź and its region there were 28 joint-stock companies dealing in textiles. The positive trends in the manufacturing of cotton boosted the woollen industry, which saw a considerable increase in production, due to the process of fast mechanization which had carried on since the mid-nineteenth century. Apart from the Łódź region, between 1870 and 1880 a new textile centre developed. This new Sosnowiec-Częstochowa centre, situated next to the Prussian 12 Gryzelda Missalowa et al., ‘Przemysł włókienniczy do 1918 r.’, in Irena PietrzakPawłowska (ed.), Uprzemysłowienie ziem polskich w XIX i XX wieku. Studia i materiały (Wrocław etc., 1970), pp. 219–276, app. III, pp. 258–259, 261, 264, 266. 13 Missalowa et al., ‘Przemysł włókienniczy do 1918 r.’, p. 220. 401

The Ashgate Companion to the History of Textile Workers border and along the Warsaw–Vienna railway, appeared as a result of the influx of Belgian, French and German capital. At the same time yet another centre for the woollen industry was set up by English capital in the town of Marki situated next to Warsaw.14 The dynamic development of the textile industry, gaining power from the 1870s, was made possible when the Russian textile market opened wider for Polish goods, after the introduction of protective customs tariffs by the Russian authorities in 1877, and also because of the influx of foreign capital to the country.15 Until the mid-nineteenth century manual labour dominated in wool production, which took the form of cottage industry or was organized in workshops. The mechanization of the woollen cloth industry in the Kingdom of Poland initially happened very slowly. On the other hand, cotton production was fully mechanized from the very beginning. Textile enterprises introduced organizational changes rapidly and concentration processes intensified. First monopolistic agreements were signed, which divided markets between companies and established the volume of production.16 The dynamic development of the textile industry in the Kingdom of Poland, characterized by the rapid growth of employment and the increasing number and power of the machines used in it, was mirrored by the fast growth of production value. Between 1879 and 1913 it increased over fivefold. With time, the production value of factories employing a small number of workers went down, whereas at the same time the production value of big and very big enterprises increased rapidly. Until the mid-nineteenth century it was very uncommon to hire peasants in textile factories. However, they would sometimes participate in supporting works for the factories in their area, such as transport, building roads and so on, especially in industrial regions.17 At that time the main workforce comprised bankrupt craftsmen, navvies, journeymen and other people of urban origin who had previously worked within guilds or in domestic industry. Moreover, in the 1820s the authorities of the Kingdom of Poland supported and even organized the immigration of weavers and spinners to the regions of Łódź, Pabianice and Zgierz from Bohemia, Austria, Prussia, Saxony, Bavaria and North Rhine. In this way about 100,000 textile workers settled in the regions.18 In the 1820s and 1830s, after having settled, the workers opened their own weaving shops. However, they were rapidly absorbed, first by the newly created manufactories, and then by factories. The majority of newcomer employees and the well-qualified ones were 14 Ibid., p. 226. 15 Władysław Rusiński, Zarys historii gospodarczej Polski (Warsaw, 1986), pp. 166–168. 16 Karol Badziak, Przemysł włókienniczy Królestwa Polskiego w latach 1900–1918 (Łódź, 1979), pp. 97–129; Wiesław Puś, ‘Robotnicy w przemyśle włókienniczym ziem polskich w XIX i na początku XX wieku. Problem rekrutacji’, in Wiesław Puś (ed.), Studia z Historii Społeczno-Gospodarczej XIX i XX wieku (Łódź, 2003), pp. 7–19, at pp. 10–11. 17 Józef Śmiałowski, Zarobkowanie ludności rolniczej w Królestwie Polskim w latach 1815–1864 (Łódź, 1973), pp. 253–255. 18 Gryzelda Missalowa, Studia nad powstaniem łódzkiego okręgu przemysłowego (Łódź, 1964), I, pp. 50–52. 402

Poland made technical staff of the new ventures. Few managed to establish their own mechanized enterprises. The enfranchisement of peasants in 1864 liberated great resources of labour, which moved towards the newly created industrial centres. In the years 1875–1895 65 per cent of the people arriving to work in the textile factories of Łódź came from villages. The shift of the rural population, previously involved in domestic production, to textile factories in Łódź, was also due to the absorption of villages by dynamically developing urban textile centres, as was the case in Łódź, Białystok and other towns and cities. In the early twentieth century the number of second and third generation workers employed in the textile industry increased.19 As textile production grew, the employment in spinning and weaving plants increased rapidly. In 1870 20,200 persons were employed in the textile industry of the Kingdom. This number doubled in less than eight years and in 1878 it amounted to 43,300 workers. In the mid-1880s almost 63 per cent of the people working in the textile industry were employed in the Łódź district. The Warsaw district totalled 9,400 (22 per cent), the Sosnowiec-Częstochowa district 6,500 (c.14 per cent) and the small Kalisz district only 700. At that time textile workers constituted 45 per cent of people employed in all industrial sectors in the Kingdom of Poland and approximately 15 per cent of all the textile industry workers of the Russian Empire. In the Białystok district about 5,700 textile workers were employed.20 In 1897 the number of labourers working in the textile industry exceeded 100,000, in 1910 150,000 and a year before the outbreak of the First World War, in 1913, this number went beyond 163,000. The dynamics of employment in the textile industry in the years 1870–1913 in the Kingdom of Poland is presented in Figure 15.1. Thanks to the rapid mechanization of textile factories at the turn of the twentieth century, a particular kind of cooperation developed between enterprises and cottage-workers. For instance, one of the largest cotton enterprises, Towarzystwo Akcyjne I.K. Poznański, which in 1901 employed 6,700 workers, cooperated with over 500 cottage-workers.21 More than 3,000 cottage-weavers from the adjacent villages and little towns worked for the plants of K. Scheibler and Poznański in Łódź and of Kindler in Pabianice. Estimates indicate that in the late 1880s and early 1890s the number of cottage-workers amounted to about 8 per cent of the total number of workers in the textile industry. Taking the workers’ sex and age into consideration, it must be pointed out that until the mid-nineteenth century, that is until the beginnings of the mechanization processes, the majority of those employed were men. From the mid-1880s the number of the women employed became stable and amounted to approximately 50 per cent of all workers. However, in some branches of textile production the prevalence of women was much greater. In Łódź in the mid-1880s, women amounted to 75 per cent of the employees of the cotton factory owned by I.K. Poznański.

19 Puś, ‘Robotnicy w przemyśle włókienniczym ziem polskich’, I, pp. 7–19. 20 Missalowa et al., ‘Przemysł włókienniczy do 1918 r.’, pp. 226–227. 21 Puś, ‘Robotnicy w przemyśle włókienniczym ziem polskich’, pp. 17–18. 403

The Ashgate Companion to the History of Textile Workers         

              



  



  



  



  



  



  







 



Figure 15.1 Employment in the Polish textile industry, 1870–1913. Source: W. Puś, ‘Robotnicy w przemyśle’, pp. 7–19, 17–18. At the same time the percentage of children working in factories decreased. Between 1885 and 1895 the fraction of children among the workers of the textile industry went down from 9.7 per cent to 2.5 per cent (from 4,300 to 2,200). In the cotton industry this percentage went down from 11.6 per cent to 4.3 per cent, in the woollen industry from 10.5 per cent to 1.8 per cent, in the flax, jute and hemp industry from 5.4 per cent to 0.2 per cent, and in the silk industry from 1.8 per cent to 0 per cent. This was a result of the tsar’s ukases of 1882 and 1884, which comprised legal regulations protecting children, and of the activities of the Factory Inspection. The percentage of technical staff and foremen among the employed was comparatively small. By the end of the 1880s in Łódź they amounted to 1.5 per cent of all workers, and in the Częstochowa district 4 per cent of all workers. Until the mid-1890s they were mostly German.22 The working hours of the employees of the textile industry fluctuated between 12 and 13 hours per day. In 1897 the working day was limited by law to 11.5 hours. Much more was done to shorten the working day of children and women. According to the legal regulations of 1882 and 1884, the children’s working day was to be a half of the adults’ shift. The regulations of 1885 prohibited the work of women and 22 Missalowa et al., ‘Przemysł włókienniczy do 1918 r.’, pp. 230, 235–236, 244–245. 404

Poland juvenile workers at night. However, due to the protests of the owners of enterprises, this ban was lifted five years later. The aforementioned regulations also banned the work of children in conditions hazardous to health, that is in dye-houses, finishing shops and on posts which required the cleaning of moving machines. These rules were, however, often breached. Examples were the constant hiring of children in the dye-house of the factory of M. Baruch in Pabianice, in the Łódź district, and the numerous accidents suffered by children in the factories owned by Scheibler and Poznański in Łódź. With the establishment of the Kingdom of Poland in 1815, employment in the factories of the textile industry was based on free hire, and labour relations were regulated by the provisions of the Code Napoleon and the decisions of the governor of the Kingdom of the years 1816, 1817 and 1823. After the failure of the January Uprising in 1864, Captain Schiemann, the commander in chief of Łódź, issued regulations which obligated textile workers to have employment books which in practice forced them to work in one factory on the terms given by the employer, without the possibility of changing their workplace. Labour contracts were concluded orally by factory owners with each worker individually. In 1891 the so-called workers’ books were introduced, intended as a written form of working agreement. However, in reality they were yet another form of pressure, making the worker dependent on the owner of the factory. Until the beginning of the First World War there were no state regulations defining minimum wages in the textile industry. There were considerable differences in the payment of different groups of textile industry workers, between different factories and types of production. The wages of women were no higher than 75 per cent of the wages of men, while those of children amounted to between 20 and 30 per cent of the payment for the work of a male adult worker. Wages were highest in the woollen cloth industry. Dyers and printers earned the most; spinners were paid less than weavers. The average weekly wage in the textile industry of the Łódź district was approximately 3.5–4.5 rs (silver rubles), with the minimal costs of maintaining a five-person family being 6 rs. What is more, in compliance with the decision of administrative authorities of the years 1866 and 1886, 1 to 3 per cent of the wages were subtracted for the Sickness Funds of the factories. In spite of difficult living conditions, textile industry workers did not have the opportunity to organize legal trade unions, as such activities were perceived by the authorities as conspiratorial and they were fiercely fought against. The newly founded labour parties which, also illegally, tried to establish contacts among workers often took advantage of this situation. In Łódź there existed the Polish Workers’ Association, and at the turn of the nineteenth century also the Polish Labour Party and the Social Democratic Party for the Kingdom of Poland. In 1892 the Polish Workers’ Association organized a manifestation in Łódź on the occasion of the First of May Holiday. Their campaigning led to the outbreak of a general strike which took place between 3 and 10 June. The strike, nowadays known as the ‘Łódź rebellion’, was brutally crushed by the army and the police. Over 200 people were killed or injured.

405

The Ashgate Companion to the History of Textile Workers The weavers and spinners of Łódź stood up for their rights once again on 23 June 1905. This event was a part of the revolution of 1905 which spread across the industrial centres of the Russian Empire. Fights with the army, and especially with Cossack units, lasted for three days. The number of fatalities reached 200 people again. The events in Łódź belong to one of the largest and most tragic demonstrations in the whole empire. During the thaw in March 1906 the authorities issued temporal regulations which legalized trade unions. However, those organizations were not allowed to appeal for wage rises, or for the shortening of working days, nor were they permitted to organize strikes. Such activities were viewed by the authorities as political and they were strictly fought against. Consequently, trade unions had no opportunities to pursue their objectives and they continued their struggle for the improvement of social and living conditions illegally.23 More detailed information about textile enterprises on Polish lands under the Austrian partition can be found in the census of enterprises, carried out in the Austro-Hungarian Monarchy in 1902. According to its records, in 1902 there were 206 mechanized factories in Teschen Silesia, which amounted to almost 68 per cent of all textile enterprises of the Austrian partition, out of which as many as 111 enterprises operated in Bielsko. Among the 206 enterprises 92 were small-sized, employing no more than five persons. Until the outbreak of the First World War there were no enterprises employing more than 1,000 workers. The largest number dealt with wool processing. They employed about 6,000 workers. At the same time only six factories produced cotton yarn and fabrics. Half of them were large-sized, each employing over 100 persons. In total they provided work for 1,500 persons. Weavers were the most numerous group among the textile workers of Galicia, which was particularly visible in the wool weaving mills in West Galicia. In East Galicia until the end of the nineteenth century a great percentage of workers were weavers of flax and jute. However, with time their number went down considerably. Spinners were much less numerous among the textile workers, both in wool- and cotton-processing (see Table 15.1). In compliance with the industrial law of 1859, workers in the Austrian Monarchy held workers’ books, which were the basis for concluding working agreements. Until 1883 the employment of children below the age of 14 was prohibited in the Austrian Monarchy. The 1883 and 1885 amendments to industrial legislation prohibited the work of children and youths at night and defined the working day of children as 8–10 hours, and of youths in the age range of 14–16 as 11–12 hours. The prohibition of women’s work at night was introduced in 1911.

23 Stefan Kieniewicz, Historia Polski 1795–1918 (Warsaw, 1975), pp. 417–418, 441–446. 406

Poland Table 15.1 Textile industry workers in Galicia, 1890 and 1910 Type of production

West Galicia

East Galicia

Total Galicia

1890

1890

1890

1910

1910

1910

Silk spinning mills

20

5

10

5

30

10

Silk weaving mills

6

11

29

17

35

28

Wool spinning mills

18

150

15

4

33

154

Wool weaving mills

1,956

5,145

83

46

2,039

5,191

Cotton spinning mills

25

33

60

37

85

70

Cotton weaving mills

119

20

50

8

169

28

Jute & flax spinning mills

148

175

80

7

228

182

Jute and flax weaving mills

672

255

1,847

26

2,519

281

Other

490

1,045

766

1,280

1,256

2,325

3,454

6,839

2,940

1,430

6,394

8,269

Total

Source: Missalowa et al., ‘Przemysł włókienniczy do 1918 r.’, pp. 262, 265.

During the First World War textile production on Polish lands decreased markedly, and textile factories incurred massive losses. The majority of the enterprises in the Łódź district were damaged either in direct war activities or as a consequence of pre-planned devastation, carried out by both German troops and the Russian army withdrawing from the Kingdom of Poland. The complete equipment of factories and their staff was evacuated to Russia. As a result, as early as the beginning of 1916, 421 factories out of the 488 operating in 1914 had closed. Employment fell to 31,500 workers, which constituted 20 per cent of the total number of those employed in 1913. In 1917 the remaining textile factories were destroyed. In contrast to the experience of Łódź, the textile industry of the Białystok district did not suffer any major losses.24

The years 1918–1945 After the revival of the Polish state in 1918, it contained three centres of textile industry within its borders: Łódź, Bielsko-Biała and Białystok. They continued to function to a large extent as they had before the war. The Łódź district remained the strongest, 24 Bajer, Przemysł włókienniczy na ziemiach polskich, p. 188. 407

The Ashgate Companion to the History of Textile Workers comprising two-thirds of all the people employed in the Polish textile industry in the period between the two world wars. This industry activated other branches of industry operating for the needs of the textile industry.25 Despite the unquestionable progress the Polish textile industry witnessed in the interwar period, it never reached the level of production it had before the First World War. The percentage of cotton processed, against the figures from 1913, amounted to 85 per cent in 1937, and of woollen yarn to 44 per cent in 1935. In the mid-1930s, Poland processed only 6 per cent of the cotton processed worldwide and only 1.5 per cent of wool.26 The textile industry of the Łódź district became extremely hampered by the loss of the previously unlimited Russian market. Adapting the sales of textile products to the requirements of the internal market called for the reorganization of production processes and for their readjustment to manufacturing yarn and fabrics of much higher quality. The concentration of the Polish textile industry in Łódź is best characterized by the number employed in the spinning and weaving mills of the city against employment in the overall Polish textile industry (see Table 15.2). Table 15.2 Textile industry workers in Poland and in Łódź, 1921–1937 Poland

Łódź

(%)

Poland

Łódź

(%)

1921

n.a.

74,179

n.a.

1930

147,000

75,529

51

1922

166,410

96,626

58

1931

133,800

64,865

48

1923

135,882

102,143

75

1932

109,843

71,164

65

1924

n.a.

77,208

n.a.

1933

128,333

77,725

61

1925

149,262

50,654

34

1934

130,453

74,055

57

1926

157,995

77,554

49

1935

139,010

79,482

57

1927

182,103

98,769

54

1936

156,160

87,607

56

1928

189,099

96,884

51

1937

157,067

n.a.

n.a.

1929

178,800

78,567

44

Sources: Missalowa et al., ‘Koncentracja przemysłu włókienniczego (1918–1939)’, Appendix V, p. 404; Mierzejewski, ‘Zatrudnienie w przemyśle’, pp. 172–173.

In 1923 existed the highest, and in 1925 the lowest employment in the textile industry of Łódź in the whole interwar period.27 The situation of the district 25 Ibid., p. 201. 26 Gryzelda Missalowa et al., ‘Koncentracja przemysłu włókienniczego (1918–1939)’, in Pietrzak-Pawłowska (ed.), Uprzemysłowienie ziem polskich, pp. 387–431, at pp. 393–395, 403; Barbara Wachowska, ‘Walka o układy zbiorowe w łódzkim przemyśle włókienniczym w 1933 r.’, Rocznik Łódzki 12 (1967), pp. 115–167. 27 Stanisław Mierzejewski, ‘Zatrudnienie w przemyśle włókienniczym Łodzi w okresie 408

Poland improved visibly beginning in 1926, and it remained in a good economic condition until 1928. Yet in the next year a depression started, and the following four-year economic crisis had a big impact on the Polish textile industry. The prices of textile raw materials plummeted, which was accompanied by an even more rapid fall in the demand for textile products, especially due to the deteriorating economic situation of Polish villages. As a consequence, the size of production was often limited, factories were closed down and employment went down. At first large enterprises producing mainly for export were closed down, for example the factories of K. Scheibler, L. Grohman and L. Geyer. In the first half of January 1930 the majority of factories did not operate, while others were open for only two or three days a week. Soon smaller enterprises, manufacturing goods mainly for the internal market, followed. The most noticeable effect of the crisis, and at the same time the most tangible for the workers, was the bankruptcies of enterprises operating in the textile industry.28 In subsequent years, the proportions between the number of women and men employed in the textile industry of Łódź remained at a steady level, about 50 per cent. In other words, throughout the whole period the number of employed women more or less equalled those of working men, and changes in the number of workers of one sex corresponded to those in the number of the other. The dynamics of the employment of women and men in the textile industry of Łódź is presented in Figure 15.2.        







PHQ









ZRPHQ

Figure 15.2 Women and men employed in the Łódź textile industry, 1921–1936. Source: Mierzejewski, ‘Zatrudnienie w przemyśle’, p. 172. międzywojennym’, Rocznik Łódzki 12 (1967): 169–191, at pp. 172–173. 28 Bajer, Przemysł włókienniczy na ziemiach polskich, pp. 203–207, 215. 409

The Ashgate Companion to the History of Textile Workers Juvenile workers of both sexes formed a comparatively small fraction of the employed. In 1925 a considerable number of juvenile workers were dismissed as a result of the law on the obligation to educate children and youths, which had been issued in the previous year. In the period of a more favourable economic situation, in the years 1926–1928, the number of juvenile employees increased sharply, although this increase was also due to the binding regulations which allowed for their employment without remuneration as vocational training. The situation changed after 1931, when such practices were banned. Consequently, in the subsequent years, until the outbreak of the Second World War, the number of juvenile employees declined systematically. The percentages of women, men and juvenile workers of both sexes employed in the textile industry of Łódź in the years 1922–1936 are presented in Table 15.3. Table 15.3 Workers in textile firms according to sex and age, Łódź 1922–1936 (%) Men

Women

Boys

Girls

1921

48.9

47.8

1.7

1.6

1922

48.8

47.6

1.8

1.8

1923

48.4

47.2

1.9

2.5

1924

50.2

46.2

1.6

2.0

1925

48.1

48.6

1.5

1.8

1926

48.4

47.8

1.6

2.2

1927

46.7

48.8

1.9

2.6

1928

46.9

48.4

2.0

2.7

1929

47.8

48.5

1.6

2.1

1930

47.6

49.8

1.2

1.4

1931

48.7

49.4

1.0

0.9

1932

50.4

48.4

0.6

0.6

1933

48.9

49.8

0.7

0.6

1934

47.7

51.2

0.5

0.6

1935

46.9

52.0

0.4

0.7

1936

46.6

52.5

0.4

0.5

Source: Mierzejewski, ‘Zatrudnienie w przemyśle’, p. 172. 410

Poland Right after the end of the First World War, labour relations in the textile industry were regulated by the newly introduced employment legislation. From 23 November 1918 the binding working day in all branches of industry was eight-hours long. At the beginning of 1919 obligatory insurance in the event of illness was introduced. From 1922 legal regulations were introduced concerning the length of leave, and in two years employing women and juvenile workers at night was banned. In the years 1919–1924 collective agreements, regulating, among other things, the amount of wages, were established.29 During the interwar period, the rising trade unions were captured by party activists, who in this way tried to obtain the support of workers. Hard living conditions were thus used by socialists, Communists and nationalists for the pursuit of their own political schemes. Workers’ appeals were most often organized by party activists and not by trade unionists. In the years 1918–1939, the period of great economic crises, serious strikes took place in the textile industry. The decrease in textile production in Poland had already begun in mid-1929, and the bottom of the crisis was reached in Łódź in 1932, when production totalled 54 per cent of the volume of that in 1928. In the autumn of 1931 the first strikes were organized in small textile enterprises that were not covered by the joint agreement of 1928. In April 1932 the Textile Industry Employers Association terminated the agreement. The attempt to organize a general strike on 16 June 1932 was a reaction to this decision. However, few enterprises and only 5,000 workers joined the action. The strike broke up. Yet the further decrease in wages, by 15 per cent in the first two months of 1933, led to the proclamation of the general strike of the textile workers of Łódź on 6 March that year. A total of 120,000 workers took part in the strike, the biggest one in Poland in the crisis period. Soon the textile factories in Żyrardów and Białystok joined the strike. On 17 March violent battles with the police took place in Pabianice. Five people died in the scuffle, and another worker was killed in Łódź on 30 March.30 The workers’ demands included the bringing back of the joint agreement of 1928, to include also small textile enterprises, as well as rejecting the so-called consolidating agreement which had caused the conditions of insurance, health care and social benefits of textile workers to deteriorate. Practically, the whole textile industry of Łódź and surrounding areas came to a halt. The strike continued and there was no agreement between the employers and the strikers. The Ministry of Labour then entered the negotiations. The compromise signed on 28 March allowed for the possibility of reducing the wages of spinners by 12 per cent, and of weavers by 15 per cent. Such decisions triggered off violent demonstrations and even fights with the police on 30 and 31 March. Workers returned to the factories on 4 April. This strike of textile workers was the biggest industrial action in Poland in the years 1918–1939.31 29 Missalowa et al., ‘Koncentracja przemysłu włókienniczego’, pp. 387–431, 393–395, 403; Wachowska, ‘Walka o układy zbiorowe’, pp. 152–153. 30 Lucjan Kieszczyński, Powszechny strajk włókniarzy Łodzi i okręgu w 1933 r. (Warsaw, 1958), pp. 8, 11–17. 31 Kieszczyński, Powszechny strajk, pp. 67, 94–95. 411

The Ashgate Companion to the History of Textile Workers

The years 1945–1990 The losses sustained by the Polish textile industry during the Second World War were not as massive as in other branches of industry. The textile factories in Tomaszów Mazowiecki, Kalisz, Częstochowa and Bielsko Biała avoided damage, yet in Łódź the German occupants of the textile factories of Osser, Rosenblatt and Silberstein completely destroyed the stock of machines. The Jews of Łódź, who constituted a considerable percentage of the inhabitants of the city, including the owners of textile factories, were murdered in the Holocaust. Approximately 30 per cent of textile factories were destroyed. German policy combined smaller factories into bigger ones, to make them more economical to manage with regard to production for war needs. As a result, according to the census of 1945, there were only 337 textile factories registered, against 1,172 in 1936.32 Factories in Białystok were destroyed completely.33 The seizure of political power in Poland by the Communists had far-reaching economic and social consequences. Among the most immediate was the adoption of the Soviet model of the economy, characterized by the introduction of the policy of command and distribution. The result was the nationalization of industry in 1946, and the implementation of national long-term economic plans, also affecting the textile industry. Therefore investments, the volume of production and employment in the textile industry were managed centrally, during the so-called Three-Year Plan (1947–1949), the Six-Year Plan (1950–1955) and the subsequent Five-Year Plans. The rule that was binding in both theory and practice until the second half of the 1950s, and in practice throughout the whole period of Communism, was from Marxist-derived economic doctrine. According to this rule, industries dedicated to increasing the means of production had to develop faster than those producing goods for consumption. As a consequence of such an approach, the investment and raw material policy of the state was focused on the development of heavy industry. Because the textile industry was among the branches producing goods for consumption, investment expenditures in the textile industry in the years 1946–1949 were comparatively low.34 The situation in this respect deteriorated further during the implementation of the Six-Year Plan (1950–1955), when the government carried out the policy of intensive development of the armament and heavy industries. At that time only 5 per cent of investment expenditures were spent on the consumer goods industry.35 Already in the final stages of the Second World War, the Polish textile industry had manufactured goods for the needs of the Soviet and the Polish army. After the war the USSR became the main consumer of Polish textile production. In 1945 32 H. Wiśniewska-Bieniek, ‘Przemysł włókienniczy Łodzi w latach 1945–1946’, Rocznik Łódzki 15 (1971): 59–75, at p. 62. 33 Andrzej Jezierski et al., Historia gospodarcza Polski Ludowej 1944–1975 (Warsaw, 1982), p. 192. 34 Andrzej Jezierski et al., Historia gospodarcza Polski (Warsaw, 1997), p. 422. 35 Jezierski et al., Historia gospodarcza Polski Ludowej, p. 192. 412

Poland cotton and woollen raw materials from the stocks left behind by the Germans were used in the manufacturing of fabrics. Beginning with the following year, cotton was imported from the USSR, to where in turn the majority of fabrics produced in Poland were exported. In 1946 Poland began importing raw materials from Western countries, and a part of the cotton used for production was provided within the UNRRA supplies.36 The main direction of the activities of the Polish textile industry after the Second World War was the production of cotton yarn and fabrics, and, on a much smaller scale, the manufacturing of woollen and flax yarn and fabrics (see Figures 15.3 and 15.4). 









 











FRWWRQ

 ZRRO









IOD[

Figure 15.3 Production of yarn in Poland, 1946–1989. Sources: Rocznik statystyczny (Warszawa 1955–1993); Rocznik statystyczny przemysłu 1945–1965, (Warszawa 1967). Due to the state’s investment policy, the production volume of cotton and woollen yarn grew steadily until the end of the 1970s. The highest production volume of both types of yarn was obtained in 1978 (225,000 tons of cotton yarn and 108,000 thousand tons of woollen yarn). The high production level in the 1970s was an element of the economic policy of the government connected with the Communist faction of Edward Gierek. It was the policy of intense industrialization 36 Jezierski et al., Historia gospodarcza Polski, p. 424. 413

The Ashgate Companion to the History of Textile Workers carried out thanks to loans from West European states and the USA. At the same time production of flax yarn decreased gradually, and its maximum volume was registered in the year 1974 (56,800 tons). This went along with a state policy aiming at the gradual termination of the production of flax goods. Similar tendencies can be observed for the production of cotton and woollen fabrics, which went up rapidly until the end of the 1970s. The production of cotton fabrics was much larger and it grew much faster than that of wool. The largest amount of cotton fabrics was manufactured in 1977 (951 million metres), and of woollen in 1976 (125 million metres).            











FRWWRQIDEULFV











ZRROHQIDEULFV

Figure 15.4 Production of fabrics in Poland, 1946–1989. Source: Rocznik Statystyczny (Warszawa 1955–1993); Rocznik statystyczny przemysłu 1945–1965, (Warszawa 1967). The deep political and economic crisis at the beginning of the 1980s also had a dramatic impact on the textile industry. In 1982 the production volume of cotton yarn amounted to less than 80 per cent of the level of production in the year 1980, of woollen yarn less than 70 per cent and of flax yarn approximately 77 per cent. In subsequent years the authorities tried to improve the situation, which resulted in an increase in production during the second half of the 1980s, especially of cotton yarn. The production of cotton fabrics was subject to similar tendencies. In 1982 it amounted to slightly over three-quarters of the production level in 1980. It was 73 per cent of the production volume of 1977, which was the year of the highest production in the twentieth century. A similar tendency could be seen in the 414

Poland production of woollen fabrics, although for this type of production the fluctuations were not as spectacular as for other types. Until the mid-1980s the growth of production in the woollen industry was to a great extent the effect of increasing investment expenditures, and until the mid1970s also of the increase in the number of people employed in the industry (see Figure 15.5).                  

Figure 15.5 Employment in the Polish textile industry, 1946–1989. Source: Rocznik statystyczny (Warszawa 1955–1993); Rocznik statystyczny przemysłu 1945–1965, (Warszawa 1967). The sharp rise in employment within the first five years following the end of the war was related to the rebuilding of the textile industry and the regaining of the production capacity from before the Second World War. On the other hand, the fall in employment in the first half of the 1950s was the immediate effect of the implementation of the so-called Six-Year Plan, according to which the priority for the government was the development of heavy industry, while the textile industry and other branches of the consumer goods industry were marginalized. During the following 20 years employment went up systematically, reaching its peak in 1975 (478,200 persons). From this moment on until today the employment rate has been gradually decreasing. This falling tendency gained speed in the beginning of the 1980s, which, as has been mentioned above, was linked to the deep crisis of the Polish economy. Over the almost 50 years of socialist economy in Poland, the number of people employed in the textile industry amounted to approximately 10 per cent of the total 415

The Ashgate Companion to the History of Textile Workers number of workers in industry: a considerable figure. At the same time the textile industry was the most feminized branch of production. Women employed in this industry constituted approximately 60 per cent of the workforce, and in 1970 even 65 per cent. After the seizure of power by the Communists in Poland, trade unions became subject to them as well. Consequently, the structures of the unions were organized by the authorities and were to be one of the tools used to supervise particular social groups. Trade unions were organized according to the industrial sector, and that was one of the reasons for establishing the Trade Union of Textile Industry Workers. All trade unions were controlled by the Central Council of Trade Unions, led by the Communist Party. According to the Communists, with the ideological assumption that workers are owners of the enterprises, there was no conflict between employers and employees. Therefore the role of trade unions was to deal solely with social issues such as organizing workers’ holidays and children’s summer vacations or New Year’s Eve parties. In the period of the Communist regime the employees of light industries, including the textile industry, belonged to the worst paid group of workers. The great number of women working in the textile industry of Łódź accounted for the fact that Łódź could withstand more hardship than other cities and it did not always join the waves of strikes which broke out in Poland. Therefore in the Communist period it was often said that if spinners, seamstresses and weavers decided to stop working, it meant that the situation was really bad. This is what happened in the years 1957, 1971 and 1980. The strikes that took place in Łódź in the years 1957–1958 were part of a series of events that began with workers’ demonstrations in Poznan in June 1956. They were a reaction to the dramatic deterioration of living conditions, during the implementation of the Six-Year Plan. The main group of strikers were municipal transport workers, but the action was joined by the workers of several textile enterprises. The biggest industrial action was witnessed by Łódź in 1971. Already in December 1970, tensions caused by the violent events in Gdansk, Gdynia and Szczecin could be sensed in the factories. However, strikes were not organized. The industrial action was taken only on 10 February 1971. It was a result of a sharp rise in prices of basic food products and a dramatic deterioration of living conditions. Subsequently textile factories and enterprises in other sectors of Łódź joined the strike on 12 and 13 February. The action reached its peak on 15 February, when all the textile enterprises were already on strike. Such a progress of events caused great alarm to the new government led by Edward Gierek, which had come to power in December 1970. The growing wave of strikes could bring down this group. Therefore the Prime Minister of the Polish government, Piotr Jaroszewicz, went to the city. His address, however, only heightened the tension. Workers started sit-down strikes in their factories and there were scuffles with the police in the city. The government, scared by these developments, withdrew all the rises in food prices. This led to the end of the wave of strikes.37 37 Krzysztof Lesiakowski, ‘Strajki robotnicze w Łodzi w latach 1957–1980’, in 416

Poland The actions taken in Łódź in June 1976, when mass strikes began in other parts of Poland, were of a comparatively small range. Still they resulted in repressions from the authorities, by whose order over 120 employees were laid off. They received help from the Committee for the Protection of Workers (KOR), which operated also in Łódź. As early as the beginning of August 1980, strikes of a very limited scope took place in Łódź. Low wages, the high cost of living, poor supply of goods to shops, no possibility of providing food for children and disastrous working conditions in textile factories led to a weavers’ strike in the Peace Defenders Cotton Industry Plant on 4 and 5 August. The reaction of the local Communist Party activists was panicky. In order to ease the tension, they announced a pay rise of 300–500 zlotys and the improvement in the stock of factory shops. Despite the promises, the strike continued in the factory for an additional two days. It also provoked similar actions in other factories. On 8 August the weaving mill in the Harnam Cotton Factory and on 11 August the ‘Arelan’ Spinning Mill stopped working. The next day, textile factories in Łódź, Pabianice, Żyrardów, Aleksandrów Łódzki and other nearby towns joined the action. The basic postulates of the strikers were always a pay rise and the improvement of stock in stores. With the outbreak of strikes on the coast, tension increased in the city. From 25 August onwards, additional factories went on solidarity strikes with the workers of the Gdansk Shipyard. The signing of the agreement with the authorities by Lech Walesa eased the tension and made the strikers return to work.38 On the wave of enthusiasm, in the enterprises of Łódź, the structures of the Independent Trade Union ‘Solidarity’ were also organized. The ‘Solidarity’ Trade Union in Łódź was established on 5 September 1980, when the Inter-factory Founding Committee was created in the city. After the registration of ‘Solidarity’ by the Supreme Court on 10 November, the Independent Selfgoverning ‘Solidarity’ Trade Union of the Łódź Region began operating in Łódź. Its chairman was Andrzej Słowik. In accordance with the accepted guidelines, the ‘Solidarity’ Trade Union was based on territorial structures. In individual regions different trade sections operated. Such was also the case in Łódź, where in the ‘Solidarity’ Union the Temporary Commission of Textile Workers was set up. Its task was to fight for the amelioration of working conditions, fair wages, to carry out negotiations on working time and occupational illnesses, and to create proper care for women, who constituted the great majority of textile workers. The dramatically deteriorating living conditions at the turn of the year 1980 resulted in the ‘hunger marches’ of 1981, in which female textile workers participated in crowds. Unfortunately, with the introduction of martial law on 13 December 1981 by the Communists, ‘Solidarity’ in Łódź lost its legal status.39 K. Lesiakowski (ed.), Opozycja i opór społeczny w Łodzi 1956–1981 (Warsaw, 2003), pp. 31–35. 38 Ibid., pp. 37–40. 39 Leszek Olejnik, ‘NSZZ “Solidarność” w Łodzi – powstanie i główne kierunki działalności w latach 1980–1981’, in Lesiakowski (ed.), Opozycja i opór społeczny w Łodzi, pp. 103–105. 417

The Ashgate Companion to the History of Textile Workers The strikes of textile workers in Łódź never turned into street demonstrations. Aware of the violent events that took place in Poznan in 1956 and in Gdansk in 1970, the workers fought for their rights on the premises of the factories. The actions with the biggest impact were the demonstrations in Łódź in 1971; they exerted a great influence on the social and economic situation of Poland. The actions of 1976 or 1980, although of not so great importance, were still a part of the conflicts that occurred in the whole country at that time.

The years 1990–2000 As a result of the abolition of the Communist regime in Poland at the eve of the year 1990, the country entered a period of economic transformation, which aimed at transition from state socialism to a free market capitalist economy. The transformation process was a very difficult operation for the whole Polish economy. The structural changes had particularly acute consequences for the textile industry. The West European countries imposed high duties on Polish textiles, and the internal market was inundated with fabrics from Turkey, Thailand and India. Companies systematically limited their production area, taking loans that they could not pay back and, consequently, often declared bankruptcy. The first severe crisis arose at the very beginning of the transformation in 1989 and continued for three years (see Figure 15.6). The production volume of cotton yarn in 1991 amounted to 37 per cent of the production volume of 1989. In the case of woollen yarn it was 48 per cent, and for flax yarn 35 per cent. This situation was brought about by numerous factors, of which the most important was the loss of the market in the former USSR, the main consumer of textile products from Poland. Simultaneously, as a result of the deterioration of the income structure of the population, domestic demand decreased. Additionally, the factories themselves, with an outdated stock of machines, were unable to adapt to the conditions of a market economy. In the 1990s the textile industry was privatized, and as a result of this process, in 2001 only 0.85 per cent of all textile enterprises were still state-owned.40 In 1992 the Polish textile industry started to climb out of the recession. However, it has never reached even half of the 1989 level of production, in the case of both cotton yarn and woollen yarn (except for cotton yarn production of the year 1994, when 53 per cent of the production volume of 1989 was reached). Even worse, in 1998, a sudden breakdown of production took place. Many enterprises went bankrupt, and their workers were made redundant. In 1999 the volume of production of cotton yarn amounted to a mere 34 per cent of the level of production of 1989, and that of woollen yarn 24 per cent. These figures were equal to 82 per cent of the production volume of cotton yarn and 67 per cent of woollen yarn of 1946. The production of flax lost practically any importance. In 2000 the production volume of cotton yarn accounted for merely one-quarter of the level of production of the previous year 40 Rocznik statystyczny przemysłu (Warsaw, 2001), p. 4. 418

Poland and 9 per cent of the level of production of 1989. It amounted to only 37 per cent of the production volume of the year 1946. These unfortunate trends affected the production of fabrics to an even greater degree (see Figure 15.7). 









             FRWWRQ

ZRRO

IOD[

Figure 15.6 Production of yarn in Poland, 1989–2000. Source: Rocznik statystyczny, (Warszawa 1993–2004).

                   

FRWWRQIDEULFV

ZRROHQIDEULFV

Figure 15.7 Production of fabrics in Poland, 1989–1999. Source: Rocznik statystyczny, (Warszawa 1993–2004). 419

The Ashgate Companion to the History of Textile Workers The decrease in the level of textile production was directly mirrored in the fall of the employment rate (see Figure 15.8). At the end of 1996 and the beginning of 1997 the level of employment in the textile industry amounted to a half of the number of workers employed in 1989, and in 2000 it went down to 31 per cent. It was approximately 50 per cent of the level of employment in 1946.                    

Figure 15.8 Employment in the Polish textile industry, 1989–2000. Source: Rocznik statystyczny (Warszawa, 1993–2004). In the period of the transformation the percentage of people employed in the textile industry against total employment in industry was decreasing at a comparatively fast rate. One of the major reasons for this decline was, firstly, the decrease in the volume of production in the Polish textile industry, and, secondly, low average wages in comparison to other branches of industry. It equalled approximately 70 per cent of the average remuneration in all branches of industry. This situation resulted not only in an outflow of trained staff from the textile industry, but also in a lack of experts’ interest in this branch of production.41 Consequently, the number of unemployed increased. In the late 1990s unemployment in Łodz reached 100,000, which amounted to 20 per cent of the working population. Soon it turned out that the city had the highest death rate in the country, including the highest infant mortality rate and the highest rate of deaths due to cancer and cardiovascular diseases.42 41 Agnieszka Bukowska-Piestrzyńska, ‘Kondycja przemysłu włókienniczego w Polsce w latach 1995–2000’, http://efektywnosc04.ae.wroc.pl/Referaty/art75.pdf, pp. 5, 9–10. 42 Wojciech Górecki, Łódź przeżyła katharsis (Łódź, 1998), p. 24. 420

16 The cotton textile industry in Russia and the Soviet Union Dave Pretty

The Soviet regime saw itself as the proletarian state and most often imagined those proletarians as metalworkers who were skilled, literate and urbanized. However, for most of modern Russian history the most significant sector of the working class both in numbers and politically were textile workers, especially cotton textile workers. Textiles were the most important component of Russian industry until the First Five-Year Plan and cotton workers were significant and at times decisive participants in the revolutions of the early twentieth century. Cotton textiles were the driving force of Russia’s industrialization, and the industry’s history in many ways encapsulates that of the Russian economy during the last two centuries: during the imperial period, tremendous growth through the adoption of Western technology without ever quite catching up with the West; autarkic stagnation under the planned economy; and catastrophic decline once it was exposed to the pressures of a global economy after 1991.

From its beginnings to 1842 As they were throughout the world, textiles had been a vital craft during the Kievan and Muscovite periods, but cotton came late into Russia; wool and especially linen  The author would like to thank Jeffrey Rossman, Alice Pate, Alexander Martin, Katya Vladimirova, Boris Gorshkov, Michael Melancon, Christopher J. Ward, Elise van Nederveen Meerkerk, Lex Heerma van Voss and Els Hiemstra-Kuperus for advice and help on this project, as well as Green Day for American Idiot, which kept me focused and motivated through both writing and revision. I have concentrated on cotton because of the overwhelming importance it assumed in Russia’s industrialization process. It should be noted that a thriving cotton textile industry existed in Russian Poland, especially around the city of Łódż; it was not well integrated, however, into the overall imperial economy. It is considered separately as part of Polish history in Chapter 15.

Map 16.1 Russia

The cotton textile industry in Russia and the Soviet Union would remain more important into the 1800s. Cotton goods first appeared, albeit sporadically, during the 1400s; imports became much more regular during the 1500s. The expense and unreliable supply of these goods led Tsar Alexei to try to institute a domestic industry after 1665; however, none of his efforts reached fruition. It was only around 1725 that domestic production began, when a Moscow linen factory owned by a Dutch merchant began to weave cloth from yarn imported from both Asia and Europe. In succeeding decades, a small cotton-weaving and spinning industry grew in Astrakhan, apparently using raw material and expertise from nearby regions of Asia. Cotton only became important in Russia towards the end of the eighteenth century, but rapidly became the engine of the nation’s industrialization. Several factors contributed to this process. Firstly, the Industrial Revolution in Great Britain greatly reduced the price and increased the availability of cotton yarn and cloth. As a result, the industry grew in Russia back from the end of the production process, with cotton-finishing and printing first establishing themselves, then weaving, followed by spinning and, finally, in the late 1800s, large-scale cotton cultivation. In many ways the cotton textile industry in Russia grew on the trellis of the linen industry. By the mid-1700s a thriving linen industry had developed throughout the area that would become known as the Central Industrial Region. As a result, when cotton became important there was a vast textile infrastructure already present, ranging from sizeable weaveries and printing establishments to myriad households spinning thread or weaving cloth on a putting-out or independent basis, backed by far-reaching commercial networks that made linen one of Russia’s major items of both domestic and international trade. The production techniques used in linen, especially in weaving and printing, were easily adapted to cotton; many of Russia’s major cotton concerns began in the 1700s as linen manufactories. A skilled workforce as well as commercial and industrial infrastructures was already in place before cotton began its ascent to textile dominance. Another factor shaping the growth of the cotton textile industry was its relative unimportance to the Russian state. The tsarist administration saw woollens and linen as vital industries, primarily for such militarily significant products as uniforms and sailcloth. The state would thus cultivate these industries and at the same time struggle to control them. Although the state lost this struggle, small-scale producers did face competition from large-scale firms with guaranteed state orders and government-granted privileges. In addition, the state preferred to lavish its favour on entrepreneurs from among the nobility and foreigners. Merchants and especially peasants found it much easier to gain entry into the cotton industry,  K.A. Pazhitnov, Ocherki istorii tekstil’noi promyshlennosti dorevoliutsionnoi Rossii: Khlopchatobumazhnaia, l’no-pen’kovaia i shelkovaia promyshlennost’ (Moscow, 1958) (hereafter Khlopchatobumazhnaia), pp. 23–27.  William L. Blackwell, The Beginnings of Russian Industrialization 1800–1860 (Princeton, 1968), pp. 42–47.  Klaus Gestwa, Die Proto-Industrialisierung in Rußland: Wirtschaft, Herrschaft und Kultur in Ivanovo und Pavlovo, 1741–1932 (Göttingen, 1999), pp. 61–65, 77–81. 423

The Ashgate Companion to the History of Textile Workers where there was little competition from protected monopolies. Furthermore, the absence of state orders for cotton products meant that demand was driven almost entirely by market forces, granting cotton a flexibility and adaptability that its textile competitors could never successfully equal. The relative ease of entry allowed many serfs to become important figures in the cotton textile industry in the years before the emancipation. Because of the poor growing conditions, most serf owners in north-central Russia had preferred to extract their feudal dues through cash payments. Many nobles thus encouraged their serfs’ business endeavours and would often use their influence at court to help these entrepreneurs. The Sheremetevs, one of whose possessions was the textile centre of Ivanovo, were among the most assiduous cultivators of peasant undertakings. Many of Ivanovo’s serfs became wealthy serf owners themselves; while factories, equipment, land and serfs all belonged de jure to the lord, the serf entrepreneurs had great de facto free rein and often built impressive fortunes, much of which would eventually be used to purchase their freedom. The modern cotton industry began in Russia in 1753 when two Britons were given permission to found a cotton-printing and finishing factory outside St Petersburg; 10 years later, a German began another printing factory nearby. Both were given permission to import foreign cloth and machinery duty free, and here many Russians would learn the secrets of dyeing and finishing that they would bring back to the central provinces; many Ivanovo linen entrepreneurs sent associates there to learn the trade. One of these learned closely guarded techniques of fixing oil-based dyes on cotton cloth and opened a printing plant on his return to Ivanovo in 1787. Within the next 10–15 years, most of the region’s linen-printing concerns – aided by the plummeting cost of imported English cotton cloth due to the Industrial Revolution – converted to cotton printing. The cotton-printing industry grew rapidly into the mid-1790s, when a short-term crisis ensued, as technological advances in Great Britain lowered the cost of British finished cloth just as a new Russian tariff instituted a fourfold increase in the import duty on unfinished cotton cloth. This rapid growth and subsequent crisis spurred the growth of a cottonweaving industry and solidified the north-central region as its Russian heartland. Many cotton-printing entrepreneurs had already begun trying to lower costs by importing foreign yarn and weaving it into cloth either at their own factories or more often putting it out to peasant households. The tariff of 1797 only increased Arcadius Kahan, The Plow, the Hammer and the Knout: An Economic History of EighteenthCentury Russia (Chicago, 1985), pp. 19–21; Mikhail I. Tugan-Baranovsky, The Russian Factory in the Nineteenth Century (1907; Homewood, Ill., 1970), p. 50.  One of the best sources for tracing the beginnings of the Ivanovo entrepreneurial dynasties and their relations with the Sheremetevs is Iakov P. Garelin, Gorod IvanovoVoznesensk ili byvshie selo Ivanovo i Voznesenskii posad (Shuia, 1884).  Pazhitnov, Khlopchatobumazhnaia, pp. 27, 28, 45, 47–50; Gestwa, Proto-Industrialisierung, pp. 81–82; P.M. Ekzempliarskii, Istoriia goroda Ivanova, Part 1, Dooktiabr’skii period (Ivanovo, 1958), pp. 49–52.



424

The cotton textile industry in Russia and the Soviet Union the economic viability of this approach by raising duties on imported cloth but not imported yarn; by 1799 the output of Russian cotton weavers already exceeded that of linen weavers. In addition, the larger accessible labour market of peasants already familiar with linen weaving gave Moscow and Ivanovo firms a greater competitive viability compared to St Petersburg. This importance was confirmed during the rapid growth the industry experienced during the first two decades of the 1800s as the dislocations of the Napoleonic Wars often cut off English competition as well as the linen industry’s primarily foreign market. Between 1799 and 1814 the production of cotton cloth and calicoes increased from 350,000 to over 11 million metres, while the production of printed linen products fell from 1.16 million metres to almost nothing. By 1814 there were 40,203 workers in weaving and printing concerns. After sputtering for almost a decade, the industry entered a period of sustained growth after protective tariffs were instituted in 1822; by 1854, there were over 126,000 cotton textile factory workers. Not all of these workers, however, worked in factories, as the industry developed in very different ways across the three major branches. While in the first few decades of the century most cotton printing was done by hand, other finishing processes took advantage of technological advances, and printing became a centralized factory industry. Although mechanized printing began to take hold after 1817, as late as 1849 61 per cent of calico in Russia’s cotton-printing capital of Ivanovo was printed by hand.10 Spinning was similarly centralized, but remained the least developed branch of Russia’s cotton industry. The first spinning factory was built in St Petersburg in 1798, but Russian yarn could not match the quality or price of British yarn. In 1831 only 19 per cent of the yarn used in Russian weaving was produced in Russia, and of this only 6 per cent was produced in factories.11 The reason Russian spinning was not competitive and that hand-block printing remained so important in finishing was the British ban on the export of textile machinery. British machinery was easily the best but could only be copied inexpertly in Russia. The branch of cotton textiles where this had the greatest effect was in weaving. While even primitive cylindrical printing machines could replace the labour of several workers, the best mechanized looms available even after the lifting of the British export ban were only twice as productive as handlooms.12 For this reason weaving long remained a decentralized activity, mostly performed in peasant homes throughout the Central Industrial Region by independent artisanentrepreneurs, on commission from either printing factories or on commission  Gestwa, Proto-Industrialisierung, pp. 83–84; Pazhitnov, Khlopchatobumazhnaia, p. 29, 31.  Pazhitnov, Khlopchatobumazhnaia, p. 38; Blackwell, Beginnings of Russian Industrialization, p. 424. 10 Pazhitnov, Khlopchatobumazhnaia, p. 51; Gestwa, Proto-Industrialisierung, p. 100. 11 Most cotton spun in Russia was spun in peasant huts; see Istoriia otechestvennoi tekstil’noi promyshlennosti (Moscow, 1992) (hereafter IOTP), pp. 121–123. 12 Pazhitnov, Khlopchatobumazhnaia, p. 41; Gestwa, in making the same point, gives a ratio of six to one in favour of mechanical looms (Proto-Industrialisierung, p. 108). 425

The Ashgate Companion to the History of Textile Workers independent middlemen who distributed yarn and then sold the finished cloth to printing factories. Peasant productivity also increased during the early 1800s, as handlooms became more available and more efficient. Peasant labour was relatively inexpensive, based on family labour with low overheads and generally integrated into an agricultural economy. With a flexible network of knowledgeable peasant weavers, cotton-printing firms had little incentive to expend capital on centralized weaving establishments; in general, only high-end grades of cloth were factoryproduced. In Vladimir province in the early 1850s 18,000 factory looms merely supplemented 80,000 peasant looms working on factory orders.13 More than in any other Russian industry at the time, cotton textile workers in the early nineteenth century were hired, not possessed by the factory owner or permanently assigned by serf owners. In 1804, 85 per cent of weaving and printing factory workers were hired; by 1825, this had grown to 95 per cent.14 Questions remain, however, as to the degree that these hired workers were freely hired, or remained free after they were hired.15 Furthermore, as the putting-out system grew through the early 1800s, many independent domestic weavers found themselves increasingly tied to particular factories or particular putting-out middlemen by loans or advances to buy yarn or more advanced looms.16 It is not possible to calculate, however, the weight of such coerced labour in the workforce. Although factory conditions were miserable, there were very few work actions through the first part of the nineteenth century. One reason may be that wages were higher than they would be later in the century; the massive growth of the industry led to great profits and a premium on retaining skilled labour. Popular dissatisfaction seems to have been expressed primarily through the use of typical peasant ‘weapons of the weak’, such as arson.17 Before the emancipation of 1861, the ‘public transcript’ of serfdom seems to have held direct conflict between capital and labour, at least in the cotton textile industry, to a minimum. In general, despite the tremendous structural problems facing the Russian economy, including horrific transport and a poor internal market, cotton textiles were able to thrive by producing an everyday necessity at cheaper prices than had ever been available. These costs would be driven even lower when the United Kingdom lifted its ban on the export of textile machines in 1842. 13 Pazhitnov, Khlopchatobumazhnaia, p. 35; Olga Crisp, ‘Labour and Industrialization in Russia’, in The Cambridge Economic History of Europe, vol. 7, pt 2 (Cambridge, 1978), pp. 308–415, at pp. 337–339. 14 Pazhitnov, Khlopchatobumazhnaia, p. 64. 15 See the discussions of various forms of coerced labor in Pazhitnov, Khlopchatobumazhnaia, p. 66; B.N. Vasil’ev, ‘Formirovanie promyshlennogo proletariata Ivanovskoi oblasti,’ Voprosy istorii 6 (1952): 99–117, at pp. 106–7; Ekzempliarskii, Istoriia goroda Ivanova, pp. 83–84; Blackwell, Beginnings of Russian Industrialization, pp. 218–19; Reginald E. Zelnik, Labor and Society in Tsarist Russia: The Factory Workers of St. Petersburg 1855–1870 (Stanford, 1971), p. 36. 16 Gestwa, Proto-Industrialisierung, pp. 103–104. 17 Tugan-Baranovsky, Russian Factory, p. 76; Garelin, Gorod Ivanovo-Voznesensk, 1, pp. 158– 159. 426

The cotton textile industry in Russia and the Soviet Union

From 1842 to 1914 The ability to import the latest state-of-the-art machinery from Great Britain after 1842 had a tremendous effect on the Russian cotton textile industry. Possibly the most affected branch was spinning, which was aided by an 1841 increase in the tariff on imported yarn. Spinning soon became the most concentrated branch of cotton textiles; in 1843 39 factories worked on 138,000 spindles, but by 1859 60 factories operated 1.5 million spindles. Spinning was the only branch that would be important in the imperial capital, as foreign capital built very large, modern factories. While in 1859 there were twice as many factories in Moscow and Vladimir provinces, St Petersburg province still had more spindles. By the time of the emancipation of the serfs in 1861, Russia possessed the sixth-largest cottonspinning industry in the world.18 The explosive growth of cotton spinning in Russia can be traced through its foreign trade. In 1821–1825, Russia on average imported 3,865 tonnes of cotton yarn annually, and only 1,147 tonnes of raw cotton. By the late 1830s, cotton imports had grown to 5,242 tonnes, but yarn had reached 9,648. Twenty years later, this imbalance had been radically reversed: 3,554 tonnes of imported yarn were dwarfed by 41,000 tonnes of raw cotton.19 In 1835, Russia produced less than 10 per cent of the yarn it used; by 1860, it produced 93 per cent.20 Cotton printing and finishing had been the most highly centralized and mechanized sectors of the industry before 1842; in that year, the introduction of the Perrotine machine, which replaced the work of 30–50 printers with two tenders, greatly accelerated this process. While there are no reliable early nineteenth-century time series for national cotton-printing production, the effect of the Perrotine and other printing innovations can be gauged from statistics for the empire’s premier cotton-printing centre, Ivanovo. The first Perrotine appeared there in 1848, and by 1850 52 medium-to-large firms produced over 608,000 units worth almost 4 million roubles; only nine years later, 45 firms produced 846,000 units worth almost 6 million roubles.21 In 1849, 61 per cent of Ivanovo’s calico was printed by hand; that had fallen to 34 per cent by 1860, and rapidly diminished thereafter.22 The tremendous growth of both spinning and printing was accompanied by a concomitant increase in weaving. However, the lifting of the British export ban did not have the same effect on weaving as on the other two branches, because, as discussed above, the best mechanized looms were not so much more productive than handlooms as to undercut the latter’s other economic advantages. The increased availability of cheap domestic yarn and demand for cotton cloth by printing firms, however, did lead to a tremendous increase in the scope of putting-out, both in 18 Pazhitnov, Khlopchatobumazhnaia, pp. 16–21; Blackwell, Beginnings of Russian Industrialization, p. 408. 19 Pazhitnov, Khlopchatobumazhnaia, p. 17. 20 Gestwa, Proto-Industrialisierung, p. 97. 21 Pazhitnov, Khlopchatobumazhnaia, pp. 94, 99, 106. 22 Gestwa, Proto-Industrialisierung, p. 100. 427

The Ashgate Companion to the History of Textile Workers the quantity and the geographical spread of homeworkers. In 1828, 6,300 weavers worked in factories in the greater Ivanovo region, while 18,224 (74 per cent) wove outside factories. In 1849, the number of factory weavers had doubled to 14,854; the number of non-factory weavers had tripled, however, to 56,980 (79 per cent). By 1869, 72 per cent of weavers still worked outside factory walls, a larger proportion than in 1814.23 In addition, non-factory production itself became better organized, more capital-intensive and more efficient in extracting labour.24 Mechanized weaving only slowly took hold in the 1860s and 1870s; in 1866, probably only 36.6 per cent of cotton cloth was produced on mechanical looms. At about that time, the Russian cotton industry began to come out of the deep crisis occasioned by the American Civil War, accompanied by a wave of mechanization. By 1879, 58.4 per cent of Russian cotton was produced on mechanical looms.25 One reason for slow growth was a stagnant level of technology had not changed: in Moscow province in 1871, the average mechanical weaving loom still only produced less than three times as much as handlooms, at a time when transportation costs and tariffs conspired to make automatic looms six times as expensive as handlooms. The real victory of the mechanized loom took place in the 1880s and 1890s; between 1860 and 1896 in the Ivanovo region the share of mechanical looms rose from 10 per cent of workers and 20 per cent of cloth production to 96 and 98 per cent respectively.26 The cotton textile industry continued to grow through the second half of the nineteenth century. Between 1860 and 1910 the production of yarn in Russia grew from under 34,000 tonnes to 267,000 tonnes. The number of workers in the cotton industry more than tripled during these 50 years from approximately 150,000 to well over 497,000.27 By 1900, Russia had the world’s fourth-largest cotton textile industry, a position it maintained until the outbreak of the First World War.28 One of the major reasons for this sustained growth was the increase in the empire’s population, which more than doubled between the emancipation and the First World War. This expanding market was more easily and cheaply penetrated because of the rapid expansion of Russia’s railroad network; it is probably not coincidental 23 Vasil’ev, ‘Formirovanie’, p. 104. 24 See the discussions in Gestwa, Proto-Industrialisierung, pp. 105–107, and Vasil’ev, ‘Formirovanie’, pp. 105–106. 25 Pazhitnov, Khlopchatobumazhnaia, pp. 42, 81. 26 Pazhitnov, Khlopchatobumazhnaia, p. 82; Gestwa, Proto-Industrialisierung, pp. 108, 158–160. Hand-weavers did not disappear altogether, however; see Crisp, ‘Labour and Industrialization’, p. 352 and Pazhitnov, Khlopchatobumazhnaia, pp. 92–93. 27 IOTP, pp. 158, 184, 186. It should be noted that the figure in Table 16.1 for yarn production seems to indicate a giant leap from 1910. This is an inconsistency in the source, and points out the many problems with economic statistics from the imperial period; there simply are no reliable or directly comparable time series. For a discussion of many of the problems involved, see A.V. Pogozhev, Uchet chislennosti i sostava rabochikh v Rossii: materialy po statistike truda (St Petersburg, 1906). 28 Susan M. Vorderer, ‘Urbanization and Industrialization in Late Imperial Russia: Ivanovo-Voznesensk, 1880–1914’ (Ph.D. diss., Boston College, 1990), p. 281. 428

The cotton textile industry in Russia and the Soviet Union that the boom periods in the cotton industry match the boom periods in railroad construction.29 Although the mass of the population was by no means well off, there was an increase in consumer spending; between the 1880s and 1914, while the population grew at approximately 1.6 per cent a year, personal consumption grew at 3.1 per cent annually.30 The rapid mechanization of the industry between 1880 and 1900, as mechanical looms replaced handlooms and ring frames replaced mules, led to falling prices for both yarn and finished calico, which meant that this growing and more easily reached market could afford to buy more cotton goods.31 These consumers were more likely to buy domestically produced Russian cotton textiles because of continued high tariffs that added more to the value of imported manufactured goods than to the value of imported raw materials.32 Another growing segment of the market for Russian cotton textiles was Asia. Already in 1860, approximately one-tenth of the industry’s output was bound for export, almost all of it to Asia, and by 1913 the weight of cotton goods being exported was five times more than that being imported.33 Major reasons for this were the construction of railroads to the border regions, especially with China and Persia, and government policy aimed at furthering Russian interests in these regions. By 1914, China imported one-third of all its cotton goods from Russia, while cotton textiles were the major component in the 50 per cent share of imports Russia enjoyed in Persia.34 The growing importance of trade with Asia, exacerbated by the crisis in raw cotton supplies during the American Civil War, were major factors leading to the Russian conquest of Turkestan in the 1860s, although production there only really began to increase in the late 1880s after the region was brought into the nation’s railroad network. In 1884, Russian Central Asia produced 164 tonnes of American cotton on 327 hectares; by 1900, Russia grew 82,000 tonnes on 272,000 hectares, almost all of it in Central Asia. By 1909, Russia produced more cotton than it imported.35 There was a sharp concentration of the cotton textile industry during the last decades of the nineteenth century, as can be seen from Table 16.1. In the more quickly modernized printing factories, the number of firms had decreased markedly already in the 1870s, while the effect of mechanized looms on putting-out firms and svetelki had an obvious effect in the 1880s. For both weaving and printing firms, however, the 1890s saw a simultaneous growth in the number of employees and a halving of the number of enterprises. Of the 588 Russian factories that employed more than six workers in 1902, the 98 with over one thousand workers had a collective labour

29 Peter Gatrell, The Tsarist Economy 1850–1917 (London, 1986), pp. 150–151. 30 Ibid., pp. 44–45, 50. 31 Ibid., p. 160. 32 Ibid., pp. 165–167. 33 Pazhitnov, Khlopchatobumazhnaia, pp. 96–97, 130–131. 34 Gatrell, Tsarist Economy, pp. 175–176. 35 Pazhitnov, Khlopchatobumazhnaia, pp. 99–101, 145–147. 429

The Ashgate Companion to the History of Textile Workers force of 257,353; the other 490 employed only 90,118. Thus one-sixth of the firms employed three-quarters of the workers. 36 Table 16.1 Russian textile production, 1860–1912 Factories

Workers

Yarn (tonnes)

Cloth (million metres)

1860

1115

150,800

33,579

2,414

1866

623

121,100

30,631

4,615

1879

614

150,100

72,563

15,549

1890

708

245,200

109,746

25,773

1900

487

378,100

173,628

54,599

1912

652

499,800

388,206

111,825

Source: Compiled from IOTP, pp. 158–160. The figures for cloth combine unfinished and finished. The cloth figures for 1900 have been altered from the source table, taking into account the amount given for unfinished cloth on p. 184.

While many start-up firms in the 1890s attempted to take advantage of good economic times, older firms were able to maintain a dominant market position. In the cotton centre of Ivanovo-Voznesensk, only one major firm had been founded after 1850, and this was a weaving factory begun by a consortium of printing concerns to supply their factories with cloth. None of Moscow’s 22 largest cotton mills in 1900 was less than 30 years old.37 The lack of penetration by new firms was reflected in the conservative nature of the industry’s structure. While few firms remained privately owned de jure, most had adopted a form of corporation designed not to attract investors but to preserve family control. While in many nations banks put themselves on industrial boards to protect investments, Russian textile firms put themselves on bank boards to funnel investments. The economic autarky of Russian cotton textiles is illustrated by an incredibly low level of foreign investment.38

36 Pogozhev, Uchet, pp. 71, 78–79, 92–93. 37 Vorderer, ‘Urbanization and Industrialization’, p. 309; Robert Eugene Johnson, Peasant and Proletarian: The Working Class of Moscow in the Late Nineteenth Century (New Brunswick, NJ, 1979), pp. 22–23. 38 IOTP, pp. 192–193; Gatrell, Tsarist Economy, pp. 181–183; Pazhitnov, Khlopchatobumazhnaia, p. 144; William B. Husband, Revolution in the Factory (New York, 1990), pp. 20–21; Alfred J. Rieber, Merchants and Entrepreneurs in Imperial Russia (Chapel Hill, NC, 1982). 430

The cotton textile industry in Russia and the Soviet Union Cotton steadily assumed the predominate role in Russian textiles. By the eve of the First World War, cotton accounted for two-thirds (by weight) of all Russian textile production; woollens and flax accounted for only 19 and 15 per cent respectively.39 Cotton alone was responsible for between one-fifth and one-fourth of all Russian industrial workers and gross value of production in 1914, easily outdistancing the combined efforts of mining, metallurgy and metalworking in production value.40 Despite this growth, however, Russia remained well behind other industrial nations in per capita cotton textile production, at one-eighth the British level.41 The industry remained geographically concentrated in the central provinces of Moscow and Vladimir, as well as in and around the imperial capital. By 1913, 80 per cent of workers – and more if Russian Poland is excluded – were located in the Central Industrial Region, in which Moscow and Vladimir provinces were located.42 This geographical concentration of cotton textiles was matched by the degree to which the industry – and Russian industry in general – was dispersed throughout the countryside. At the turn of the last century, 61 per cent of all Russia’s factories and 59 per cent of its factory workers were located outside of cities.43 These rural mills were often among the largest; of the 174 textile factories that employed more than one thousand workers in 1902, 60 per cent – accounting for 63 per cent of the workers – were located outside of cities.44 Even in factories located in the big cities, workers were mostly drawn from the peasant population. It is very difficult to measure exactly how many workers were natives of the village milieu, since under the Russian system of estates one could be a fifth-generation city dweller and still be legally ascribed to a village that one had never seen. In samples from Ivanovo-Voznesensk factories in the period 1887– 1895, fully 80 per cent of employees were juridical peasants; in a sample from the immediate pre-war years, this proportion had grown to over 90 per cent.45 Peasant migrants accounted for 93 per cent of all factory workforces in Moscow in 1902, and most of them were in textiles.46 The catchment areas for all three major textile centres were focused on the Central Industrial Region. Even in distant St Petersburg, over 41 per cent of peasant migrants came from that region, a distance of hundreds of miles.47 Moscow and 39 Pazhitnov, Khlopchatobumazhnaia, p. 127. 40 Crisp, ‘Labour and Industrialization’, p. 354. 41 Pazhitnov, Khlopchatobumazhnaia, p. 131. 42 Ibid., p. 132. 43 Joseph Bradley, Muzhik and Muscovite: Urbanization in Late Imperial Russia (Berkeley, 1985), p. 16. 44 Pogozhev, Uchet, p. 59. 45 B.N. Vasil’ev, ‘K kharakteristike formirovaniia promyshlennogo proletariata v Rossii: (Po materialam Vladimirskoi, Kostromskoi i Iaroslavskoi gubernii)’, Shakhtinskii pedagogicheskii institut. Uchenye zapiski 2 (2) (1957): 217–33. 46 Johnson, Peasant and Proletarian, p. 31. 47 Robert B. McKean, St. Petersburg between the Revolutions: Workers and Revolutionaries June 1907–February 1917 (New Haven and London, 1990), p. 18; the bulk of these migrants were from the Central Industrial Region’s north-westernmost – that is, closest to 431

The Ashgate Companion to the History of Textile Workers Ivanovo’s catchment areas were not as extensive, but were steadily growing. The continuing growth of Russian industry in general, and textiles in particular, drew more and more peasants from much farther afield into factories.48 It has been a major point of contention among historians of Russia as to whether a proletariat – a class of industrial workers conscious of themselves as a class, capable and willing to work in their own class interest – had actually formed by 1917. The iconic image of the Russian proletariat was the St Petersburg metalworker. However, the most important component of Russian factory labour had from the beginning been – and would remain well into the Soviet era – the textile worker. To what extent did textile workers fit the mould of the proletarian? The most important gauges for this in the historiography have been the degree of hereditariness and of ties to the land. The assumption has been that the longer the family heritage of factory work and the more attenuated the workers’ bonds to their village parcel, the more they were becoming workers and the further they were leaving peasanthood behind. As far as they can be reliably measured, however, these indices give a complicated representation of Russian textile workers. In the 1890s, a survey of the giant Tsindel’ cotton-printing factory in Moscow found that 55 per cent of male workers were hereditary workers; however, 90 per cent of Tsindel’ workers still retained an allotment of land in the village.49 Evidence from the Ivanovo region seems to indicate that the most successful families were those that combined agriculture and factory work. The local population was highly acclimatized to the factory, but in doing so did not give up their ties to the village. Indeed, even the majority of those who did not sow any land stubbornly held on to village allotments, although that was not in their short-term economic interest (they had to pay more taxes); the village was a social welfare safety net, where sustenance could be found during industrial downturns.50 The close interweaving of the loom and the plough can also be seen in the age structure of Russian industrial cities and the industrial region’s countryside. In Ivanovo’s Shuia county, while male–female ratios are roughly the same between city and countryside for those below 10 years of age and those older than 50, they are radically different for all groups in between. Among those in their thirties, there were approximately 67 men for every 100 women in the countryside, but 124 men for every 100 women in Ivanovo. Over two-thirds of rural males were 19 or younger or 50 and older; 55 per cent of men in Ivanovo were between 20 and 49. The figures are almost perfectly complementary.51 The implication of this data is that many St Petersburg – provinces of Tver and Yaroslavl; this was still a distance, however, at the nearest approach of almost 300 km. 48 Johnson, Peasant and Proletarian, p. 32; Vasil’ev, ‘K kharakteristike’, pp. 216–233. 49 Johnson, Peasant and Proletarian, pp. 62–63. 50 See Pretty, ‘Neither Peasant nor Proletarian: The Workers of the Ivanovo-Voznesensk Region, 1885–1905’ (Ph.D. diss., Brown University, 1997), pp. 93–108, for a fuller discussion of these issues. 51 Pervaia vseobshchaia perepis’ naseleniia Rossiiskoi imperii 1897 goda, vol. 6, Vladimirskaia guberniia, part 2 (St Petersburg, 1904), pp. 8–9. 432

The cotton textile industry in Russia and the Soviet Union peasant workers – perhaps especially textile workers – were hereditary workers, coming from families with industrial work experience, but still very much part of the village milieu. They were born in the village, worked in the factory for a decade or two, and then returned to the village.52 Outside of the smaller rural factories, where most workers could just walk to work, textile workers generally lived either in rented quarters, usually just a corner of a room, or in factory-provided housing, whether large barracks or apartments for families. Private housing was most common in St Petersburg and Ivanovo while through most of the Central Industrial Region barracks were the norm well into the twentieth century.53 The most important change in the profile of textile workers over the course of the late imperial era was the growing importance of women. From its beginnings, factory work had generally been a male activity. During the early stages of industrialization, 80 per cent of Ivanovo’s male population had been engaged in cotton printing, and printing would remain a male-dominated profession. The first generations of factory weavers were male, while domestic weaving was seen as a largely female occupation.54 However, the introduction of power-looms in the second half of the nineteenth century changed this radically. By 1883, the only weaveries in Vladimir province that still employed more men than women were those using older selfacting looms; 63.5 per cent of power-loom weavers were women, and labour in weaving divisions as a whole was only 43 per cent male.55 In spinning, while selfacting mule spinners remained overwhelmingly male, the new ring frames were operated almost solely by women; after the turn of the century ring frames were virtually the only spinning machines being installed in Russian factories, and by 1913 only one-third of Russian cotton was being spun on self-actors.56 While power-looms and ring frames were easier to use and required less strength, there was no intrinsic reason that this would make them more suitable for female labour; some of the most demanding jobs in textiles both physically (wool cleaning) and in terms of skill (warping) were traditionally woman’s work. It was a given in Russian culture, however, that women should always receive less pay, 52 See Johnson, Peasant and Proletarian, and the discussion in Crisp, ‘Labour and Industrialization’, pp. 364ff. On the important role that chain migration played in the transition between field and factory, see Johnson, Peasant and Proletarian, p. 69; Reginald E. Zelnik (trans. and ed.), A Radical Worker in Tsarist Russia: The Autobiography of Semën Ivanovich Kanatchikov (Stanford, 1986), pp. 7–13, 398, n. 1; Vorderer, ‘Urbanization and Industrialization’, pp. 411–412; Pretty, ‘Neither Peasant nor Proletarian’, pp. 91–92. 53 Johnson, Peasant and Proletarian, pp. 84–85; William J. Chase, Workers, Society, and the Soviet State: Labor and Life in Moscow, 1918–1929 (Urbana, Ill., 1987), pp. 191–192; Vorderer, ‘Urbanization and Industrialization’, pp. 388ff., 406. 54 Gestwa, Proto-Industrialisierung, pp. 101, 105, 110, 519–520, and 519, n. 309. 55 P.A. Peskov, Fabrichnyi byt Vladimirskoi gubernii: Otchet za 1882–1883 g. fabrichnago inspektora nad zaniatiiami maloletnikh rabochikh Vladimirskago okruga (St Petersburg, 1884), 1, p. 9; 2, pp. 7–8. 56 Chris Ward, Russia’s Cotton Workers and the New Economic Policy: Shop-floor Culture and State Policy 1921–1929 (Cambridge, 1990), pp. 77, 80. 433

The Ashgate Companion to the History of Textile Workers and employers basically took advantage of the introduction of new technologies to switch from men to lower paid and, it was assumed, more pliant female labour. In 1884, female weavers in Moscow province earned only one-half of what a male weaver did. By the eve of the First World War, it was 90 per cent of the male wage, although this was mostly a result of men’s wages being reduced through the effects of competition.57 By 1913, over 60 per cent of Ivanovo workers were women, although printing divisions remained stubbornly (about 90 per cent) male domains.58 Child and adolescent labour played an important role in cotton textiles, accounting for possibly 30 per cent of employees in the 1870s. This was severely limited in 1885 when children under 12 were forbidden to work in factories; at the same time, minors and women were prohibited from working night shifts.59 The percentage of adolescent labour in Ivanovo-Voznesensk seems to have remained at a steady 5 per cent of the workforce between 1885 and 1914; there does not seem to be any reason to believe those figures were different in other regions of the country.60 The majority of Russian textile workers were paid by the piece. There seem to have been few regional differences in pay, although workers in the capital earned a bit more, undoubtedly due to St Petersburg’s relatively high cost of living and lack of factory barracks.61 In 1896, weavers earned 10–12 roubles a month, somewhat more than ring-frame operators, while mule spinners made almost twice as much, a difference that reflects the greater feminization of weaving and water frames.62 Both the best and worst paid positions tended to be in the printing divisions; some job descriptions only earned between 9 and 10 roubles a month, while metal workers and mechanics in textile factories earned over twice as much, and engravers and pattern designers might earn four times as much.63 There is also evidence that rural factories had to pay higher wages to keep workers from moving to bigger cities.64 Through much of the textile industry, wages were often 10 per cent higher in the summer months to entice workers off the farm during the peak farming months. In the 1850s, workers often laboured 14 hours a day; such hours could still be found in 1896, although many worked for as few as nine. This changed in 1897, when the work day was legally cut to 11½ hours.65 The many fines imposed by management were so unpopular and caused so much unrest that the state eventually forced owners to funnel fine revenue into funds from which those injured in the factories would receive compensation. The government set up a 57 See Rose L. Glickman, Russian Factory Women: Workplace and Society, 1880–1914 (Berkeley and Los Angeles, 1984), pp. 105–114 (wage differentials) and pp. 215–218 (female passivity). 58 Gestwa, Proto-Industrialisierung, p. 519. 59 Ibid., pp. 521–522. 60 Vorderer, ‘Urbanization and Industrialization’, p. 502; Pretty, ‘Neither Peasant nor Proletarian’, p. 83. 61 Pazhitnov, Khlopchatobumazhnaia, pp. 104–105. 62 Ibid., p. 104. 63 Pretty, ‘Neither Peasant nor Proletarian’, pp. 82–83. 64 Vorderer, ‘Urbanization and Industrialization’, p. 555. 65 Pazhitnov, Khlopchatobumazhnaia, pp. 69, 103–104. 434

The cotton textile industry in Russia and the Soviet Union factory inspectorate in the 1880s that was progressively given more responsibilities to mediate between labour and capital. Inspectors were hemmed in by restrictive laws that if not explicitly favouring ownership certainly distrusted workers, and were never really trusted by either workers or owners.66 The government itself was ambivalent in its attitude towards the working class. Since at least Peter I, the regime had been torn between the need to modernize to match the dynamism of its western neighbours and the fear that changes would either occur too fast for the state to control or be too fundamental for the regime to survive. Its approach to industrialization was thus ambivalent: it desired the increased state power and military might that the process had brought to Western Europe, while it feared the social and political conflicts that had accompanied it there. Thus the finance ministry generally backed industrialists, but the interior ministry found itself trying to stifle worker unrest while arguing that abuses by ownership had to be curbed. Although strikes and unions would remain illegal through 1905, this did not, however, mean that strikes and other forms of worker unrest did not occur. What might be called the beginning of a continuous strike movement in Russia began in 1870, during a brief economic upturn.67 While the frequency and size of strikes would increase steadily to the crescendos of 1905 and 1914, it was a staccato, not continuous growth. Strikes generally occurred during economic boom years, when workers had more leverage; each successive upturn through the end of the century saw more and more labour activism. And through the end of the century, it was in cotton textiles that worker unrest was concentrated.68 The first major strike occurred in May 1870 at the largest factory in St Petersburg, the Nevsky spinning mill. Although only 800 workers were involved, it punctured many hopes that Russia could avoid the labour unrest of the West; one newspaper declared: ‘And a strike has befallen us, and God has not spared us!’ It set many patterns for future strikes; after extensive press coverage and light judicial sentences for the strike leaders, the government resorted to administrative measures to punish four ringleaders and tighten press censorship of labour issues. In the future, very few strike cases would ever be dealt with in the regular judicial system; the most typical punishment for those deemed guilty of ‘instigating’ strikes would be internal exile to the far north or Siberia.69 The Nevsky strike was followed by a major strike in 1872 at the Kreenholm cotton mill – at that point Europe’s largest

66 Many of these issues are dealt with in S. Gvozdev, Zapiski fabrichnogo inspektora: Iz nabliudenii i praktiki v period 1894–1908 gg., 2nd edn (Moscow and Leningrad, 1925). 67 For earlier strikes, see Pazhitnov, Khlopchatobumazhnaia, pp.70–73; Ekzempliarskii, Istoriia goroda Ivanova, pp. 111–112; Zelnik, Labor and Society, pp. 163–169. 68 On the coincidence of strike activity and economic booms, see Johnson, Peasant and Proletarian, pp. 127–131, and Pretty, ‘Neither Peasant nor Proletarian’, pp. 122–123. For the concentration of labour unrest in textiles for Moscow province, see Johnson, Peasant and Proletarian, pp. 144–145, where it is shown that between 1880 and 1900, well over half of the strikers in that province were cotton workers, although they accounted for less than one-third of the industrial workforce. 69 Reginald E. Zelnik, Labor and Society, pp. 340–367; the newspaper citation is on p. 341. 435

The Ashgate Companion to the History of Textile Workers – 85 miles west of St Petersburg, involving 5,000 workers.70 Many of the complaints that would be important throughout the pre-Revolutionary period were present at Nevsky and Kreenholm: irregular pay practices, manipulation of piece-rates, unfair fines and abusive supervisory personnel. The international depression that began in 1873 brought an end to this brief flurry of labour activism, which picked up again with boom years accompanied by increasing worker unrest that lasted from 1879 to 1880, then 1885 to 1889, and the sustained economic growth from 1893 to 1899. Each of these periods saw an escalation in the number of strikes and strikers. According to one standard account, between 1870 and 1894 there were 85 strikes in the cotton textile industry in which 53,341 workers took part. In the next six years (1895–1900), there would be 188 strikes with 139,154 participants, a tenfold increase in the number of strikers per year.71 These years marked the high point of the ‘Witte industrialization’, when Russian industrial expansion reached unprecedented rates, sometimes achieving double-figure annual growth rates. There had never been such a demand for labour, and workers had never had such leverage. Throughout the imperial period, the lead in cotton strikes was generally taken by weavers. Printing plants were divided into so many separate workshops and had such wide disparities in income that very little commonality of purpose could ever be established. Even in general strikes, printing divisions tended to stop work last, once supplies of cloth began to run short. While spinners had the most strategic position in the production process, since a stoppage there shut down all other divisions, spinning plants were by the 1890s split into the very different universes of patriarchal mule spinners and underpaid, generally female ring-frame operators. Weavers, on the other hand, were the single largest job description in cotton textiles, often accounting for a full quarter of the workforce even in plants that incorporated every stage of the production process. They were gathered in the largest work spaces in the factories, and, after taking account of the gender divide, had very homogeneous work experiences, incomes and backgrounds. While spinners could most easily strangle the production process, weavers were in many ways the profession with the critical mass necessary to organize and lead large-scale work stoppages. From very early in the industrialization process, weavers assumed that they had both a right and a responsibility to lead; indeed, textile workers were generically referred to as tkachi, or weavers.72 Most strikes had a spontaneous character. Given the illegality of strikes and labour organization before 1905, even if they were not spontaneous, it was best if they appeared that way. From very early on, there were usually no obvious leaders, and owners and government officials asked workers to elect deputies so as not 70 Reginald E. Zelnik, Law and Disorder on the Narova River: The Kreenholm Strike of 1872 (Berkeley and Los Angeles, 1995). 71 Pazhitnov, Khlopchatobumazhnaia, pp. 116–117. 72 See the discussion in Pretty, ‘Neither Peasant nor Proletarian’, pp. 75–82, and Zelnik, Law and Disorder. 436

The cotton textile industry in Russia and the Soviet Union to be forced to negotiate with a crowd. Strikes usually began in one division and spread as strikers ‘took out’ other divisions or even other factories. Before 1917, strikes were almost always a male activity; even as women were accounting for an ever greater majority of textile workers, initiating and leading strikes remained an overwhelmingly male prerogative. The worldwide depression that began at the end of 1899 brought a sudden end to the strike movement among textile workers; the total number of cotton textile strikers for the first five years of the century was only marginally more than in 1897 alone, and the last year in this period – 1904 – would see only four cotton strikes nationwide.73 This decline in textile worker activism sharply reversed itself in 1905, when a full-scale social revolution broke out. Between 1895 and 1904 there had been 270 strikes at cotton factories, with 197,139 participants who left work for a total of 945,686 worker hours. In 1905, just in the cotton textile industry, there were 1,008 strikes, 784,058 strikers and 8,329,352 worker hours lost.74 Textile workers showed a greater propensity than metalworkers to strike in St Petersburg and during the key moments of the revolution from October through December in Moscow.75 The textile centre of Ivanovo-Voznesensk was the site of what was to that point the greatest strike in Russian history in terms of worker hours lost (well over a million) when the entire city shut down from May into July. This strike would see the formation of what later became enshrined in Soviet hagiography as the first soviet of worker deputies.76 The most violent and tragic revolutionary event, the December uprising in Moscow, was centred on one of the city’s largest textile mills.77 In general, Russian textile workers – especially cotton textile workers – demonstrated a great deal of initiative and innovation, first in the 1890s strike movement and then in the revolution of 1905. This was despite a relatively low literacy rate (39 per cent in 1897, 52 per cent in 1918, compared to 66 and 77 per cent among metalworkers),78 low skill level and persistent ties to the countryside, all qualities which were classically presumed to lessen worker activism and class consciousness. Some historians, however, believe that the peasant nature of the textile workforce may well have facilitated strikes, as workers with a safety net in the village would have been more willing to take risks in the factory. Similarly, there 73 Pazhitnov, Khlopchatobumazhnaia, p. 150. 74 Ibid. 75 Laura Engelstein, Moscow, 1905: Working-Class Organization and Political Conflict (Stanford, 1982), p. 229; Gerald D. Surh, 1905 in St. Petersburg: Labor, Society, and Revolution (Stanford, 1989), p. 68, n. 37. 76 Pretty, ‘Neither Peasant nor Proletarian’, pp. 416–592. 77 Engelstein, Moscow, 1905, pp. 202–225. 78 Gattrell, Tsarist Economy, p. 162. This was in large part because of the greater weight of female labour in textiles. Among men, the rates in 1897 were 67 for metalworkers and 54 per cent in textiles, and 81 and 76 per cent respectively in 1918; for women, 32 per cent and 12 per cent in 1897, 38 per cent in 1918 (the 1897 figures are for textile workers as a whole, the 1918 ones are specifically cotton workers). 437

The Ashgate Companion to the History of Textile Workers are signs that news of strikes and protests spread throughout the Central Industrial Region along networks of migrants from the same regions. Peasant modes of protest, which emphasized unanimity and anonymity, were well suited to an industrial setting in which strikes and unions remained illegal; even acts of violence reflected a calculation, directedness and even etiquette that belied roots in the countryside. Even the soviet itself, which spread with great speed and instant legitimacy throughout Russia, has been seen by some as an institution grounded in the sel’skii skhod, the village assembly with which all Russians with rural ties were familiar.79 Of course, by 1905 there was another variable, namely the influence of political agitators identified with various socialist parties. Dissident intellectuals first attempted to organize industrial workers in the early 1870s, but consistent efforts were only made from the mid-1890s on. Among textile workers political activists tended to be concentrated in the least likely sector to strike, the printing divisions. It was here that the most literate and skilled workers were found, both in certain printing occupations such as engraving and in the metalworkers found in virtually every factory who repaired the machines and made spare parts, and who, in vertically integrated factories, were usually housed in the printing divisions.80 The 1905 revolution left a regime that was only tenuously liberalized. Strikes and unions were legal, but instigating a strike was not, and unions laboured under restrictions that rendered them quickly moribund. The overall economy remained in the doldrums through the end of the decade, which had an inhibiting effect on labour protest. Workers, however, had made solid gains from the 1905 revolution. During the inter-Revolutionary period most cotton workers worked nine hours days, and in 1910 were making almost 10 per cent more in real terms than in 1900; these gains would be lost by 1913, however, possibly because of the growing feminization of the profession, itself possibly a conscious attempt by industrialists to obtain a less rebellious workforce. By 1914, women constituted 63 per cent of the cotton labour force.81 The hopes management placed in female complacency would be sorely disappointed by events immediately before and during the First World War. Sparked by a massacre of striking gold miners in Siberia in early 1912, and supported by a booming economy, a formidable strike movement erupted in Russia that built to a crescendo in the summer of 1914. During the first half of 1914, cotton workers struck for more worker hours in six months than in any year except 1905.82 This 79 Johnson, Peasant and Proletarian, pp. 152–154 and passim; Vladimir Andrle, A Social History of Twentieth-Century Russia (London, 1994), p. 100; Daniel R. Brower, ‘Labor Violence in Russian in the Late Nineteenth Century’, Slavic Review 41 (3) (Fall 1982): 417–431; Pretty, ‘Neither Peasant nor Proletarian’; N.V. Mikhailov, ‘The Collective Psychology of Russian Workers and Workplace: Self-Organization in the Early Twentieth Century’, in Michael Melancon and Alice K. Pate (eds), New Labor History: Worker Identity and Experience in Russia, 1840–1918 (Bloomington, Ind., 2002). 80 Pretty, ‘Neither Peasant nor Proletarian’, pp. 383–384, 418, 429–430 and passim. 81 Pazhitnov, Khlopchatobumazhnaia, p. 129; IOTP, p. 189; Crisp, ‘Labour and Industrialization’, pp. 357–358; Husband, Revolution in the Factory, pp. 16–18. 82 Pazhitnov, Khlopchatobumazhnaia, p. 150. 438

The cotton textile industry in Russia and the Soviet Union strike wave was markedly centred on St Petersburg, however, and was driven by metalworkers. While textile workers in the capital enthusiastically participated, the Central Russian cotton heartland was relatively quiescent.83

Cotton textiles in war and revolution The great strike in St Petersburg was brought to a sudden stop by the outbreak of the First World War. This war, for which Russia was unprepared and which would eventually bring down the autocracy, did not have a marked effect on the textile industry through 1915, but during 1916 and 1917 total production in the textile industry fell almost by half, and pay in constant roubles fell almost 20 per cent. The economy was hard pressed to wage both a major war and maintain a functioning economy. At the most prosaic level, there simply were not often enough railroad cars to supply the factories, feed the cities and maintain the war effort. While most cotton textile plants were able to operate through the war without having to close, many of them operated at much less than capacity. For those firms that were operating, however, the war was very profitable, as average profits increased from 4.9 per cent of capital in 1913 to 14.7 per cent in 1916.84 The strike movement had already picked up during the war. Beginning in May 1915, a wave of strikes swept Vladimir and Kostroma provinces, which were repressed with deadly force.85 More cotton workers struck in 1915 than the year before and the totals for 1916 approached the revolutionary year of 1906 in number of strikers and worker days lost.86 The wartime strike wave was dominated by cotton workers, although Petrograd’s cotton workers were less restive than others in the capital, and Moscow’s were quiescent.87 By the beginning of 1917 it was obvious that the present rulers had lost their credibility, and the only suspense was when the expected coup would take place. What no one foresaw was that a strike by women textile workers would bring down the entire political system. On 23 February 1917, several thousand female weavers struck in the renamed imperial capital, Petrograd. That day 23,000 would strike; two days later, 10 times as many had left work. The war had greatly increased women’s share of the industrial workforce; even the traditionally male bastion of metalworking was one-third female in the capital, and the percentage of men in textiles fell from 42 per cent in 1914 to 32 per cent in February 1916.88 With men at the front, women 83 McKean, St. Petersburg, pp. 198, 205, 297–317; Ekzempliarskii, Istoriia goroda Ivanova, pp. 273, 305–307. 84 Pazhitnov, Khlopchatobumazhnaia, pp. 134–136, 138, 144, 149. 85 Ekzempliarskii, Istoriia goroda Ivanova, pp. 318–324. 86 Pazhitnov, Khlopchatobumazhnaia, p. 150. 87 Diane Koenker, Moscow Workers and the 1917 Revolution (Princeton, NJ, 1981), p. 31; McKean, St. Petersburg, pp. 406–429. 88 Koenker, Moscow Workers, p. 81. Note that the Russian calendar in 1917 was 13 days 439

The Ashgate Companion to the History of Textile Workers were increasingly likely to be the sole breadwinners for a family. As the supply system deteriorated many woman workers were getting in bread lines at two or three in the morning and then going to work a 10-hour day. Troops in the city demonstrated sympathy with striking women from the beginning, and a full-scale mutiny of the garrison on 27 February sealed the regime’s fate. The fall of the monarchy led to a flurry of organization throughout Russia. From the centre to the local level, two parallel hierarchies formed: one, whose apex was the Provisional Government, composed of members of the privileged classes that wanted to establish a liberal republic, the other composed of soviets of worker deputies that espoused a more radical, socialist outcome. While the soviets had popular support from the beginning, the moderate socialist leadership of the Petrograd soviet insisted that Russia was not ready for socialism and deferred to the Provisional Government. For over six months, the moderate socialists refused to take power, and even attempted to strengthen the Provisional Government by participating in it. Instead, the moderate socialists themselves became identified with the government’s deeply unpopular measures, especially continuation of the war. The only political party willing to take a position against the war, for land distribution and for an immediate transfer of power to the soviets was the Bolsheviks, and their support grew steadily from April to October. If the Bolsheviks had not taken power violently on 25 October, they almost certainly would have headed a coalition government within a day or two. Outside the capitals, however, soviets were much more assertive; in the textile heartland, the Ivanovo-Voznesensk soviet had declared itself as possessing full power at the beginning of July. The union movement also revived. Right after the February Revolution, some Bolshevik activists took over the shell of the Central Industrial Region Textile Workers Union, which had only 64 members when the revolution occurred, with the aim of building it into a viable national organization. By the end of June, it represented textile workers as a national union at an allRussian congress of trade unions, claiming to represent to some degree 178,000 workers.89 This was against a background of continued economic decline, as the nation’s distribution and supply system stumbled towards complete collapse. Both the unions and the local soviets were partly driven by increasing shutdowns of textile factories. Throughout the Central Industrial Region, possibly 20 per cent of textile factories had closed by late October. Workers did not see this as a result of macroeconomic chaos, however, but rather as part of a management plan to retain control in a post-Revolutionary situation. This distrust not only drove increasing demands for soviet power and support for the Bolsheviks but also sparked support for – and more radical redefinitions of – workers’ control of the factories.90 From the beginning of the revolution, factory-level committees attempted to implement more radical understandings of workers’ control if only to keep factories open. At the same time, the Provisional Government’s establishment of a central textile behind the calendar used in the West. 89 Husband, Revolution in the Factory, pp. 19, 24, 37. 90 Ibid., pp. 27–28, 35; Koenker, Moscow Workers, pp. 204, 215, 258, 311. 440

The cotton textile industry in Russia and the Soviet Union cartel was seen by workers as a gift to owners to manipulate prices and wages with greater impunity. Local union leaders in the Ivanovo region began a general strike in October that eventually embraced 300,000 workers throughout central Russia. It did not come to an end until mid-November, when the Bolshevik leadership itself negotiated an end with owners. The workers themselves were not satisfied, and continued to fight for weeks.91 This began a process that would last through the civil war, as the Bolshevik (soon renamed Communist) regime attempted to establish its authority, creating central economic institutions with supposedly more and more authority even as local authority continued to institute its own visions of soviet power. In many ways workers’ control may well have been embraced so completely by textile workers because it so closely reflected the peasant agenda for the revolution: the establishment of an economic autarky free of outside political manipulation. It would take the central government, after 1918 in Moscow, several years to establish control over local textile unions, textile enterprises and textile-region soviets, in large part because of the chaos of the civil war. From 1918 through the end of 1920, the new regime found itself in a vicious struggle to survive, and with its coercive capabilities directed against external enemies, the central government had only sporadic control over local institutions. This period saw the complete disintegration of the Russian economy. Market exchange with the countryside completely broke down, and the regime depended on armed detachments to gather enough grain to feed the cities and army. Market exchange in the cities also ceased; as currency became meaningless, a barter economy took over, and the big cities lost up to half their population. By 1919, industrial production had fallen to 22 per cent of its pre-war levels. Textile production was even harder hit; by May 1920, only eight of Ivanovo-Voznesensk’s 130 factories were open, and by June 1921, only 1 per cent of Russia’s spindles were working. The food situation was so dire that by 1920 over 18,000 hectares of factory land had been converted to food production.92

Cotton textiles in the Soviet Union The Soviet regime had won a military victory by the end of 1920. However, once peace had been achieved, a population that had been able to rationalize privations because of the counter-revolutionary threat were no longer willing to put up with food shortages and grain seizures. By February and March 1921, faced with a combination of peasant rebellions, industrial strikes and military mutinies reminiscent of February 1917, the Communist regime reacted with much greater 91 Husband, Revolution in the Factory, pp. 30, 33–44. See also William B. Husband, ‘Local Industry in Upheaval: The Ivanovo-Kineshma Textile Strike of 1917’, Slavic Review 47 (Fall 1988): 448–463. 92 Husband, Revolution in the Factory, pp. 93, 146; Ward, Cotton Workers, p. 11. 441

The Ashgate Companion to the History of Textile Workers dexterity than its predecessors. At the Tenth Party Congress, Lenin announced the New Economic Policy (NEP): while the state would hold on to the ‘commanding heights’ of industry and commerce, small-scale commercial and artisanal operations would be tolerated, as would a free market in grain. The war on the countryside to feed the cities, which had been going on to some degree since 1916, would cease, and a bond (smychka) would be forged between the peasantry and the workers. In this plan, textile workers were to play a central role. Peasants were expected to grow more grain if there were consumer goods to buy; as one state publication stated, ‘the products of cotton mills are the basic means of exchange between town and countryside’.93 As a result, cotton textiles would recover from their nadir with stunning speed. By the 1926/27 fiscal year, 90 per cent of all cotton equipment was back on line; in 1927/28, production exceeded 1913 levels. Even in the metalworking bastion of Leningrad (née St Petersburg), cotton textiles were the largest industry by 1928; employment in textiles was at 98.9 per cent of its pre-war levels, metalworking at only 60.3 per cent.94 Table 16.2 Russian and Soviet textile production, 1913–1940 Total yarn (1000s tonnes)

Total cloth (million linear metres)

Cotton

Linen

Wool

Silk

1913

370

2,848

91

4

4

2

1920

40

a

139

72

2

11

0.3

1925

195b

1,860

91

7

3

0.1

1930

433

2,679

88

7

4

1

1935

533

2,978

89

7

3

1

1940

846

4,436

89

6

3

2

a

Adjusted from source to match sectoral inputs.

b

1924.

Percentage of total cloth

Source: IOTP, pp. 275, 291, 295, 298, 301.

93 Na novykh putiakh. Itogi novoi ekonomicheskoi politiki, 1921–1922 gg., vyp. 3 (1923): 9, quoted in Ward, Cotton Workers, p. 5. 94 Ward, Cotton Workers, pp. 12–14. 442

The cotton textile industry in Russia and the Soviet Union This period saw a large turnover in the workforce; it would seem that close to half of the textile labour force in the 1920s was new to the trade. While women’s share of the industrial workforce had grown during the years of war and revolution – although apparently not in cotton textiles – once peace was restored, women were often fired and replaced with men re-entering the labour market. Through the 1920s, women made up approximately 60 per cent of the cotton textile workforce, slightly less than they had before and during the war.95 Despite their numerical predominance in the industry, women only accounted for about a quarter of the sector’s party membership. A similar situation prevailed in the union, where membership on factory committees was overwhelmingly male although the membership was mostly female.96 Unions, however, had relatively little power after the civil war, and were treated as ‘transmission belts’ of government policy and not as organizations for expressing worker demands. Cotton workers would have many discontents that unions could have expressed. Consumer goods prices were generally kept artificially low in order to attract peasants into the market economy. At the same time, enterprises were expected to make a profit. As a result, many of them were forced to find ways of cutting costs. One method was to decrease production. This resulted in a worsening of the exchange problem with the peasantry, since fewer goods were put on sale at artificially low prices and were usually snapped up by urban speculators before they could percolate to the countryside. The state’s preferred solution was to increase efficiency in production, leading to a love affair with Taylorism and ‘the scientific organization of labour’. At the same time, the importance of piecerates grew again as a way of tying wages to productivity. However, a major state campaign to increase productivity in 1925 led to a widespread strike campaign that was temporarily successful.97 The final crisis of NEP began in early 1927 and would escalate through 1928 and 1929. Several factors played a role in this process. There was a perception that the economy had recovered to pre-war levels and that further growth would require greater amounts of capital investment than NEP could easily provide. A war scare made the need seem more desperate, while diminished grain collections raised fears that the surplus grain needed to trade for foreign currencies, and thus Western technology, would not be available. Many sectors of the population, especially youth and the working class, resented NEP because of the chronic unemployment and underemployment engendered by trying to preserve ties with the peasantry. Distrust of the market led to a belief that further industrialization must be planned 95 Wendy Z. Goldman, Women at the Gates: Gender and Industry in Stalin’s Russia (Cambridge, 2002), pp. 11–12. 96 Ward, Cotton Workers, pp. 24–26, 109–117. 97 On the 1925 strike wave, see Ward, Cotton Workers, pp. 176–198; for another strike wave in 1922–1923, see pp. 162–167. During the Soviet era, labour disputes were seldom reported in the press and were off limits for historical research until the last 20 years or so. As a result, there is very little information on these; the Ivanovo region strikes of the early 1930s, discussed below, are the exception that proves the rule. 443

The Ashgate Companion to the History of Textile Workers and not left to the chaotic workings of elemental economic laws. All these factors came together in the First Five-Year Plan (1928–1932), the hallmarks of which were madcap industrialization – eventually targeted at a 144 per cent increase over the course of the plan – and the forced collectivization of the peasantry to assure enough surplus grain to pay for the plan. The First Five-Year Plan freed Soviet cotton textiles of the need for importing raw supplies. One reason the civil war had affected cotton textiles so adversely was that for much of the period Central Russia was cut off from both Central Asia and foreign imports. Through the 1920s, the importance of domestic cotton grew, and by 1928 it had surpassed its pre-war levels, accounting for over 70 per cent of domestic consumption. However, between 1929 and 1931 the area of cotton sown within the USSR more than doubled, and by 1937 the Soviets exported more cotton than they consumed.98 The chaos of these years adversely affected textile workers. Attempts to rationalize production continued, but more harmful than those were the shortages of food and raw materials that caused serious hardships for cotton workers. Many of these hardships were caused by the high consumer prices instituted to help pay for industrialization. We know that at least in the Ivanovo region there were waves of strikes in 1928 and 1932, and the second wave may have led the regime to shift more resources into consumer goods production.99 However, it also led to increased secret police interference in labour matters. After 1932, strikes seem to have been very rare in the Soviet Union until the regime’s death throes in the late 1980s. The paradoxical effect on textiles of the industrialization drive was to make them much less important. The First Five-Year Plan was meant to make the Soviet Union self-sufficient industrially. As such, heavy industry received priority: precious foreign exchange would be spent on building factories that would build machines instead of buying machines that would only make perishable items. Only after heavy industry was firmly established would the Soviet Union be able to build its own looms and ring frames. It is telling that while the plan kept hiking the targets for heavy industry (from a 99 per cent increase in machinery to a 382 per cent increase), the targets for textiles actually fell, from 48 per cent to 21 per cent.100 In fact, textile production declined slightly between 1928 and 1932. While textiles accounted for 30 per cent of all Soviet industrial production in 1927, by 1930 its share was less than 18 per cent, falling from first to third place.101

98 IOTP, pp. 292–293, 315–316. 99 Ibid., p. 314; Jeffrey J. Rossman, ‘Worker Resistance under Stalin: Class and Gender in the Textile Mills of the Ivanovo Industrial Region, 1928–1932’ (Ph.D. diss., University of California, Berkeley, 1997). 100 Alec Nove, An Economic History of the U.S.S.R, rev. edn (Harmondsworth, 1982), p. 190. For a good account of the Soviet period in general that concentrates on social and economic affairs, see Andrle, Social History. 101 IOTP, pp. 312–313. 444

The cotton textile industry in Russia and the Soviet Union In addition, it seems that textiles now became gendered as female. Even though women had long predominated numerically, the profession was generally represented as male; textile workers in general were still called tkachi, not tkachikhi. This changed after 1930, as the Soviet regime began an intentional campaign to designate certain industries as purely ‘woman’s work’. The industrialization drive had produced a major labour shortage, and the regime wanted to muster as many women as possible into the industrial workforce. This influx was funnelled, however, into low-skilled and semi-skilled positions that did not demand too much training. It was projected that 90 per cent of the jobs in textiles would be filled by women; only the most hazardous and heaviest equipment would be operated by men. In any event, the proportion of textile and sewing employees who were women increased from 61 to 72 per cent from 1928 to 1936. While 5 per cent fewer women worked in cotton textiles in 1933 than in 1929, 27 per cent fewer men did. Men especially left for better-paying, higher-status positions.102 The First Five-Year Plan became the founding passion play of the Soviet economic system, and its altars were the foundry ovens of Magnitogorsk, not the carding rooms of Ivanovo. Steelworkers were considered heroes, and would remain so through the rest of Soviet history.103 It was heavy industry that defeated the Nazis. It was heavy industry that stared down the United States. Textile workers were, at best, the sweet girls waiting for steelworkers back home. Just as peasants were becoming more firmly constructed as a feminine other after 1930,104 textiles also became solidly gendered as female, as an ‘other’ that did not participate in the heroic acts of socialist industrialization. The dominant image was no longer the unyielding tkach but the literal and figurative spinster.105 Ivanovo was no longer the ‘Russian Manchester’ but the ‘city of brides’. By 1992, an official history of the textile industry could write of cotton production: ‘This industry usually employs women, and thus in the regions in which it is concentrated it is difficult for men to find work’; the possibility that men could find work in the same factories is not even contemplated.106 Beginning with the First Five-Year Plan, textiles received little investment and little attention. By 1965, they employed just over 5 per cent of all industrial workers.107

102 Goldman, Women at the Gates. 103 On the ‘heroism of steel’, see Hiroaki Kuromiya, Stalin’s Industrial Revolution: Politics and Workers, 1928–1932 (Cambridge, 1988), pp. 305–306. 104 Elizabeth Waters, ‘The Female Form in Soviet Political Iconography, 1917–32’, in Barbara Evans Clements, Barbara Alpern Engel and Christine D. Worobec (eds), Russia’s Women: Accommodation, Resistance, Transformation (Berkeley and Los Angeles, 1991), pp. 225–242, at pp. 240–242. 105 A character in Natalya Baranskaya’s famous 1969 novella ‘A Week Like Any Other’ loves to sew but became a researcher because ‘Nobody marries seamstresses nowadays’ (A Week Like Any Other: Novellas and Stories (Seattle, 1989), p. 25). 106 IOTP, p. 376. 107 Donald Filtzer, Soviet Workers and De-Stalinization: The Consolidation of the Modern System of Soviet Productive Relations 1953–1964 (Cambridge, 1992), p. 193. 445

The Ashgate Companion to the History of Textile Workers This feminine gendering and loss of status make it harder to write about textiles and textile workers because they virtually disappear from the historiography. Textiles were no longer interesting, and even in monographs that have looked at the Soviet working class, textile workers are barely ever disaggregated from the whole. Indeed, even in the better studies textile workers are often only treated at length in sections on women workers. The history of textiles during the last halfcentury of the Soviet period is one in which textiles constantly lag several steps behind heavy industry.108 In the recovery from the Second World War, gross industrial production in 1950 was 73 per cent greater than in 1940, producers’ goods 105 per cent greater, consumers’ goods only 23 per cent greater, but the production of cotton fabrics actually declined a fraction. Much more attention was given to consumer goods immediately after Stalin’s death, and between 1950 and 1955 cotton fabric production increased 51 per cent and woollens rose 65 per cent; this still, however, trailed the 85 per cent increase in gross industrial production. Even with the tremendous importance Khrushchev gave to butter over guns, cotton fabric production during his SevenYear Plan (1958–1965) only increased 22 per cent, while gross industrial output went up 80 per cent.109 Very little changed during the Brezhnev years. According to official statistics, industrial production increased 50 per cent from 1965 to 1970, 43 per cent from 1970 to 1975 and 24 per cent from 1975 to 1980. Production of fabrics, on the other hand, increased only 18, 12 and 7 per cent respectively in these same periods. On the whole, between 1928 and 1970, the production of all fabrics increased almost three-and-a-half times. During the same period, steel production increased 27 times, coal over 17 times, machine tools 120 times; even other consumer industries experienced greater increases, including leather footwear (almost 12 times) and the other grand dowager of early Russian industrialization, beet sugar (over seven times).110 One other noteworthy development was the slow but steady decline of cotton’s dominance in textile production since 1950, primarily to the benefit of silk and non-woven fabrics.

108 See, for example, Filtzer, De-Stalinization, pp. 193–196, the only multiple-page reference for ‘textile workers’ in the index. Filtzer’s works on the Soviet working class, including Soviet Workers and Late Stalinism: Labour and the Restoration of the Stalinist System after World War II (Cambridge, 2002) and Soviet Workers and the Collapse of Perestroika (Cambridge, 1994), remain the best monographs on the subject. 109 Nove, Economic History, pp. 293, 342, 355. 110 Ibid., pp. 377, 398. 446

The cotton textile industry in Russia and the Soviet Union Table 16.3 Soviet and Russian textile production, 1940–2003 Total cloth x 1 million m

2

Percentage of total cloth

Per capita Cotton

Linen

Wool

Silk

Non-woven

1940

3,320

17

82

12

5

2

–

1945

1,353

12

85

8

5

2

–

1950

3,374

19

81

10

6

3

–

1955

5,347

28

79

7

6

8

–

1960

6,636

31

73

10

7

10

0.01

1965

7,498

33

73

9

6

11

0.3

1970

8,852

37

70

9

7

13

1

1975

9,956

39

67

9

7

15

2

1980

10,746

41

66

8

7

17

3

1986

12,310

44

63

8

5

16

8

2000

2,314

79

5

2

8

7

2003

2,889

82

5

2

5

Source: IOTP, p. 396; Parshukova, ‘Textile Production’; Techtextil Rossija, Industries and Markets: Market Information, 2004, http://techtextil.messefrankfurt.com/rossija/en/branchen_ market.html; Federal’naia sluzhba gosudarstvennoi statistiki, Informatsiia o sotsial’noekonomicheskom polozhenii Rossii, Operativnaia informatsiia 2003 goda, Proizvodstvo produktsii v otdel’nykh otrasliakh promyshlennosti, http://www.gks.ru/scripts/free

Thus, while textile production increased dramatically during the period of the planned economy – almost doubling per capita output – the industry’s relative importance steadily declined, and this told most on the position of textile workers in Soviet society. Textiles were seen as a relatively low-pay, low-status, high-stress profession with little continuity between generations; the daughters of weavers were more likely to enter the fast-growing service sector, while textiles became highly dependent on short-term migrants from the countryside. Clerking in a shop was considered clear upward mobility from the shop floor, leaving textile work for migrants desperate for any position that would allow them to stay in the city.111

111 Filtzer, De-Stalinization, pp. 193–196; Filtzer, Perestroika, pp. 27–31. In 1965, textile workers made an average of 81 rubles a month; while above the average for light industry, this was well under most other industrial professions (coal-mining, 191; iron and steel, 124; construction, 106; all industry, 101) (Filtzer, De-Stalinization, p. 104). 447

The Ashgate Companion to the History of Textile Workers The Gorbachev years saw an increase in textile production, but the dissolution of the Soviet Union at the end of 1991 was followed by a catastrophic collapse of the industry. According to one measure, fabric production rose 6 per cent between 1985 and 1990, but fell 79 per cent between 1990 and 1995. Many of the mills that were still open in the late 1990s were operating at only 20 per cent capacity with equipment that was old and outdated. In 1996 underemployment and unemployment were estimated to total 70 per cent for the sector. While most of the country’s mills were still operating, two-thirds of them were doing so at a loss. As one observer noted, ‘in several regions including Ivanovo almost every town is dependent on a single textile factory. If they close, people will be without work. People come to work because there is nowhere else to go. At some factories, employees have not been paid in six months, and even in this situation their monthly salary is 230,000 roubles (US  $40).’ Privatization after 1992 was very haphazard, and most enterprises lacked any working capital or experience in marketing. What investment was available was provided by large trading companies interested more in short-term profits than in long-term investments. Fixed capital had already been deteriorating before the breakup of the USSR; afterwards, very little improvement took place. At the same time, the traditional sources of raw material, such as Central Asian cotton, were now located in other countries that were charging much higher prices than had obtained in the days of Soviet control. In addition, cheaper imports from South and South East Asia have taken a firm hold in the Russian market, growing to approximately 75 per cent of textile and apparel sales in 2003.112 While there was a marked increase in domestic production following the collapse of the rouble in late 1998 (15 per cent in 1999, 44 per cent in 2000), and despite hefty duties on imports, only one-fifth of textile factories showed a profit in 2000. By 2002, the textile and apparel industry contributed only 1.5 per cent of Russia’s gross domestic product, compared to 12 per cent in Soviet times. It was only the twelfth largest of Russia’s 15 major industries, employing 800,000 people (80 per cent female) in 16,000 enterprises. Cotton mills were only operating at 52 per cent of capacity, silk at 30 per cent, linen at 28 per cent.113 The one-time engine of Russian industry, cotton textiles fell back to caboose during the period of the command economy. Protected from foreign competition, but benignly ignored as far as investment was concerned, the industry has been largely incapable of competing in the brave new world of globalization. 112 Maria Breiter, ‘Textile Equipment Market in Russia’, U.S. Foreign Commercial Service (November 1997), http://www.bisnis.doc.gov/bisnis/isa/9711text.htm; Marina Parshukova, U.S. Foreign Commercial Service (1 December 2003), http:// www.apparelandfootwear.org/data/isa0312russia.pdf; ‘Light Industry’, http://www. russiansabroad.com/russian_history_231.html. The quote is from Breiter. 113 Marina Parshukova, ‘Textile Production in Russia – A Snapshot’, U.S. Foreign Commercial Service (20 June 2001), http://www.bisnis.doc.gov/bisnis/country/ 010627rustextiles.htm; Marina Parshukova, ‘Apparel and Textiles Market – Industry Sector Analysis (ISA)’, U.S. Foreign Commercial Service, U.S. Embassy – Moscow (1 December 2003, expires 30 September 2005); John Gibbon, ‘Russia’s Textile Industry’, www.director-e.com (July 2002), http://www.director-e.com/director-e/features/index. asp?view=display&featureID=14. 448

17 Spain Angel Smith, Carles Enrech, Carme Molinero and Pere Ysàs

Within a Western context the Spanish cotton and woollen textile industries by no means stood out because of their size and dynamism. They attained considerable importance in Catalonia during the nineteenth century, but they tended to be rather inward looking; largely unable to compete in foreign markets they sheltered behind increasingly high tariff barriers. Where, in the present context, a study of these industries can be very useful is with respect to the light they shed on processes at work on a global scale. In this chapter we have focused our attention on three broad, interrelated areas: the economic development and evolving industrial structure of the industries; their changing gender and ethnic composition; and labour relations, union organization and worker protest.

From the pre-industrial era to the antecedents of factory production, 1650–1814 The seventeenth century crisis saw a spectacular process of de-industrialization in Spain, during which the destruction of urban manufacturing was not compensated for by any shift in production to rural areas. Nevertheless, rural Spain did become the principal focus for what remained of the textile industries. In the second half of the seventeenth century they were above all centred on the manufacture of woollen cloth, followed by linen and silk. They were characterized by their dispersion, though in some parts of Castile, Catalonia, Valencia and Galicia pockets of more concentrated manufacture were to be found. Simultaneously, the economic decadence of the cities produced a fall in demand among the elites and the middle 

Angel Smith and Carles Enrech have focused on the period 1650–1939. Carme Molinero and Pere Ysàs have written the section on 1939–2000, with additional statistical data by Carles Enrech.

Map 17.1 Regions of Spain

Map 17.2

Catalonian textile regions

Spain classes and, consequently, a decline in the quality of woollen cloth. Meanwhile, the market for fine cloths was covered by foreign suppliers, whose wares inundated both Spain and its colonies. As we shall see, industrial structure varied considerably from area to area, but some common characteristics could be discerned. As in Europe as a whole, the sexual division of labour was very marked. Masters and artisans, who worked in what remained of the urban manufactures, were men who had usually undertaken an apprenticeship, formed part of a guild and for whom textiles frequently provided the main source of income. The journeymen worked either from home or in small workshops owned by their masters. Their wives would also be involved in the productive process, working as assistants on looms within the workshops, while from home they would sow lace and silks (blondas y encajes) for putting-out merchants. In rural areas, on the other hand, textiles complemented agricultural labour. And while men were normally handloom weavers, women undertook spinning and carding, which were easier to combine with the household chores. Under the new Bourbon monarchy the first decades of the eighteenth century were to see the introduction of protectionist measures, the setting up of a Junta de Comercio in order to promote manufacturing and trade, and the establishment, in Castile above all, of a series of Royal Factories. For the most part these factories were designed to satisfy luxury consumption at the royal court, but in textiles they operated as pilot projects, aimed at re-enforcing the monarchy’s re-industrialization policy by substituting imports, fomenting high-quality production and spreading technical and product innovation. In textiles about a dozen Royal Factories were set up, the first and most important of which was the Royal Woollen Cloth Factory of Guadalajara, founded in 1719, which by 1791 had 500 looms and employed close on 4,000 artisans on the premises, and over 18,000 women and girls, who either worked from home or from some 190 spinning ‘factory schools’ established by it in rural areas. Yet rather than a road to modern factory-based industry the Royal Factories were to prove a dead end. They were bureaucratic and did not use the factors of production effectively, with the result that they were loss-making and, in general, failed to mechanize. The transition to a modern factory-based industry would come from the private sector, but the structure and fate of the textile industries in the various Spanish territories would be very different. The linen industry, centred on Galicia, represents an extreme example of atomized production by small, family units, with hardly any intervention by merchant manufacturers (the so-called Kaufsystem). The industry was undercapitalized and production techniques rudimentary, and in these circumstances it would be impossible to undertake the transition to

 Pierre Vilar, Cataluña en la España Moderna (3 vols, Barcelona, 1978), vol. 1, pp. 443ff.  Agustín González-Enciso, Estado e industria en el siglo XVIII. La fábrica de Guadalajara (Madrid, 1980), pp. 337ff; Juan Huelguera Quijada, ‘Empresas y empresarios manufactureros en la España del siglo XVIII’, in Francisco Comín and Pablo Martín Aceña (eds), La empresa en la historia de España (Madrid, 1996), pp. 115–140. 451

The Ashgate Companion to the History of Textile Workers factory-based production. In other areas merchant manufacturers played a more important role in organizing production (the Verlagssystem), though this in itself did not guarantee the emergence of modern industry. This was the case of both Valencian silks and Castilian woollen textiles. In Valencia manufacturers put out the silks to domestic artisans, the latter organized in guilds. However, much more effort was put into the highly profitable production and export of raw silk, a fact which helps explain why production was again atomized, investment very limited and the final product lacked standardization. Some similar problems could be seen in the Crown of Castile. Woollen manufacture in this territory was, in terms of volume, the most important in Spain. In 1751 there were 45,000 workers employed in the industry, and in the last quarter of the eighteenth century it produced a minimum of 7,149,103 metres of cloth per annum. Low-quality cloth remained predominant, though rising rural incomes did result in some growth in the high-quality end of the market. Within the Castilian industry a system emerged through which a mercader (equivalent to the Verleger) both financed and oversaw the process of production. Its origins are to be found in the 1515 Ordenanzas Generales de Obraje de Paños, which established the owner of the cloth as the central figure within the industry, even if he had not been examined in any particular craft. Henceforth, the master weavers, carders, shearers and dyers could only undertake a particularly craft, while the mercader oversaw the entire production process. This, as we shall see, was a common feature of the most dynamic sectors of pre-industrial textile manufacture. Yet it was not sufficient to guarantee a successful transition to factory production. In the late eighteenth century the Castilian mercaderes did not provide a sufficiently high-quality, costeffective alternative to Catalan textiles and succumbed under the competition. It was only in this latter territory that a transition to factory production took place, both within woollen textiles and in the new sector of cotton textile manufacture. In comparison with the Crown of Castile the size of the woollen textile industry was modest (with 2,088 looms compared to 8,902 in 1760), but over the century it was to prove more dynamic. There were similarities with the situation in the Castilian towns, with so-called pelaires or paraires, who had originally bought the wool and undertaken the processes prior to spinning, emerging as key players, both organizing the production process and marketing the final product through networks of muleteers, traders and shopkeepers. As part of this process they reordered the guild system to their advantage and subordinated other trades to  Joám Carmona Badía, El atraso industrial de Galicia. Auge y liquidación de las manufacturas textiles (Barcelona, 1990).  Vicente Martínez Santos, Cara y cruz de la sedería Valenciana (Siglos XVIII–XIX) (Valencia, 1981).  For this section on the Crown of Castile we have used Emiliano Fernández de Pinedo, ‘Coyunturas y política económica’, in Centralismo, Ilustración y agonía del Antiguo Régimen (1715–1833), vol. VII: Historia de España, dirigida por Manuel Tuñón de Lara (Barcelona, 1980), pp. 98–99; the articles by Angel García Sanz and Helguera Quijada in Comín and Pablo Aceña, La empresa, pp. 93–113 and 117–119; and John Lynch, La España del siglo XVIII (Barcelona, 1999), p. 194. 452

Spain their interests. In the seventeenth century, while they employed female spinners from rural areas on piece-rates, in the towns and villages weavers had formed separate guilds from which the pelaires had to buy the cloth. However, during the eighteenth century they were able to break the power of the weavers’ guilds and employ journeymen weavers directly. The difference with Castile was that the manufacture of high-quality woollens was concentrated in several larger towns (above all Sabadell, Terrassa and Igualada) and that the pelaires were able to take advantage of a more dynamic economic climate. This encouraged a process of social differentiation and the emergence of a rich and powerful group of ‘industrialists without factories’, who concentrated key production processes in their own hands (especially the tasks prior to spinning and finishing), kept a strict eye on costs, intervened directly in buying the wool and marketing the final product, and put further emphasis on high-quality production. Yet the key industry of the modern era was to be cotton textiles. The first socalled indianas factories appeared in Barcelona between 1728 and 1742, dedicated largely to the printing of contraband cloth. However, it was only from the 1760s, with the liberalization of the import of raw cotton and cotton thread, that the development of a full-fledged industry became possible. Weaving and printing were from the first undertaken by male artisans operating in proto-factories of considerable dimensions, in which, despite the pre-industrial technology, modern features such as the introduction of a management hierarchy, a relatively high intensity of labour and the use of the division of labour could be discerned. On the other hand, spinning remained dispersed throughout the countryside, and left to women working in a domestic environment. The expansion of the industry needs to be placed within the context of the rapid economic development which took place in Catalonia, especially from the second half of the century, on the back of the export of wines and spirits to the American colonies, and a consequent growth in demand for cotton cloth, which was viewed as a more ascetic and hygienic alternative to other textiles. Its consolidation would take place with the liberalization of commerce with the American colonies  For Igualada, see Jaume Torras Elías, ‘Fabricants sense fàbrica. Estudi d’una empresa llanera d’Igualada (1726–1765)’, Reçerques 19 (1987): 145–160, and ‘Gremios, familias y organización del trabajo. Las cofradias de oficio en los siglos XVII y XVIII’, in Santiago Castillo (ed.), El trabajo a través de la historia (Madrid, 1996), pp. 171–180; for Sabadell and Terressa, see Josep Maria Benaül Berenguer, ‘Los orígenes de la empresa textil lanera en Sabadell y Terrassa en el siglo XVIII’, Revista de Historia Industrial 1 (1992): 39–62.  Eloy Martín Corrales, ‘La importación de telas de algodón levantino’, Revista de Historia Industrial 6 (1994): 47–74.  According to a late eighteenth-century census cotton textile establishments in Barcelona had an average of 108 workers: see Isabel Miguel López, ‘El censo de Manufacturas de 1784. Una nueva fuente para el análisis de la industria catalana’, Revista de Historia Económica 14 (1) (1996): 125–183. Establishments were so big above all because large numbers of workers were needed to carry out the finishing processes, especially bleaching, which took place in fields outside the city walls. 453

The Ashgate Companion to the History of Textile Workers – which had previously been funnelled through Cadiz – in 1783. This would lead to a further specialization, with some factories weaving and/or printing cotton and others stamping (largely imported) cotton and linen cloth. Whereas the former predominated in the internal Spanish market the latter were particularly sought after in the American colonies, making possible both a process of rapid capital accumulation and the import of American raw cotton on the return leg, which in turn gave a great boost to cotton spinning. This process was accompanied by the introduction of new technological innovations – the first spinning jennies in 1774 (followed by a larger Catalan variant, the Bergadana), carding machines in 1792 and the water frame in 1793. At first the jennies were still operated in a domestic environment, but they would lay the basis for the incorporation of spinning into factories.10 Overall, between 1776 and 1790, in Catalonia as a whole, output probably grew over sevenfold. Just in Barcelona, in 1784 the industry produced 5,745,600 cotton pieces (indianas), along with 3,396,790 pieces of stamped cotton and linen cloth (lienzos pintados).11 This made cotton the most innovative – though not the largest – sector within the Catalan textile industries (see Table 17.1). Table 17.1 Catalan textile industries in 1784 Factories Looms

%

Workers

%

Cotton (incl. stamped and painted pieces)

147

3,054

44.4

12,279 23.8

Silk scarves and handkerchiefs

n.a.

1,842

26.8

6,918 13.4

Wool

n.a.

1,147

16.7

17,463 33.9

Linen and hemp

n.a.

840

12.2

5,183 10.1

Total (weaving)

n.a.

6,883 100.0

41,843 81.2

Knitted wool and cotton

53

690

18.0

2,059

4.0

Knitted silk

n.a.

1,098

28.7

3,439

6.7

Silk ribbons

n.a.

2,044

53.3

4,189

8.1

Total (knitting and ribbons)

n.a.

3,832 100.0

9,687 18.8

Source: Isabel Miguel López, ‘El censo de Manufacturas de 1784. Una nueva fuente para el análisis de la industria catalana’, Revista de Historia Económica 14 (1) (1996): 125–183. 10 For the cotton industry in the eighteenth and early nineteenth centuries, see also the studies by Alejandro Sánchez, ‘La era de la manufactura algodonera en Barcelona, 1736–39’, Estudios de Historia Social 48–49 (1989): 65–113 and ‘Crisis económica y respuesta empresarial. Los inicios del sistema fabril en la industria algodonera catalana, 1797–1839’, Revista de Historia Económica 18 (3) (2000): 485–522. 11 A discussion of levels of production in the industry is to be found in Alexandro Sánchez, ‘La indianería catalana: ¿mito o realidad?’, in Revista de Historia Industrial 1 (1992): 213–232, at p. 225. 454

Spain In cotton textiles this industrial model would be challenged by the colonial wars with Britain in the late eighteenth and early nineteenth centuries, which led to the interruption of the colonial trade. Textile manufacturers had no choice but to abandon the production of stamped cotton and linen and focus on indianas for the home market. This led to a slowdown in growth, but in a context in which large amounts of capital had already been accumulated they responded positively and undertook further technological modernization, with, from the 1790s, the transfer of spinning, which had previously been an activity largely carried out by female peasants, to factories operated by both horse and water power. These were located above all in Barcelona, Manresa, and on the banks of the Llobregat and Ter rivers in northern Catalonia.12

The rise of a mechanized factory system, 1814–1885 In the new century industrial Catalonia (above all the area around Barcelona) rose to become the ‘factory of Spain’; a major industrial centre in what was still to a large extent an agrarian economy. Because of the difficulty of developing heavy industry this industrial growth would to a remarkable degree be based on textiles, with the cotton textile industry to the forefront. By 1900 there were about 80,000 workers in cotton textiles, and another 25–30,000 in the territory’s other textile industries. This represented about one-third of Catalonia’s entire industrial working class. In Spain as a whole in 1900 it is probably not unrealistic to talk of one million plus industrial workers, but outside Catalonia textiles had very much withered away. In 1920 there were 40,000 workers left in the rest of Spain, most still working in a domestic environment.13 It is in the light of these figures that the sections dealing with the period 1814–1939 will centre on Catalan cotton textiles. For this industry the early nineteenth century represented a difficult period of adjustment. Most of the country’s industrial structure survived the War of Independence (1808–1814). More damaging was the loss of Spain’s colonial empire combined with the constant social and political conflicts which afflicted the country between 1815 and 1839. In response textile manufacturers successfully pushed for higher protective tariffs. This would set the tone for the next 150 years. Spain’s own variant of the German ‘pact of iron and rye’ was forged, whereby both industrialists and cereal producers were protected from foreign competition and dominated the home market. The difference was that it would become a defensive alliance, with Spanish cotton manufacturers, for example, only able to compete effectively with the major industrial powers in low-quality fabrics. 12 Sánchez, ‘Crisis económica y respuesta empresarial’, pp. 494–497. 13 For Catalonia, see Pere Gabriel, ‘La població obrera catalana ¿una població obrera industrial?’, Estudios de Historia Social 32–33 (1985): 191–231. The data for 1920 has been calculated by Angel Smith from Ministerio de Trabajo y Previsión, Servicio General de Estadística, Censo de población de 1920, Tomo V (Madrid, 1929), pp. 421–424. 455

The Ashgate Companion to the History of Textile Workers Nevertheless, despite the handicaps they faced, industrialists were able to consolidate a factory-based industry. From the 1810s the mechanization of spinning proceeded apace. The industry’s modernization was further encouraged by the rapid switch to steam power in Barcelona and nearby textile towns during the 1840s, which made possible the introduction of large mule-jennies in spinning along with power-looms, and stimulated vertical integration of spinning and weaving within larger plants. By 1860 spinning had been totally mechanized while power-looms accounted for half the industry’s total (for spinning see Table 17.2).14 In addition, the number of workers also grew rapidly; over 72,000 cotton textile workers in Barcelona by 1841 and close on 105,000 in Spanish cotton textiles as a whole in 1867 (for the 1867 figures see Table 17.3).15 During this period, consequent upon technological change not only did the labour process undergo frequent modifications, the sexual division of labour was also to suffer profound changes. And in the new liberal age the industry’s transformations became closely bound up with labour organization and protest. As in the rest of Western Europe, the belief that men were the primary breadwinners outside the domestic environment, and that the role of women should be complementary and combined with child-rearing and domestic chores, was transposed onto the new factory environment. As a result, supervisory work, and what was considered skilled labour, remained the preserve of men, while female labour was conceptualized as low-grade and poorly paid.16 Yet rapid technological change would call into question the type of work undertaken by male and female workers. Thus, in weaving, while no one doubted that handlooms were the preserve of male journeymen, in the 1830s industrialists began to employ some women on the new power-looms. These machines required a short period of training and were relatively easy to operate, and could, therefore, be regarded as requiring ‘only’ low-paid female labour. The industry’s highly competitive environment also led to other concerns among the male handloom weavers. There were complaints at industrialists’ attempts to increase the length of pieces woven, supervise the production process more closely and water-down apprenticeship restrictions.

14 Jordi Maluquer de Motes, ‘Las técnicas hidráulicas y la gestión del agua en la especialización industrial de Cataluña. Su evolución a largo plazo’, in María Teresa Pérez Picazo and Guy Lemeunier (eds), Agua y modo de producción (Barcelona, 1990), pp. 311–348, at p. 336. 15 For the 1841 data, see Llorenç Ferrer i Alòs ‘Notas sobre la familia y el trabajo de la mujer en la Cataluña Central (siglos XVIII–XX)’, Boletín de la Asociación de Demografía Histórica 12 (2–3) (1994): 199–232, at pp. 209–10. 16 Though the fact that wage differentials were to a large extent based on job differentiation meant that on occasions when women performed similar tasks to men they could earn relatively high wages. This was the case, for example, of female self-acting mule spinners in the large Barcelona factory La España Industrial in the late nineteenth century. 456

Spain Table 17.2 Spindles in the Catalan cotton textile industry, 1807–1929 1807 Hand spindles

1829

1836

1841

1850

1857

82,870 290.700 457,200 446,400 183,778

1907

1929

0

0

0

Mechanical spindles Water frames Throstles Mulejennies

10,890

0

0

0

0

0

0

0

0

7,028

12,800

19,800

51,040

70,360

0

0

70,285 181,047 326,323 475,490 388,650

0

0

740,000

436,344

2,040

Self-actors

0

0

0

0

Ring frames

0

0

0

0

Total spindles % mechanical

96,328 437,054 0

1,315,567 1,655,630

95,800 386,013 651,047 792,723 622,858 896,064

2,055,567 2,101,974

13.5

21.0

29.7

43.6

0

77.2

100.0

100.0

100.0

Sources: Sánchez, ‘Crisis económica y respuesta empresarial’, p. 508. Jordi Maluquer de Motes, ‘La gran transformació. Industrialització i modernització a la Catalunya del segle XIX’, in Història Econòmica de la Catalunya Contemporània (4 vols, Barcelona, 1994), vol. I, pp. 41–186, at p. 168; Federich Rahola y Trémols, ‘Del Comerç y de la industria de Catalunya’, in Francesc Carreras y Candi (ed.), Geografia general de Catalunya (4 vols, Barcelona, 1980 [1913–18]), vol. I, pp. 323–464, at p. 397; Lucas Flórez Beltrán, La industrial algodonera española (Barcelona, 1943), p. 93.

It was in this context that during more liberal periods of government (1841–1843, 1854–1856) these weavers played the leading role within the expanding industrial towns in forging the new labour movement, and would be at the forefront of demands for trade-union rights to be respected. Their importance was based on two key factors. On the one hand was their numerical weight within the rapidly growing industrial towns and, on the other, their strong tradition of independence on the shop floor, which included the right, as they saw it, to talk on the job, leave work when they wished and celebrate ‘Saint Monday’.17 17 For the weavers and their union organization during these years, see Manuel Reventós, Assaig sobre alguns episodis històrics dels moviments socials a Barcelona en el segle XIX (Barcelona, 1925), pp. 20–21; Josep M. Ollé Romeu, El moviment obrer a Catalunya. Textos i documents (Barcelona, 1973); Genís Barnosell, Orígins del sindicalisme català (Vic, 1999). 457

The Ashgate Companion to the History of Textile Workers In the 1850s spinners also became increasingly prominent in labour protest, a fact which again needs to be related to the changing structure and sexual division of the trade. At first within the new factories spinning, along with the processes prior to spinning and weaving, was seen as women’s work. This was called into question with the rapid introduction of the new mule-jennies after 1814. They were large semi-automatic machines, unlike anything that had gone before, which required considerable physical effort and the spinner to work alongside a team of youngsters, known as piecers, who changed the spools and mended the broken thread. As in other European countries this encouraged the replacement of women by men – ‘heads of families’ in the spinners’ own words – within the trade.18 The process has not been studied in any detail. Crucial, no doubt, was the physical strength required to operate the new machines. In addition, they made it possible for the spinners to play a supervisory role and train their piecers. Hence their position could be assimilated to that of an artisan. Indeed, we know that informal apprenticeship systems were established in which male spinners acted as subcontractors, paying piecers out of their wages.19 The situation was again to change in the 1850s with the introduction of the self-acting mules. These were fully automatic and required less physical effort. Hence the temptation among employers to cut costs by replacing male with female spinners. By 1856 in Barcelona, while men operated 66.7 per cent of the mulejennies, they were employed on only 21.2 per cent of the self-acting mules.20 In 1854 this prompted a violent general strike which obliged the captain general to issue a proclamation ordering the employers to turn their self-acting mules into mulejennies.21 The social history of cotton textiles over the next 25 years is still in need of further study. Nevertheless, several elements clearly stand out. The evolving sexual division of labour needs to be seen in the context of a series of economic and sociocultural factors. Much of the textile industry in small-town Catalonia coincided with relatively prosperous wine growing areas, and here the male head of household tended the family’s land. In addition, in the area around Manresa on the Llobregat river (the Bages), where scarves and ribbons were still woven on handlooms, men generally preferred to undertake such work along with agricultural tasks rather than enter the factories. On the other hand, in the industrial towns which sprang up around Barcelona and were incorporated into the city in 1897 (especially Sants In Barcelona itself in 1856 workers in textiles made up over 30 per cent of the industrial working class, and the city’s 5,195 handloom weavers represented a little over 50 per cent of the total number of men in the textile industries: see Ildefonso Cerdà, ‘Monografía estadística de la clase obrera en Barcelona en 1856’, in Teoría general de la urbanización, reforma y ensanche de Barcelona (3 vols, Madrid, 1968), vol. 2, pp. 595–598. 18 Reventós, Assaig, pp. 77–78 and 96; Josep Mª Vilà, Els primers moviments socials a Catalunya (Barcelona, 1935), pp. 50–53. 19 Cerdà, ‘Monografia’, pp. 595–598. 20 Ibid. 21 Reventós, Assaig, pp. 20–21; Josep Benet and Casimir Martí, Barcelona a mitjan segle XIX (2 vols, Barcelona, 1976). 458

Spain and Sant Martí), in those close to Barcelona (above all Terrassa and Sabadell) and those which grew up around the Ter river to the north (with Manlleu the most prominent example), the very rapidity of industrial development, combined with heavy migration, meant that such alternative work was not available. In those areas in which men worked on the land and/or there were plentiful alternative jobs, they disdained all but what were considered the most elite positions within textiles, while in other areas they sought a broader range of occupations. Spinning retained its elite status, with the result that over much of Catalonia men ‘captured’ the self-acting mules and turned them into the equivalent of a highpaid artisan trade. By the early 1880s they had re-established a dominant position, even it seems in the areas around Manresa.22 In Barcelona and the surrounding industrial towns women were employed only on small self-acting mules, as in the case of La España Industrial. Furthermore, with some exceptions (such as the Batlló factory in Barcelona and Reus in the south of Catalonia, where groups of women did most of the actual spinning and piecing), men transplanted the subcontract system from the mule-jennies, retaining both relatively high wages and a powerful position in the production process.23 In weaving, on the contrary, the ease of the power-loom’s operation meant that the job remained both more poorly paid and of a lower status than spinning. Hence, in areas where there were abundant alternative sources of employment men did not go into weaving, but where such work was scarce they managed to a large degree to ‘capture’ the power-looms and apply informal apprenticeship systems. The overall result was important local variations in the gender balance within the industry. Where there were more attractive alternatives to working in the textile factories the percentage of women workers was high (at its highest in Manresa, the capital of the Bages, with 79.2 per cent of the workforce comprising adult female workers in 1868).24 On the other hand, where they were limited more male workers were employed. Thus, in Sants, one of the most extreme examples, in Spain’s largest cotton factories El Vapor Vell and La España Industrial in 1866 42.6 per cent and 65 per cent respectively of the workers were adult males.25 The men’s ability to 22 See the comments in this respect in Angel Smith, ‘Industria, oficio y género en al industria textil catalana, 1833–1923’, Historia Social 45 (2003): 79–99, at p. 84. 23 Cristina Borderías, ‘Women’s Work in the Barcelona Labour Market, 1856–1936’, in Angel Smith (ed.), Red Barcelona: Social Protest and Labour Mobilization in the Twentieth Century (London, 2002), pp. 142–162, at pp. 151–156; Carles Enrech, Indústria i ofici. Conflicte social i jerarquies obreres en la Catalunya tèxtil (1881–1923) (Bellaterra, 2005); Albert García i Balañà, La fabricació de la fàbrica. Treball i política a la Catalunya cotonera (1784–1874) (Barcelona, 2004). 24 Ferrer i Alòs ‘Notas’, pp. 209–210; Albert García and Raimon Soler, ‘La formación de un centro industrial en Cataluña. Vilanova i la Geltrú durante la primera etapa de la Revolución Industrial’, Comunicación al VI Congreso de la Asociación de Historia Económica, Girona, 15–17 de septiembre de 1997, pp. 173–187 (unpublished manuscript). 25 Calculated by Carles Enrech from Información sobre el derecho diferencial de bandera y sobre los de aduana exigibles a los hierros, el carbón piedra y los algodones (4 vols, Madrid, 1867), 459

The Ashgate Companion to the History of Textile Workers monopolize more elite trades and also, if need be, undertake a wider range of jobs, meant that in Catalonia between the 1840s and the 1860s there may well have been an increase in the percentage of male workers employed. Data on cotton textiles in Catalonia indicate that adult males made up 38 per cent of the workforce in 1841 and data for Spain as whole for 1867 put the figure at 52 per cent (see Table 17.3). Meanwhile, according to Ildefonso Cerdá’s detailed study, men comprised 48 per cent of workers in the textile industries as a whole in Barcelona in 1856.26 Table 17.3 Workers in the Spanish cotton textile industry in 1867 Men

Women

Children

Total

%

Spinning

4,000

7,000

3,000

14,000

13.4

Weaving

25,400

6,605

1,025

33,030

31.5

Finishing

6,650

800

200

7,650

7.3

Weaving mixes

18,662

23,625

7,892

50,179

47.9

Total

54,712

38,030

12,117

104,859

100.0

52

36

12

100

%

Source: Calculated from Información sobre el derecho diferencial de bandera y sobre los de aduana exigibles a los hierros, el carbón piedra y los algodones (4 vols, Madrid, 1867), vol. 4, ‘Algodones’.

These male workers were able to ensure that the organization of factory production favoured them in other respects. Although working hours remained long (in Barcelona a 12-hour day until 1873, 11 hours from 1873 through to 1913) and piece-rates operated throughout the industry, all the indications are that between the 1850s and the mid-1880s there were some modest increases in real wages, especially during periods in which unions were able to operate relatively freely. Moreover, in the major industrial towns during economic downturns work was shared out among the labour force as a whole, waiting lists were established for vacant posts, promotion was to an important degree dependent on length of service, workers had a say in the setting of piece-rates and industrial discipline was not that tight.27 Much more work is needed on these questions, but several factors probably permitted these male ‘journeymen’ to retain a degree of influence on the shop floor and prevented their substitution. In the first place, from the 1830s through to the vol. 4, ‘Algodones’. 26 Ferrer i Alòs ‘Notas’, pp. 199–232; Cerdá, ‘Monografía’, pp. 595–598. 27 Carles Enrech, ‘Estructures laborals i jerarquies obreres en la industria tèxtil catalana, 1881–1923’ (Ph.D., Universitat Autònoma de Barcelona, 2000), pp. 330–349. 460

Spain early 1880s (with the exception of the ‘cotton famine’ of 1861–1865) the economic climate was very favourable. This period ended with a great boom known as the ‘gold fever’ between 1875 and 1883.28 This no doubt made industrialists little disposed to provoke conflicts through attempts to cut costs further. Secondly, in smaller industrial centres it seems that the workers’ position was strengthened by labour shortages. On the one hand, ‘mixed’ peasant-industrial workers were unreliable because they left the factories at harvest time; on the other hand, larger industrial centres exercised a strong pulling power on full-time workers.29 Finally, there was probably a degree of complicity among industrialists, who would have been influenced by dominant cultural paradigms regarding the employment of male and female labour. As for the women themselves, their subaltern status was consolidated. Male heads of household were to a large extent able to command the best paid posts within the industry with the result that patriarchal family relations were transposed to the factory, with male foremen and spinners overseeing the work of youngsters, women and children. For most women factory work remained a complementary activity and they tended to leave the factory when they began to have children (though, given low real incomes, they would continue to undertake paid labour from home).30 Nevertheless, jobs in textiles were relatively well paid and coveted. Hence, in the larger factories in Barcelona, and one suspects in other areas, families of long-standing within the neighbourhood monopolized employment opportunities. More recent female migrants, on the other hand, worked above all as domestic servants.31 Labour’s position was reinforced after the ‘Glorious Revolution’ of September 1868, which inaugurated a five-year period during which unions were able to operate quite freely. The year 1869 saw the birth of the Federación de las Tres Clases de Vapor (TCV), which grouped together factory workers in the preparatory processes, spinning and weaving. Its practice echoed that of the British New Model Unions, with its appointment of paid officials and emphasis on prudence and negotiation. It was formally apolitical but was close to the republican movement and suspicious of the new Spanish branch of the International Working Men’s Association. This became particularly clear when, after several years of heavy repression between 1874 and 1880, it re-emerged and refused to join the anarchistled Federación Regional de Trabajadores de la Región Española. The union was built by male spinners and weavers in Barcelona and the surrounding industrial towns, and aimed to perpetuate artisan structures within 28 Jordi Nadal, El fracaso de la revolución industrial en España, 1814–1913 (Barcelona, 1973), pp. 207–208. 29 On patterns of migration in the nineteenth century, see Enriqueta Camps, La formación del mercado de trabajo en la Cataluña del siglo XIX (Madrid, 1995). 30 See Borderías, ‘Women’s Work’. 31 Angel Duarte, ‘Mayordomos y contramaestres. Jerarquía fabril en la industria algodonera catalana, 1870–1890’, Historia Social 4 (1989): 3–20; Borderías, ‘Women’s Work’, pp. 157–159. 461

The Ashgate Companion to the History of Textile Workers the industry. Nevertheless, the male leaders of the TCV did at least attempt to affiliate women (unlike some skilled workers – including spinners – in English manufacturing in the late nineteenth century) and improve their wages and working conditions. This occurred within a context in which there was considerable state and employer hostility to union organization, encouraging textile workers to build industrial unions which, if necessary, could mobilize the entire workforce.

The offensive against the cotton textile trades, 1885–1939 Despite the growth of these years the Catalan cotton textile industry was never as competitive as those of the major European industrial powers. Given the lack of coal reserves industrialists who used steam power had to import relatively high cost British coal through the port of Barcelona. Furthermore, they were dependent on the small and diverse internal Spanish market, in which demand fluctuated wildly according to the success or otherwise of the summer harvest. Indeed, in order to ensure capital was not idle for long periods, rather than expand plant size, at times of peak production the manufacturers put out weaving to dependent producers, who, in rural areas and small villages, frequently still used handlooms. The result was that Catalan factories were relatively small compared to their British counterparts and also failed to specialize to such a degree. In 1919 there was a smattering of large integrated weaving-spinning factories which employed between 500 and 2,000 workers, but the average number of workers in a Catalan cotton factory (excluding the finishing trades) was 87 (130 workers per spinning plant and 73 workers per weaving shed).32 These weaknesses were exposed from 1883 when the import of cheap grain from the United States and Russia provoked a crisis in Spanish agriculture, thereby hitting demand for textiles. Moreover, industrialists in Barcelona and its surroundings, who were reliant on steam power, faced a growing competitive threat. From the early 1870s improving transport links with the towns on the Llobregat and Ter rivers (the hub of the area known as the Muntanya) had given a great boost to spinning and joint spinning factories in these areas and tempted some such enterprises based in Barcelona, the surrounding industrial towns and littoral (the so-called Pla) to join them (see Table 17.4).

32 Calculated by Angel Smith from Cámara Oficial de Industria de Barcelona, Memoria reglamentaria del año 1919 (Barcelona, 1920), pp. 137–162. One suspects these figures underestimated the number of small establishments. 462

Spain Table 17.4 Distribution of the Catalan cotton textile labour force, 1850–1919 Industrial area

1850

1861

1890–1892

1919

Pla

68%

69%

45%

51%

Muntanya

32%

31%

55%

49%

75,432

40,279

73,000

101,146

Total number of workers

Note: The Pla includes those areas in which steam power was predominant: Barcelona, the Maresma, Garraf, Vallés, Baix Camp and Baix Llobregat. The rest of Catalonia is Muntanya. Sources: Carles Enrech has calculated the data for 1850, 1861 and 1890–1892; Angel Smith has calculated the data for 1919. The original sources are Guillermo Graell, Historia del Fomento de Trabajo Nacional (Barcelona: Viuda de Luís Tasso, n.d. [1910?]); M. Giménez Guited, Guía fabril e industrial de España (Madrid, 1862); Joan Sallarés y Pla, El trabajo de las mujeres y de los niños (Sabadell, 1892); Cámara Oficial de Industria de Barcelona, Memoria reglamentaria de 1919 (Barcelona, 1920).

Such a strategy had advantages and disadvantages. Factors such as the industrialists’ political power within a rural/small-town environment, the factories’ location far from Barcelona and their use in part of ‘mixed’ peasant/industrial workers allowed them both to pay up to 25 per cent less and operate a working week between four and eight hours longer than in the Catalan capital. However, while peasants from the surrounding countryside could be paid less they had little or no experience of factory production and continued to be very much linked to the land. Furthermore, the industrialists could not rely on a quiescent labour force. Much of the workforce came not from the countryside but from Manresa and from a series of smaller industrial towns along the Ter and Llobregat rivers (which by the end of the century had between 1,500 and 6,000 inhabitants), and between 1868 and 1873 these areas became foci of unions and strikes. In order to deal with these problems the largest factories took the first steps along the road to constructing company towns (colònies industrials) by building housing alongside the factory and, in a few cases, erecting a church and opening company shops for the workers.33 The intense competition faced by the Pla industrialists, who found themselves with both antiquated machinery and relatively high energy and labour costs, greatly aggravated the crisis of the mid-1880s. In response the big factories around Barcelona, La España Industrial, Güell and Batlló, led a movement to cut costs. There were two ways in which they could do this: firstly, via the intensification of labour, increasing the number of power-looms and spinning machines each worker 33 For an overview, see Carles Enrech, ‘La formació del sistema de colònia industrial i la crisi del model fabril tèxtil dels grans vapors urbans. “Les colònies no neixen, es construeixen” (1872–1890)’, in Lluís Virós (ed.), Actes de les V jornades d’arqueologia industrial (Barcelona, 2002), pp. 201–223. An excellent micro-study is Gràcia DorelFerré, Les colònies industrials a Catalunya. El cas de la Colònia Sedó (Barcelona, 1992). 463

The Ashgate Companion to the History of Textile Workers minded (the so-called ‘English system’); secondly, though the introduction of new machinery, which would allow them to reduce the number of workers and de-skill those who remained. The first course met with staunch working-class resistance. The Güell factory tried to follow this path, but given the hostile reaction of its workers moved production to a company town it had built outside the city. Faced with similar problems Batlló simply shut down in 1888. Nevertheless, over the next 20 years industrialists would continue to make sustained efforts in this direction.34 La España Industrial, on the other hand, chose the second route, cutting the number of personnel, replacing self-acting mules by ring frames, modernizing its weaving shed and finishing section, and renewing its antiquated steam engines. The introduction of the ring frame was the key technological innovation of these years. It had a cost advantage over the self-acting mule, most especially in the lower thread counts in which much of Catalan industry specialized. Furthermore, industrialists argued that given that it was easy to operate it was more suited to female workers. Hence it was introduced by large numbers of industrialists over the next 20 years and combined with attempts to replace men by lower paid women. By 1907 60 per cent of Catalan spindles were ring frame, the highest percentage in the world behind Italy (see Table 17.2).35 These measures would be accompanied by efforts both to replace male by female power-loom weavers and to hold down the wages of the men that remained. Hence, by the 1900s in Barcelona men tended to work only on the largest machines and on jacquard looms. The overall result would be a degree of feminization of the industry, above all in the factories on the Pla. The 1920 population census put the number of men over the age of 20 working in textiles in Barcelona at just over 22 per cent, and the percentage in Catalonia as a whole at just under 28 per cent.36 This reorganization proved easier for the industrialists to manage because, from the 1890s in particular, the Catalan (and especially Barcelona) economy began to diversify away from textiles and in some larger towns men in any case began to migrate into work in somewhat more highly paid areas, most notably in the metal, chemical and construction industries.37 Women industrial workers, however, enjoyed no such alternatives. It has been argued that as part of this feminization process women also tended to stay on longer in the factory (until their second or third child was born) in a context in which birth rates were falling and compulsory schooling was being more effectively policed.38 34 For these attempted reforms and the conflicts they brought, see Enrech, El Pla contra La Muntanya, pp. 160–186, and ‘La reforma de la organización de trabajo en La España Industrial a finales del siglo XIX’, Sociología del Trabajo 29 (Madrid, 1997): 135–155. 35 Angel Smith, ‘Social Conflict and Trade-union Organization in the Catalan Cotton Textile Industry, 1890–1914’, International Review of Social History 36 (1991): 331–376, at pp. 337–338. 36 Censo de población de 1920, vol. 5, pp. 30–253. 37 Borderías, ‘Women’s Work’, p. 145. 38 Enriqueta Camps-Cura, ‘Transitions in Women’s and Children’s Work Patterns and Implications for the Study of Family Income and Household Structure: A Case Study from the Catalan Cotton Textile Industry’, The History of the Family 3 (2) (1998): 464

Spain In the 1880s the reaction of the TCV was to try to negotiate common tariffs of wages and working hours throughout Catalonia in order to reduce the pressure on the Pla industrialists. In addition, it accentuated its class collaborationism, forming a common front with employers for further increases in tariff protection for the industry. However, the resistance of the river-bank industrialists would lead to the failure of negotiations and a period of intense industrial conflict. At its heart was the employers’ determination to break the unions in order to cut costs and the male weavers and spinners’ attempts to defend their working conditions and what was still in many respects their informal artisan status. On the Pla, of symbolic importance was the defeat of the workers of La España Industrial in 1887, thereby opening the way to the reforms of 1888 and 1889. More important, however, were events between 1890 and 1891. The river-bank industrialists had not been slow to follow their colleagues on the Pla and attempt to reduce costs, culminating in mass lockouts in the Muntanya. This confirmed the industrialists’ hostility towards trade unionism. They had only reluctantly entered into collective bargaining with the TCV when times were good, and once the bottom fell out of the industry were quick to attack. They were emboldened because the Spanish state failed to play a positive role in encouraging the consolidation of stable union-based labour relations. The Conservative Party in particular contained a vociferous lobby of Catalan industrialists and sided with them. This was only to change from the turn of the century when the Restoration regime began to pay more attention to Western European models designed to bring social stability through stable collective bargaining. Yet, the problem would then be that more reforming governments would not be strong enough to impose themselves on recalcitrant and hard-line Catalan industrial interests.39 On the river banks the resistance to the 1890–1891 employer offensive was centred on the workers from the Llobregat and Ter industrial towns. It is important to note that despite the fact that the TCV was male dominated and that the strikes were, in part at least, about defending men’s status and jobs, female textile workers participated enthusiastically. This indicates that in an age in which modern feminism had yet to make its mark women, who it should not be forgotten were often family members, by and large accepted their subordinate role within the world of work. It was an attitude no doubt reinforced because once all the players had accepted the cultural ground rule that women should be paid less than men then the replacement of male workers would lead to a devastating decline in family income. The result was an extremely high level of worker solidarity in the textile 137–153; Montserrat Llonch, ‘La feminizació del treball tèxtil a Catalunya (1891–1959)’, in Montserrat Llonch (ed.), Treball tèxtil a la Catalunya contemporània (Lleida, 2004), pp. 77–94, at pp. 77–83. 39 Relations between Catalan industrialists and the Spanish State are analysed in Borja de Riquer, ‘Burgesos i caçics a la Catalunya de la Restauració’, L’Avenç 85 (1985): 16–33; Soledad Bengoechea, Organització patronal i conflictivitat social a Catalunya (Barcelona, 1994); Angel Smith, ‘The Catalan Counter-revolutionary Coalition and the Primo de Rivera Coup, 1917–1923’, Europe History Quarterly 37 (1) (2007): 7–34. 465

The Ashgate Companion to the History of Textile Workers towns. And, in these circumstances, in the larger Ter Valley towns the industrialists were forced to desist in their attempt to employ women on the new ring frames. Nevertheless, over much of Catalonia the TCV was totally undermined, making it far easier to push through the attack on working conditions.40 The offensive was accompanied by intense efforts by the industrialists to impose social control through a mix of strict labour discipline and paternalism, most notably in the Muntanya where the company towns were expanded and consolidated. The opportunity was provided by a new economic boom which followed the protectionist tariffs of 1891, which further protected the home market and gave Spanish textiles a monopoly in the Cuban colonial market (allowing exports to rise from 3 per cent of total Spanish cotton textile production in 1885 to 20 per cent in the peak year of 1897). With one or two exceptions the biggest company towns were set up on the Llobregat River, where relatively large mixed weaving-spinning factories predominated. New housing was built, walls were put up, neo-gothic churches and sumptuous mansions were erected (emphasizing the loci of power), and schools, shops and entertainment provided. These towns were drenched in conservative Catholicism and made every effort to keep unions at bay.41 Throughout Catalonia these changes were accompanied by the rise, in conservative circles, of a discourse which portrayed the industrialist as a paternalist father figure who was owed obedience by his worker-children.42 The events of the 1890s also had far reaching political consequences. The TCV’s fall presaged a conflict-ridden social landscape and opened the door to more radical political alternatives. This became clear at the turn of the century when union organization in the industry revived in the context of the somewhat more open political climate which followed Spain’s defeat in the war of 1898. On this occasion Barcelona remained a black spot, for the labour force was now to a large extent made up of female workers who did not have the power to take on their industrialists. Rather, the new textile federation, the Federación de la Industria Textil Española (FITE), was led by male spinners and weavers from the river-bank towns who incorporated large numbers of female workers from these localities into the union. However, as in 1890–1891 union growth during 1900 was met by a series of employer lockouts, leading to an attempted general strike. On the ground class hatred reached incredible heights, as exemplified by events in Manlleu where a mob of workers burnt down a number of employers’ houses in early 1901. Within the FITE the Spanish Socialists (PSOE/UGT) were the dominant players. They did not adopt the TCV’s class collaborationist approach but did try to resurrect its organizational structure (including the employment of paid officials) and enter into collective bargaining with employers. Yet the latter’s attitude once again made 40 For these strikes and their aftermath, see Smith ‘Social Conflict’, pp. 347–349; Enrech, El Pla contra la Muntanya, pp. 132–139. 41 Enrech, ‘La Formació del sistema de colònia industrial’, pp. 204–208. 42 Jordi Sole Tura, Catalanismo y revolución burguesa, 2nd edn (Madrid, 1974), pp. 233– 263; Gary Wray McDonogh, Good Families of Barcelona: A Social History of Power in the Industrial Era (Princeton, 1986), pp. 53–57. 466

Spain this impossible. This led to widening rift within Spanish organized labour, with the final desperate decision to call a general strike in December 1900 denounced by the moderate Madrid-based leadership of the Socialist party, leading to claims by Catalan anarchists that the Socialists had betrayed the working class.43 The failure of the general strike represented the death knell of the FITE. It coincided with a growing recession in the industry as the impact of the loss of the Cuban market made itself felt. The depressed climate was prolonged over the rest of the decade by a number of poor harvests. Following the strike wave some of the biggest company towns – who despite the building programmes of the 1890s had still been seriously hit by the strikes – made great efforts to stabilize social relations by housing a relatively small number of well-treated families in the town and bringing the rest of the workforce in from the surrounding countryside (thereby cutting out, as far as possible, the troublesome town workers).44 In this climate, the indications are, workers saw little improvement in living standards from the 1890s. Indeed, data provided by the Barcelona local authorities make clear that in textiles, as in the other major industries, the majority of male workers failed to earn a ‘family wage’ (enough to keep a family with two children). There was no equivalent to the situation in Europe’s major industrial countries, where the period between the 1860/70s and 1900s had seen large improvements in working-class living standards. This meant that only a few of the highest paid workers – most notably foremen – could try to ape middle-class values and prevent their wives from working.45 Unions were only effectively to revive from 1910 as the economic picture became rosier. Prior to the First World War the most remarkable aspect of this reorganization was the key role women workers in Barcelona were to play, leading a strike for a 50-hour week in July 1913. Several factors were involved here. Though the unions were led by men (with, in Barcelona, journeymen from the finishing trades playing an increasingly central role) they remained very much aware of the need to mobilize women in order to strengthen union organization as a whole. Furthermore, the hope still existed in some circles that women might be replaced by men, and it was felt that this would be easier if they were unionized and their wages increased (apart from the fact that this served to raise working-class family income). Finally, the female workers were subject to intense exploitation; they worked an 11-hour day (compared to between 8 and 10 hours in most crafts) and were increasingly forced to handle two ring frames or between three and four looms. Those who were married had to try somehow to juggle this burden with child-rearing and housework. When the new union, called La Constància, struck for a 10-hour day the strike spread through much of the Pla. It was to last for almost two months and at first 43 For this period see Smith, ‘Social Conflict’, pp. 363–365. 44 Dorel-Ferré, Colònies industrials, pp. 404–412. 45 For more details, see Angel Smith, Anarchism, Revolution and Reaction: Catalan Labour and the Crisis of the Spanish State, 1898–1923 (New York and Oxford, 2007), pp. 36–40, 47–49. 467

The Ashgate Companion to the History of Textile Workers proved a considerable success, with the Government publishing a Royal Decree establishing a 60-hour week/3,000-hour year in the industry. Women proved every bit as militant as their men folk, contradicting stereotypes which portrayed them as meek and submissive. But subsequent events did not bode well for the future. The Government, which had moved under pressure from the strike, then backtracked, agreeing that the industrialists could operate a 62-hour week or cut the number of workers’ holidays (because of a number of unofficial holidays that were celebrated), and did not enforce its decree in the river-bank factories. This was typical of the Restoration regime’s inability to set a course and stick to it and emphasized its weakness when it came to dealing with industrial interests. Nevertheless, outraged that the state had interfered at all, many cotton industrialists founded a new hard-line Catalan employers’ federation. As for the women textile workers, disappointed at the final outcome of the strike most were to leave the union.46 The war years were to bring elements of radical change and continuity. Spain did not enter the First World War and cotton industrialists benefited enormously from the export opportunities it provided. At the same time, with the rapid substitution of steam power for electricity the Pla manufacturers found themselves in a stronger competitive position, and began to focus more on high-quality garments, like corduroys and velvet, which actually required more skilled labour. This was reinforced by the government decree of January 1919 establishing an 8-hour day throughout Catalan industry, which the workers in the northern river-bank towns were now better able to enforce. The result was an inversion of the tendency for the industry to shift to the northern river-bank factories (see Table 17.4).47 Yet this economic boom did not lead to any stabilization of labour relations in the industry. On the one hand, it provoked an inflationary spiral as industrialists took advantage of export opportunities and left the home market denuded. This prompted workers to organize and declare strikes just in order to defend their, already meagre, real wages. On the other hand, the tight labour market allowed the rapid expansion of the anarcho-syndicalist labour confederation, the Confederación Nacional del Trabajo (CNT). Catalan conditions, characterized by at times fierce social conflict, lack of institutionalized collective bargaining and generally ineffective state intervention, had favoured the growth of anarchism, and in 1918/19 the CNT was able to mobilize much of Barcelona’s industrial working class on the back of local industry-wide union federations (Sindicats Unics). In the city, within textiles the male-dominated finishing trades became the fulcrum around which union organization revolved. Unfortunately we have no social history of this group of workers, but the indications are that this occurred in a context in which (like many other groups of journeymen) they were under pressure from employer attempts to intensify production and manage the labour process 46 For female textile workers, see also Carles Enrech, ‘Conflictivitat, gènere i racionalització dels temps de treball (1891–1919)’, in Llonch (ed.), Treball tèxtil, pp. 95–112, at pp. 105– 109. 47 Jaume Marquès Mir, Història de l’organització sindical tèxtil ‘El Radium’ (Barcelona, 1989), p. 24. 468

Spain more closely. In 1915 they were joined by a new group of workers, the weaving warp dressers (contramaestres), who changed the cotton pieces and repaired the machinery. Most worrying for industrialists was that this was a group of relatively well-paid workers from whom the foremen were largely drawn.48 From late 1918 La Constància, under the control of the workers from the finishing trades, was able to mobilize most of the female textile workers in the city. Most significantly, for the first time the union was able effectively to organize the workers of the biggest textile factories, which were largely situated in Sants. During 1919, as inflation eased off, the boot was now on the other foot. Though industrialists were able to maintain social control in the company towns, in urban areas, a general strike to cut working hours, launched by a new Catalan textile federation – founded in 1913 – was widely seconded. And in Barcelona, under pressure industrialists signed a collective agreement in September conceding substantial wage increases.49 This was symptomatic of the new conjuncture. Between 1919 and 1920 for the first time since the ‘gold fever’ period, throughout Catalonia textile workers were able to achieve very substantial improvements in real wages, more than making up for the falls of the war years.50 This process by which the defences of the largest factories were breached was apparent in a number of industries and constituted a key feature of these years. Industrialists reacted by forging an alliance with the military (centred on Barcelona), which used expedite measures to break the back of the CNT between 1920 and 1922, and, more broadly, played a key role in undermining the Restoration regime – which was seen as being insufficiently protective of employer interests and too soft on the CNT – and ensuring its replacement by the Primo de Rivera Dictatorship in September 1923. In this climate social conflict took an especially violent turn, with CNT gunmen (a significant percentage of whom were drawn from the textile finishing trades and the warp dressers) targeting employers and foremen, and the militarized local authorities striking back though mass repression.51 Spain emerged from seven years of dictatorship in 1931 with the establishment of the Second Republic. In fact the dictatorship had proved relatively benign, and while the CNT was banned arbitration boards were established on which the Socialists were allowed to sit. With respect to working conditions, the dictatorship saw a high degree of stability, with the favourable economic climate of the mid1920s (boosted by even higher protectionist tariffs and heavy state expenditure) counterbalanced by the workers’ weaker bargaining position. While industry was hit by the Great Depression of 1929 the damage was limited because most production was aimed at the home market. In subsequent years it was buoyed up 48 Ibid., pp. 27–29, 41ff. 49 Ibid., p. 55. 50 Carles Enrech, Industria i ofici, pp. 388ff; Raimon Soler i Becerro, ‘La evolución del salario en una empresa textil algodonera. La fábrica de la Rambla de Vilanova i la Geltrú (1891–1925)’, Revista de Historia Económica 15 (2) (1997): 400–411. 51 For more details, see Smith, ‘The Catalan Counter-revolutionary Coalition’. 469

The Ashgate Companion to the History of Textile Workers by good harvests and a general rise in purchasing power. The years between 1931 and 1933 were also marked by a rapid union reorganization, with, in Barcelona, workers in the finishing trades and warp dressers once again leading strikes for improved working conditions, resulting in additional if unspectacular rises in real wages. However, by 1935 textiles was being increasingly affected (rather later than most industries) by the depressed international climate, leading to a significant growth in unemployment. The lot of some workers at least was improved in other ways too. In Barcelona from the 1920s in working-class areas birth rates began to outstrip death rates for the first time, and higher wages allowed families with a regular income to participate in the first stirrings of consumer culture.52 There was no major restructuring of the cotton textile industry during the 1930s, but important changes in both the composition of the working class and the political context. From 1918 there was a massive increase in internal migration to Barcelona, with many new migrants finding low-skilled industrial work. The institutional framework was much more favourable to labour, though both the new Spanish Republican authorities and the Catalan autonomous Government distinguished carefully between anarchist and non-anarchist unions, penalizing the former and favouring the latter. Finally, in Catalonia the labour movement became both more diverse and more divided. The anarchist CNT/FAI launched an all out offensive against the Republic, while more moderate syndicalist-dissidents, Socialists and eventually Communists were more accommodating. The argument has been put forward that it was mainly skilled workers who turned their back on the ‘direct action’ tactics of the CNT, integrated into the arbitration boards and showed some sympathy for the new Republican institutions. On the other hand, the poorer, less skilled working class, which enjoyed none of the relative prosperity of the Republican years, remained more closely wedded to anarchism.53 A serious social history of Barcelona textiles during the Second Republic would certainly allow us to test such hypotheses more effectively. Unfortunately, none has been written. We know that in Barcelona the CNT remained strongest in major industrial sectors in which there was a high degree of industrial conflict. This included cotton textiles. On the other hand, workers in the textile towns outside Barcelona tended to gravitate towards other political options. This was not a new phenomenon. From the late nineteenth century in the Ter Valley, Manresa and the hosiery centre of Mataró, Socialists had built up a considerable following. One cannot rule out the operation of chance in such events, but there seem also to have been some more concrete factors. In these towns, workers linked to the Socialists 52 Albert Balcells, Crisis económica y agitación social en Cataluña (Barcelona, 1971); Carles Sudrià, ‘Una societat plenament industrial’, in Història econòmica de la Catalunya contemporània (Barcelona, 1989), vol. 4, pp. 13–265, at pp. 32–34; Esteve Deu i Baigual, ‘L’esgotament del model del segle XIX, 1914–1939’, in Història econòmica de la Catalunya contemporània (Barcelona, 1989), vol. 6, pp. 13–42, at p. 23; Nick Rider, ‘The New City and the Anarchist Movement in the 1930s’, in Smith, Red Barcelona, pp. 66–87. 53 Rider, ‘New City’; Christopher Ealham, ‘“Outcast Barcelona” Anarchism and Illegality, 1931–1937’, Contemporary European History 4 (1995): 133–151. 470

Spain and republicans were able to make an impact in local elections and go some way to ensuring social legislation was respected. While it would be an exaggeration to talk of the operation of ‘municipal socialism’ in these areas, the workers’ successes would give anarchist ‘antipoliticism’ less legitimacy.54 The Civil War brought the collectivization of all Catalan industry. This resulted in important changes in the process of production. On the one hand, the tendency for the differential between the wages of skilled male and more unskilled female workers to grow was checked, but, on the other, the number of wage-grades was reduced. The collectivist experiment was, however, brought to an end by the victory of the forces of General Francisco Franco in 1939, leading to the re-establishment of the old wage scales.

From dictatorship to democracy, 1939–2000 The Franco regime was a brutal dictatorship whose aim was the eradication of class-based union organization. To achieve this all ‘producers’ (both workers and employers) were integrated into a single union, the Organización Sindical Española (OSE), which was dependent on the regime’s single party, the FET y de las JONS. The effect, when combined with the new labour legislation, which until 1958 would put the fixing of wages and working conditions in the hands of the state, was to subordinate workers to the authorities and their employers. Most notably, the regime imposed a harsh, authoritarian climate within the workplace, based on hierarchy and obedience.55 This new order was theoretically designed to overcome the class struggle and achieve harmony within the world of work. It operated a policy of economic autarky, which would have disastrous consequences for the Spanish economy. Interventionism was inefficient and bureaucratic, while a combination of import restrictions and corruption produced severe shortages and a burgeoning black market.56 For the textile industries the result would be contradictory. On the one hand, autarkic restrictions, along with the favourable treatment given to heavy industry and production for the military, led to shortages of raw materials and electricity supplies. On the other, in textiles (as in other industries) the level of state protection meant that the industrialists monopolized the home market. This, together with low real wages and the subordination of the workforce, ensured high

54 For which, see Smith, Anarchism, Revolution and Reaction, pp. 136–137. 55 Carme Molinero and Pere Ysàs, ‘Patria, Justicia y Pan’. Nivell de vida i condicions de treball a Catalunya, 1939–1951 (Barcelona, 1985), pp. 25–91; José Babiano, Paternalismo industrial y disciplina fabril en España (1938–1958) (Madrid, 1998). 56 José Luis García Delgado, ‘La industrialización y el desarrollo económico de España durante el franquismo’, in Jordi Nadal, Albert Carreras and Carles Sudrià (eds), La economía española en el siglo XX. Una perspectiva histórica (Barcelona, 1987), pp. 164–189. 471

The Ashgate Companion to the History of Textile Workers profits, a fact which was to lead to the growth of textile manufacture in other parts of Spain, most notably in Valencia. It was an economic model which had only a limited lifespan. Growth during the 1940s and 1950s was slow. Despite low wages, costs were high in a context in which the internal market remained very small, raw materials were expensive, factories modest in size and antiquated machinery predominated. In the cotton industry, for example, in 1952 25 per cent of the machinery was over half a century old, and only 30 per cent was less than 25 years old.57 And with an average of 135 workers per spinning establishment and 60 per weaving shed in 1942 factory size was about the same as it had been in 1919.58 When under both internal and external pressure the regime took the decision to liberalize the economy in the 1950s a thoroughgoing restructuring became inevitable. From 1959 plans were prepared to raise productivity and prepare for a hypothetical entry into the Common Market. Using the British model the authorities aimed to subsidize the substitution of older machinery. Yet the results were contradictory, for while larger factories benefited considerably, many smaller employers did not take advantage of the scheme because the credits were not generous enough.59 The industries would, at the same time, undergo other important changes. At first cotton textiles remained predominant, but the its weight within the sector declined from the 1950s as competition from developing countries intensified and new artificial fibres came onto the market (for cotton textiles see Table 17.5). By 1973 it employed 43 per cent of the country’s textile workers (excluding the finishing trades).60 These restructuring plans continued into the 1970s, but as the international economic crisis bit their objective changed from the promotion of modernization to the removal of excess capacity. This involved subsidies to close factories, destroy machinery and provide redundant workers with compensation and higher rates of unemployment benefit than was the rule. Workers, it should be noted, were better treated in the context of the great mobilization that accompanied the fall of Francoism and the transition to democracy. The result was a downward readjustment of productive capacity, together with an important injection of financial resources for those companies that participated in the scheme, allowing for significant technological modernization. This, together with continued protectionist measures and export subsidies, explains why the Spanish trade balance in textile manufacture was still in surplus in the first half of the 1980s. Spain’s final integration into the European Community in 1986 changed all this. Subsequently, competition both 57 Carme Molinero and Pere Ysàs, Els industrials catalans durant el franquisme (Vic, 1992), p. 22. 58 Flórez Beltrán, La industria algodonera española, pp. 149–171. 59 For an overview of the 1940s and 1950s, see Muriel Casals i Couturier, ‘“Cupos”, recuperació i perdua de protagonisme, 1939–1985’, in Història Econòmica de la Catalunya Contemporània, vol. VI (Barcelona, 1989), pp. 43–80, at pp. 52–54. 60 Direcció General d’Ocupació, Monografies sectorials sobre l’atur no subvencionat a Cataluña. Sector textil (Barcelona, 1986), p. 25. 472

Spain from within and outside the Community put enormous strain on these industries and called their survival into question. The situation only stabilized in the mid1990s when they specialized in the higher quality end, which was less affected by competition from the developing world.61 Table 17.5 Development of the Spanish cotton textile industry, 1942–1979 Spindles

Looms

Workers

1942

2,039,000

76,698

161,748

1954

2,363,000

66,456

n.a.

1960

3,611,000

67,326

131,147

1970

2,175,000

53,088

96,196

1979

2,014,000

26,327

58,372

Sources: Flórez Beltrán, La industria algodonera española, p. 171; Joaquin Maluquer, La política algodonera 1940–1970, Servicio de Estudios de Banca Catalana (Barcelona, 1973), pp. 61–63; Direcció General d’Ocupació, Monografies sectorials sobre l’atur no subvencionat a Cataluña. Sector tèxtil (Barcelona, 1979), p. 16.

Between the 1940s and 1960s the percentage of the active industrial workforce within the textile industries remained high. The 343,169 workers employed in the industries in 1958 represented – including mining and transport – a little over 10 per cent of Spain’s industrial working class (as against just under 13 per cent in 1920).62 However, its gender and ethnic composition changed significantly over time. At the outset of the regime, labour shortages of men provoked by the Civil War meant that the percentage of adult female workers was higher than ever – 75 per cent in cotton textile in 1942.63 From the 1950s the flood of Spanish migrant workers into Catalonia – first seen, as we have noted from the end of the First World War – was resumed and intensified. The impact on cotton textiles was dramatic. Migrants from Valencia and Aragón above all, began to find employment in factories on the Ter and Llobregat from the turn of the century, but their penetration was, as we have already noted, hindered by the operation of family networks in the textile towns. This, it seems, began to break down from the 1950s as social mobility among Catalan workers accelerated. Most notably, large numbers of Catalan women were 61 Asociación de Colectividades Textiles Europeas, El sector textil. Una aproximación territorial (Terrassa, 1998), p. 79. 62 Jordi Nadal (ed.), Atlas de la industrialización de España, 1750–2000 (Barcelona, 2003), p. 338. 63 Flórez Beltrán, Industria algodonera, p. 171. 473

The Ashgate Companion to the History of Textile Workers now able to find employment in the service sector. The overall result was a rapid influx of southern Spanish migrants, mostly women, into the textile factories.64 This situation was to change with the crisis of the mid-1970s. There would be a large contraction of the workforce over the next decade (see Table 17.6), which would have an especially pronounced impact in Catalonia itself for two reasons. Firstly, state-owned companies set up in other parts of Spain did not reduce the size of the workforce as quickly in order to guard against the possible social consequences in areas in which there was limited alternative employment. Secondly, higher wages in Catalonia made it difficult to compete in a context of contracting demand. This contraction would continue throughout the 1980s, with a decline of 28 per cent in the total number of Spanish workers engaged in textiles and clothing between 1988 and 1995.65 Table 17.6 Active workforce in the Spanish textile industries, 1958–1997 1958 Catalonia Spain

1962

1973

1975

1977

1979

1981

1990

n.a. 232,344 195,671 172,337 157,044 138,821 124,433 116,741 343,169

1997 n.a.

n.a. 275,095 249,395 237,931 216,113 195,687 273,802 229,735

Note: Data referring to the whole of Spain for 1990 and 1997 refers to textiles and clothing. Source: Direcció General d’Ocupació, Monografies sectorials sobre l’atur no subvencionat a Cataluña. Sector tèxtil, Generalitat de Cataluña (Barcelona, 1986), p. 7; Nadal (ed.), Atlas de la industrialización de España, pp. 458–459.

The unemployment provoked by restructuring above all affected women, who were the first to be dismissed. Several elements explain this. Firstly, it should be seen in the context of the regime’s patriarchal discourse, which emphasized that a woman’s place was in the home. Such a discourse then extended to the business sector. Secondly, in times of crisis firms tended to keep on more highly skilled male workers, with whom a closer, paternalist relationship had often been established. Hence, within the Catalan cotton textile industry while in 1973 56.7 per cent of adult workers were women, in 1980 48.9 per cent of such workers were women. Women also suffered from what might be termed sub-employment. This was the result of the putting-out of work to small workshops as the economic recession intensified in the late 1970s. From the 1980s in particular, clandestine workshops were a boom area, with labour legislation ignored, work paid by the piece and working hours 64 Some data is to be found in Elizabet Oliver Frauca, ‘El treball feminí a la indústria tèxtil de Sabadell, 1939–1960: La Marcet SA’, in Dones i treball tèxtil: Sabadell, 1900–1960 (Sabadell, 1999), p. 100. 65 Asociación de Colectividades Textiles Europeas, El sector textil, p. 82. It should be noted that we have no data for child labour for this period. 474

Spain often extremely long. Recent estimates put the level of production within the black economy at between 15 and 25 per cent of the total.66 Textile workers were only able to make their voice heard in the 1960s, following the regime’s decision to champion a policy of economic liberalization, which allowed Spain to take advantage of accelerating economic growth in Western Europe. By the mid-1960s growth rates were high, leading to constant conflicts in the dynamic sectors of metal, mining, construction and the chemical industry, in which workers were able to bring considerable pressure to bear. In textiles labour protest first raised its head in 1966, but it was not until 1970 that it became quantitatively important (representing 10 per cent of the total number of conflicts). The reason was that the new restructuring plan, which was being pushed through against a backdrop of an economic boom, was radicalizing attitudes (see Table 17.7). Table 17.7 Conflicts in the Spanish textile industries, 1966–1976 Conflicts

(1970 = 100)

Workers

(1970 = 100)

Hours lost

(1970 = 100)

1966

9

100

2,426

100

42,862

100

1967

3

33

737

30

3,587

8

1968

7

78

1,691

70

1,800

4

1969

9

100

1,643

68

357

1

1970

109

1,211

21,201

874

313,800

732

1971

40

444

11,323

467

147,700

345

1972

68

756

11,941

492

n.a.

n.a.

1973

51

567

30,336

1,250

853,847

1,992

1974

66

733

24,163

996

381,367

890

1975

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

1976

64

711

135,666

5,592

4,296,319

10,024

Source: Carme Molinero and Pere Ysàs, Productores disciplinados y minorías subversivas.Clase obrera y conflictividad laboral en la España franquista (Madrid, 1998), p. 126.

As Table 17.7 shows, 1976 was a year of particularly intense conflict, which was stimulated both by the expectations generated by the transition to democracy and by the workers’ fear that they would bear the cost of the economic recession. Strikes would peak in 1979, but by then they were much more defensive, often fought 66 Ibid., p. 480. 475

The Ashgate Companion to the History of Textile Workers simply to prevent redundancies, even if this meant accepting worse working conditions. From the 1980s, in a context of the proliferation of small workshops and with employment precarious, collective action rapidly tailed off. The growing labour conflict of the 1960s stimulated the rise of Comisiones Obreras (CC.OO), a new form of labour organization which was to adapt to the conditions under Franco by taking advantage of the legal opportunities to penetrate the OSE and also to operate clandestinely outside its boundaries. It became linked to the Communist Party but also inherited the mobilizing tradition of the CNT. From the mid-1960s it was at the heart of labour opposition to the regime, and during the transition would lobby for the formation of a single labour confederation, which would incorporate the various political tendencies within its ranks. However, the transition would see the revival of the Socialist labour confederation (UGT), linked to the growing political prominence of the Socialist Party (PSOE), and it successfully defended a pluralist trade-union model. The new democratic institutions established a system of ‘union elections’ organized at plant level, through which representatives on a factory committee were elected (giving workers the right to vote even if they were not union members). Within the textile industries the two major union confederations would, between 1978 and 1986, take an increasing share of the vote and prove quite evenly matched. In more recent years, despite the crisis of the Left and the fact that the process of ‘globalization’ has weakened the negotiating position of the workers, unions have remained an important instrument in defending workers’ interests.67

67 Maluquer, La política algodonera 1940–1970, pp. 61–63; Direcció General d’Ocupació, Monografies sectorials sobre l’atur no subvencionat a Cataluña. Sector tèxtil (Barcelona, 1979), p. 16. 476

18 The Ottoman Empire, 1650–1922 Donald Quataert

Vast numbers of Ottoman subjects worked in either subsistence or commercial textile production, drawing on the ready availability of cotton, wool, silk and flax to make goods both for their own use and for the market. There is little doubt that most rural dwellers spun, wove or embroidered and this likely also was true of urban families. The number of persons spinning and weaving for personal consumption declined with the rise of imported textiles after 1800 (see below) while market production likely increased during this same period. A few crude statistics give some notion of the importance of commercially made textiles in the overall economy at the very end of the Ottoman period. On the eve of the First World War, workers making textiles for the marketplace accounted for about one-half of the enumerated labour force in the total manufacturing sector. Manufacturing overall contributed about 10 per cent of ‘national income’ while agriculture, according to these very incomplete surveys, reportedly contributed more than 50 per cent. This former figure surely is a gross underreporting since the survey missed many urban locations of manufacture and most of the rural-based industry. Despite their lacunae, these late surveys reinforce the impression of textile production as a very large manufacturing sector – rivalling that of food processing – that employed many hundreds of thousands of workers in rural and urban areas alike. Further, I would hazard a guess that per capita Ottoman consumption of textiles rose in the period 1650–1922, thanks to new technologies that drove down their costs while incomes increased due to mounting sales of agricultural goods.

 Vedat Eldem, Osmanlı İmparatorluğu’nun İktisadi Şartları hakkında bir Tetkik (Istanbul, 1970), p. 208. The figures presented in these surveys understate the reality and are not consistent with those presented in this chapter. The Ottoman state carried out surveys of the economy and of industry in the early twentieth century. These were quite incomplete and are quoted here to offer some overall notion of textile’s importance.  Eldem, Osmanlı İmparatorluğu’nun İktisadi Şartları, pp. 224–226.

Map 18.1 Textile centres in the Ottoman Empire, c.1650–1914

The Ottoman Empire, 1650–1922

The decline of Ottoman manufacturing? The causes, formation, volume and chronology of Ottoman textile production and of domestic and foreign demand are not certain but quite old. Already in medieval times, the rich textiles of many Middle Eastern towns and cities were sought eagerly, both locally and abroad, by caliphs, sultans, bishops and kings, as well as lesser notables. European demand for labour-intensive goods such as the heavy silk brocades of Bursa and the gold threads, silk and gold weaves of Aleppo is well known. For centuries, surely until at least about 1750, these and some other Middle East textile products – ranging from ultra-luxury high end fabrics to simple cotton yarns – were avidly sought after by consumers outside of the empire. But, during the seventeenth and eighteenth centuries, textile producers from the Middle East, now under Ottoman domination, lost their place, both in the European economy and in the European imagination, as the age-old providers of these rich, heavy brocades and other silver-gold fabrics. These losses in the export trade have often been noted in the historical literature and their passing is incontrovertible. The decline of Ottoman manufacturing, however, is quite another matter and has been the subject of some debate. The older paradigm portrayed the manufacturing sector as one in steady, irretrievable decline, unable to compete with the growing might of an industrializing Europe. More recently, however, in common with textile production in many other areas of the globe such as India and South East Asia, the fate of Ottoman textile manufacturing has been understood differently. The question nowadays is concerned less with whether or not Ottoman textile producers adopted steam-powered, factory-based technologies (which they did only belatedly and modestly) and more with the other steps that they actually did take to compete for international and domestic markets. When the inquiry about Ottoman textile producers is phrased in this fashion, a rather different picture than one of failure emerges. In the new age of European hegemony, Ottoman textile makers quite often successfully adapted and survived with a combination of strategies. Thus, new textile styles and fabric mixes appeared in many regions of this large empire, seeking to create niche markets difficult for competitors to penetrate. Also, cheaper labour sources were found. Male labour, often guild-organized and urban based,

 For an extensive discussion of the mythical decline of nineteenth century Aleppo textile production, see Donald Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution (Cambridge, 1993), pp. 77–79.  Charles Issawi, An Economic History of the Middle East and North Africa (New York, 1982); Şevket Pamuk, The Ottoman Empire and European Capitalism, 1820–1913 (Cambridge, 1987).  Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution; John Chalcraft, ‘The Striking Cabbies of Cairo and Other Stories: Crafts and Guilds in Egypt, 1863– 1914’ (unpublished Ph.D. thesis New York University, 2001). Also, T. Roy, Traditional Industry in the Economy of Colonial India (Cambridge, 1999) and K. Pomeranz, The Great Divergence: Europe, China, and the Making of the Modern World Economy (Princeton, 2000). 479

The Ashgate Companion to the History of Textile Workers gave way in many locales to non-guild, less expensive, female and child labour, frequently located in rural areas. And finally, new technologies appeared. These included better methods of hand-spinning, hand-weaving and cloth finishing (see below). Moreover, the early and widespread adoption of machinemade yarns and synthetic dyestuffs stands at the very centre of successful adaptations in a changing competitive environment. The shift from hand-spun to machine-spun yarns was among the most important of these survival stratagems. Between c.1810 and 1910, the volume of Ottoman imports of machine-made yarns rose by more than fiftyfold, to some 12,550 tons. The extent to which the rise of imported yarns, dyestuffs and cloths and of Ottoman machine-spuns and -weaves affected auto-consumption is not clear. Around 1900, hand-spuns still accounted for about one-quarter of all cotton yarns consumed in the empire. At that time, in many villages, spinning wheels and looms were still found in virtually every home, where family members, usually women, spun and wove. On the one hand, the dislocations at the personal level were real as many hand-spinners lost incomes once gained from commercial spinning. And yet, on the other, there is no doubt that the import (as well as the local manufacture) of machine-mades, particularly yarns, provided a boon to commercial production, freeing many tens of thousands of labour hours to move into more productive and remunerative tasks, such as clothmaking, instead of spinning. For the economy, the move away from hand-spinning meant a shift from less productive to more productive tasks. It meant that workers who once laboriously had spun, a labour-intensive, very un-remunerative task, now more profitably could spend their time weaving and printing these imported yarns and fabrics (or shifting over to commercial agriculture). The import of yarns thus meant vast savings for the Ottoman textile industry and an enhancement of its ability to compete. Less important among the new technologies (except in the case of silk reeling) was the formation of factories to spin yarn and weave cloth. With these stratagems, Ottoman textile producers succeeded in finding new commodities to sell to foreign consumers. These international markets included the Italian peninsula and France and the rest of Western Europe as well as central and Eastern Europe and as far as India and the United States. More significantly, Ottoman producers successfully struggled with Indian and then European competitors and retained most of their domestic consumers. Their retention of domestic markets, even in the nineteenth-century heyday of West European industrial hegemony, is the key to understanding Ottoman textile production. In sum, the post-1650 history of Ottoman textile producers is one in which they gained, lost, retained, sometimes regained and sometimes acquired new domestic consumers as well as foreign buyers. In some areas, textile production for export and perhaps for domestic consumption took a downturn during the late eighteenth century, along with the rest of the economy. Where it occurred, the decline lasted 60–80 years, and  Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, p. 40.  For example, Manoog B. Dzeron, Village of Parchang: General History (1600–1937) (Boston, 1984). 480

The Ottoman Empire, 1650–1922 then began a recovery cycle during the 1830s and 1840s. The downturn derived from various causes including state oppression and war, and the shockwaves from the new, mechanized competitors of Western Europe. As the Ottoman state fought against Russia, Napoleon and others from the 1760s through the turn of the nineteenth century, the wars imposed a crushing burden. In its need for revenues to support the war effort, the central state imposed crushing tax burdens that bankrupted some producers and prompted the flight of others to smaller towns and the countryside to escape the taxman. This sequence of events occurred in Tokat, for example, where many cloth dyers fled the exactions of the lifetime tax on their workshops in the town centre and relocated in rural enclaves and smaller nearby towns. In other areas, however, the downturn is certain but the causes less so. In Aintab, near the present-day Turkish-Syrian border, cotton textile producers had won new markets in the seventeenth century, both in Europe and within the empire, copying Indian textiles, which were winning such international favour among consumers. This industry flourished until the late 1770s but then declined, perhaps because of border disputes between the Ottoman Empire and a weakened Iranian state.10 In many regions, the entry of machine-made fabrics, mainly from England, attracted consumers away from Ottoman textiles. In the Balkan provinces, there may have been prosperity rather than any late eighteenth-century downturn. The town of Ambelakia, in present-day Greece south of Salonika, boomed in the final two decades of the eighteenth century, thanks to its exports of red cotton yarn to east and central Europe.11 There is additional, indirect evidence that the late eighteenth century indeed may have been an era of growth and prosperity for many textile producers in the Balkan provinces of the empire, such as the areas of present-day Bulgaria, Rumania and Greece.12 Moreover, the political reforms of the Ottoman state, c.1830–1870, brought new prosperity in textile production, which was destroyed, however, once these regions broke from Ottoman control during the later nineteenth century.13 Overall, the crisis of the era 1760–1840 faded. Wars ended and relative stability ensued. Ottoman textile producers recovered from the bad times and the initial



Mehmet Genç, ‘Ottoman Industry in the Eighteenth Century: General Framework, Characteristics and Main Trends’, in Donald Quataert (ed.), Ottoman Manufacturing in the Ottoman Empire and Turkey, 1500–1950 (Albany, NY, 1994), pp. 59–86, points to the role of state policies but likely exaggerates their importance.  Yüksel Duman, ‘Notables, Textiles and Copper in Ottoman Tokat, 1750–1840’ (unpublished Ph.D. thesis, Binghamton University, State University of New York, 1998). 10 Suraiya Faroqhi, Approaching Ottoman History: An Introduction to the Sources (Cambridge, 2000), p. 227. 11 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, p. 34. 12 This is my interpretation and not necessarily that of Faroqhi, Approaching Ottoman History, p. 228. Pamuk, The Ottoman Empire, p. 110, argues for an eighteenth-century decline in production. 13 Michael Palairet, The Balkan Economies, circa 1800–1914 (Cambridge, 1997). 481

The Ashgate Companion to the History of Textile Workers shock of Western competition and began adapting to the new goods. The worst was over and recovery began.

Production The extant literature offers a number of clues suggesting the importance and widespread nature of textile production during the seventeenth century. Moreover, the textile production traditions hinted at during these early centuries often persisted, to be documented more completely in later years. Following are some examples of textile production drawn from the various – European, Anatolian and Arab – provinces of the empire. These are intended to illustrate, but not exhaustively, the nature and scope of production activities over the long term, 1650–c.1922. During early Ottoman centuries, scarce documentation persuasively suggests a significant volume in certain export commodities. Famed among these are ‘Oriental carpets’, mainly from the west Anatolian region around Uşak, which were already flowing to West European consumers in impressive quantities around 1650. The volume can only be guessed at: for example, during the period 1540–1700, Dutch painters portrayed nearly 1,000 different Ottoman carpets in their paintings.14 During the later eighteenth century, carpet exports continued and then, during the early nineteenth century, began to increase sharply. By the end of the nineteenth century, middle- and working-class customers in Europe and the United States bought these rugs in near mass consumption quantities.15 Altogether, in western Anatolia, the industry in its spinning, dyeing and knotting phases employed about 60,000 workers. Technological innovation, on both the demand and the supply side, radically altered rug-making. In the first place, the boom in Western demand for these handmade Ottoman manufactures, especially evident as the nineteenth century developed, derived from European and American efforts to escape from the increasingly mechanized tedium of their everyday lives. Technology – in the form of factory-made yarns and dyes, innovations coming from the West – also changed the very production of rugs and, in fact, made possible the explosive growth of the industry at the end of the nineteenth century. Both innovations sought to circumvent the crucial problem of labour shortages which plagued Ottoman manufacturing, at least in the nineteenth century and likely earlier as well. As demand for carpets boomed, Ottoman entrepreneurs during the 1840s sought to adopt machine-made wool yarn to solve the labour problem. Technical difficulties persisted and it was not until the 1890s that yarns more suitable for carpet knotting could be machine-made. Thereafter, mechanized factories in the carpet-making 14 Faroqhi, Approaching Ottoman History, p. 153. 15 Donald Quataert, ‘Machine Breaking and the Changing Carpet Industry of Western Anatolia, 1860–1908’, Journal of Social History, Spring (1986): 473–489; Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution; Donald Quataert, ‘The Age of Reforms, 1912–1914’, in Halil Inalcık with Donald Quataert (eds), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge, 1993), pp. 761–943. 482

The Ottoman Empire, 1650–1922 regions supplied all the yarns needed.16 Synthetic and artificial dyes, beginning around 1850, offered producers the means to obtain virtually unlimited quantities of fashion colours to meet the changing tastes of Western customers and, at the same time, avoid the labour bottlenecks of naturally produced dyestuffs. Although the advantages of synthetic dyestuffs, including cheapness, made them popular, problems of misuse persisted. Issues of quality and taste maintained demand for the natural dyestuffs. Villagers in the Plovdiv area (in modern Bulgaria) were already producing a coarse wool cloth around 1550, marketing their wares in nearby small towns and villages. During the next century, these woollens grew from local specialty to interregional good.17 Coarse wool cloth production endured and evolved. In the eighteenth and nineteenth centuries, weavers also made a somewhat fine wool cloth. At this time, producers adopted new and larger spinning wheels and also made some improvements in small-scale hand-weaving technologies. Producerpeddlers then sold these two textiles in the imperial capital, all over Anatolia and in India. The quantities were vast. In 1870, perhaps the peak era, producers in the region of Bulgaria made over one million square metres of coarse and fine wool cloth. Notably, production collapsed following the withdrawal of Ottoman authority, falling by two-thirds in slightly more than a decade.18 The famed mohair cloth production around Ankara/Angora suffered ups and downs during the political instabilities of the seventeenth century. During good years of this era, 1,000 looms were working to meet foreigners’ demands for the cherished mohair cloths of the town. Around 1800, the industry reportedly employed 10,000 workers and annually exported 20,000 pieces. Permanent troubles began when imitations from Amiens eliminated the French market and, later on, when the English developed South African sources of the valued wool. By the mid-nineteenth century, there was scarcely a mohair loom left in the city.19 In this instance, an Ottoman textile industry did collapse and vanish under the impact of international competition. Both cotton and wool cloth production were flourishing at Manisa in western Anatolia during the sixteenth century.20 More generally, widespread cotton cloth manufacturing was present throughout the Aegean region in the 1670s, when dyeing was concentrated in a few towns. Producers worked both for merchants

16 Quataert, ‘Age of Reforms’. 17 Suraiya Faroqhi, ‘Crisis and Change, 1590–1699,’ in Inalcık with Quataert (eds), An Economic and Social History, pp. 413–636, at p. 457; Quataert, ‘Age of Reforms’. 18 Palairet, Balkan Economies, pp. 69–72; see tables on pp. 71 and 192. 19 Faroqhi, ‘Crisis and Change’, p. 457. Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, pp. 92 and 92, n. 60. 20 Daniel Goffman, Izmir and the Levantine World, 1550–1650 (Seattle, 1990), pp. 70–71 shows that both the woollen and cotton textile industries at Manisa were flourishing in the sixteenth century but suffered in the early seventeenth century as foreign merchants were disrupting raw material supplies. The damage certainly was not permanent: this region remains an important textile production centre down to the present. 483

The Ashgate Companion to the History of Textile Workers and on their own, selling cloth to merchants.21 At this time, some local weavers were making significant amounts of cloth for the state.22 Powerful weaving traditions continued until the end of the empire in a string of towns and villages – Denizli, Buldan, Kadıköy, Manisa – in this region. For example, 15,000 cotton cloth weavers and dyers worked at Kadıköy around 1900, now using imported British yarn rather than locally made hand-spun.23 Weaving traditions were similarly well established and long-lasting in Aleppo, one of the great Ottoman textile centres. During the 1660s, the booming market for Aleppo textiles was sufficient to attract skilled weavers from the neighbouring towns of Mardin and Diyarbakir.24 Aleppo and Damascus alternated in the topranking position among textile production centres of northern Syria throughout the Ottoman period. One estimate suggests that around 1914, there were perhaps 120,000 textile workers in Syria and Lebanon.25 In Damascus, c.1840, at least 10,000 women, children and men worked in the textile industry while, in the 1870s, there were 6,000–7,000 weavers.26 At Aleppo, machine-spun imported yarns played a vital role in helping weavers meet their competitors but 6,000 urban workers still spun by hand around 1850. Throughout the nineteenth century between 3,000 and 6,000 looms worked in Aleppo. In about 1875, the city accounted for up to three-quarters of textile production in northern Syria. Already by 1850, textile weavers in Aleppo had shifted away from luxury cloths to the lower end of the market. In about equal numbers, they made plain cotton cloths for the bottom of the market and mixed cotton-silk cloths that offered the appearance of luxury but at a much lower price than pure silk textiles. By the end of the century, only vestiges of the famed luxury silk industry remained, in the form of some 60 workers making gold and silver thread.27 Mixed cloth production also had fallen but cotton cloth weaving, serving the lowest end of the market, flourished.28 The Ottoman silk industry had deep roots and longevity, as evidenced by the Crusaders’ admiration of the silk cloths of the Levant. In about 1650, silk cloth 21 Faroqhi, ‘Crisis and Change’, p. 459. 22 Ibid. One locale in west Asia Minor and another in the southernmost Balkans together wove 82,000 pieces for the state. 23 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, p. 58. At nearby Buldan there were 1,500 looms. 24 Bruce Masters, ‘Aleppo: The Ottoman Empire’s Caravan City’, in Edhem Eldem, Daniel Goffman and Bruce Masters (eds), The Ottoman City Between East and West: Aleppo, Izmir, and Istanbul (Cambridge, 1999), pp. 17–78, at p. 38. 25 Issawi, Economic History, p. 152. 26 Sherry Vatter, ‘Militant Textile Weavers in Damascus: Waged Artisans and the Ottoman Labour Movement, 1850–1914’, in Donald Quataert and Erik Jan Zürcher (eds), Workers and the Working Class in the Ottoman Empire and the Turkish Republic, 1839–1950 (London, 1995), pp. 35–57, at pp. 36, 39. 27 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, pp. 78–79, 102–104. 28 Ibid., pp. 71–73. 484

The Ottoman Empire, 1650–1922 production was centred in the regions of Bursa and Aleppo, with important locales elsewhere in the empire. Bursa’s role as a silk cloth production and raw silk re-export centre was already set in 1400 and at the beginning of the period of interest here the town contained numbers of substantial weaving workshops, many holding between 30 and 60 looms.29 Near the end of the era, around 1914, the Ottoman silk industry employed about 400,000 workers – three-quarters of them (in about equal shares) in the Bursa region of west Anatolia and in the Lebanon. Most of these workers raised cocoons. At least 36,000 worked in silk reeling factories, and a lesser number wove cloth. Mechanized silk reeling experienced explosive growth during the 1840s and 1850s, mainly to satisfy the burgeoning demand of European factories. The centuries-old hand-reeled silk was found to be too inconsistent and unreliable for use on the mechanized looms of Europe. And so, virtually simultaneously in the major Ottoman silk regions, steam-powered reeling mills emerged. Following their establishment, however, these mills endured a topsy-turvy history, heavily damaged by silkworm diseases and competition from East Asian suppliers. Raw silk production suffered in a trough until the 1870s and then rose eightfold in the next half century. The chequered history of Bursa silk cloth production during the nineteenth century is worth recounting for the light it sheds on the myth of decline in Ottoman textile manufacturing. Between 1750 and 1810, production remained steady but then, in the subsequent two decades, reached record levels for the 1750–1850 era. Both demand and supply factors were at work: namely, the end of a long and devastating series of international wars and the adoption of an improved cloth finishing technique. Subsequently, some weavers adopted European cotton yarn and turned to weaving less costly mixed fabrics made of cotton and silk. Whatever successes they enjoyed were fully undercut by silkworm diseases, which eliminated their raw materials. Silk cloth production was devastated by the 1860s and fell 90 per cent from its peak levels of the 1810s–1820s. But then the disease was finally conquered and in the early twentieth century some entrepreneurs adopted mechanized looms. Some workshops used both mechanized and handlooms. Around 1914, weavers working on 700 looms made approximately 120,000 pieces of silk cloth, about 20 per cent more than during the peak years of the early nineteenth century.30 Ottoman subjects, it should be noted, purchased virtually the entire cloth output. Thus, it should be emphasized, the silk cloth industry of Bursa lost its foreign markets but successfully adapted and survived, serving domestic customers.

29 Halil Inalcik, ‘The Ottoman State, Economy and Society, 1300–1600’, in Inalcik with Quataert (eds), Economic and Social History, pp. 9–409, at pp. 223–229. 30 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, pp. 109–116, which also presents a comparison with other silk weaving centres. Around 1914, there were up to 1,400 looms in the Bursa area weaving linen and cotton fabrics. See ibid., p. 59. 485

The Ashgate Companion to the History of Textile Workers

Organization of production: introduction Merchants likely played the leading role in adopting the new materials and technologies that account for much of the dynamism of the textile sector and for most of the new sources of employment in Ottoman textiles. Guilds also may have been active in instances when these innovations occurred in extant industries. But when a wholly new textile-making activity emerged in an area, merchants most likely were the agents of change. Thus, merchants established new cloth and print making industries, based on the use of machine-made thread, factory-produced dyestuffs and sometimes cloth, in numerous towns and cities and in the rural countryside. These were at the centre of the textile sector’s response to international competition and to its survival. They founded many other textile branches as well. In the imperial capital, for example, a significant ready-made clothing industry had emerged around 1900, organized by merchant houses and utilizing imported cloth. Also, in an outlying district of the city, merchants organized the Ottoman ‘Irish lace’ industry that developed in the later nineteenth century, employing more than 7,000 women and girls. Merchants founded nearly all of the mechanized silk reeling factories in the Balkan, Anatolian and Arab provinces. The case of carpetmaking is slightly more complicated since guilds initially may have organized the industry; but merchants from Istanbul and later from Izmir prompted its explosive expansion, both in the home and in the workshop, during the nineteenth century. Similarly, merchants and not guilds apparently capitalized all of the mechanized spinning and weaving factories in wool and cotton.

Guild-organized production Over the centuries, guilds, merchants and unorganized labour competed for control of the workplace and of textile production. A clear overview of guilds – their origins, prevalence, structures and significance – is impeded by a lack of information. Historians’ earlier views that guilds derived primarily from state initiatives recently have yielded to arguments that stress the agency of the workers themselves.31 That is, most guilds came into existence because workers banded together in various locations to organize production and defend their common interests. Much of our information is about guilds in Istanbul, where their structure and organization may have been idiosyncratic because of its status as a mega-city 31 For the state as the central actor, see Gabriel Baer’s studies, for example his ‘The Administrative, Economic and Social Functions of the Turkish Guilds’, International Journal of Middle East Studies I (1970): 25–50. For workers’ agency, see Suraiya Faroqhi, for example her ‘Eighteenth Century Ottoman Craftsmen: Problematiques and Developing Sources’, unpublished paper, c.1997, pp. 1–36, courtesy of the author. For another study that gives workers’ agency, see Engin Deniz Akarlı, ‘Law and Order in the Marketplace: Istanbul Artisans and Shopkeepers, 1730–1840’, pp. 1–37, unpublished paper, courtesy of the author. 486

The Ottoman Empire, 1650–1922 and imperial capital. Even with these caveats, it is safe to assert that in the midseventeenth century, craft guilds were active in all three of the major regions of the empire. Thus, either they already existed prior to Ottoman rule or formed quickly thereafter.32 Guilds in various locations responded differently to change. Some guilds in early eighteenth-century Mosul, as demand for their goods increased, moved their shops away from the town centre or went outside guild structures to find cheaper raw materials and services. As early as the end of the seventeenth century, Mosuli cotton weaving guilds were complaining bitterly about rural and urban women who had taken over their yarn spinning functions.33 More generally, as local manufacturers struggled in an increasingly competitive environment during the late eighteenth and early nineteenth centuries, many Ottoman guilds became more highly structured and organized. For example, by the end of the eighteenth century, guilds appear in the records for the first time as institutional donors to pious foundations.34 Also, where they existed, many guilds became not only more formal but more restrictive than previously, an unfortunate development at the very moment when survival demanded flexibility and adaptability. Hardly surprisingly, guilds in many areas decayed or disappeared. But in some regions they hung on and were vital, even at the end of the Ottoman era. During the seventeenth and early eighteenth centuries, entry into or withdrawal from a guild was simple, apparently involving only the payment of the market taxes of that particular guild. At this time, guild structures were probably less formal and hierarchical than later on (see below). Since movement in and out was easy, this meant the ready participation of migrants into Ottoman cities. But restrictions on guild membership mounted in the later eighteenth century, as economic crisis loomed. Many guilds began building barriers to participation. Apparently in desperation, many guilds abandoned the small workshops where they had carried out their craft and gathered together in large collective workshops involving several dozen workers. And, sometimes, they formed collectives in which hundreds of artisans used common dyeing vats or other expensive equipment. There they hoped to more readily control access to the means of production and thus better protect their livelihoods, which were being threatened by war, increasing tax burdens and mounting international competition. At the centre of the new restrictive atmosphere stood the instrument of gedik – rights which could be transferred only to members of the same guild. Gedik was both the legal right to do business in a certain place and, sometimes, the rights to the tools and implements needed for work. This entity was born in the early eighteenth century but became more commonplace during the era of crisis, after about 1750. Its possession, following a long waiting period, became 32 Suraiya Faroqhi, ‘Eighteenth Century Ottoman Craftsmen’, pp. 4–5. 33 Dina Rizk Khoury, State and Provincial Society in the Ottoman Empire: Mosul, 1540–1834 (Cambridge 1997), pp. 136–138. 34 Faroqhi, ‘Eighteenth Century Ottoman Craftsmen’. Here I side with her view that impoverishment and not prosperity led to increasing restrictions. 487

The Ashgate Companion to the History of Textile Workers a prerequisite for practising a craft and remaining a guild member.35 Such gedik registration led to monopolistic guild practices during the second half of the eighteenth century. Moreover, guild masters gradually manipulated these gediks to their own benefit. The government of Sultan Mahmud II (1808–1839) supported their efforts, driven by the desire for higher tax revenues and greater political stability. And so the guild masters gained fuller control over the workshops, at the expense of journeymen and other guild members.36 Thus, in the case of at least Istanbul, the spreading use of gedik helps explain both the growing monopolistic practices of guilds after about 1750 and the increasing domination of the masters.37 Whether guilds outside of the imperial capital followed this pattern is uncertain: at Damascus, by the late 1820s, the gedik procedures and practices described above were already said to have vanished.38 Can we generalize about Ottoman guilds since generalizations after all too often lead to stereotypes? The following sweeping statements might be true for most guilds during the period 1750–1900. As stated, guilds in Istanbul likely possessed more formal structures than previously but the applicability of this statement elsewhere is not certain. Some guilds possessed an organizational hierarchy of master, journeyman and apprentice. Stewards, normally elected by the masters but sometimes appointed by the government, represented the guild and mediated among members. Some guilds provided financial support to sick members and to widows. Guilds were organized according to the crafts pursued by the members and subdivisions also existed according to location. In very large cities there might be a number of guilds making the same commodity while in smaller communities a single guild sufficed. Their efforts to control access to the craft and to the needed raw and semi-processed materials varied considerably. Many, probably most, guilds contained members of more than one religious denomination – Muslim, Christian, Jewish. To obtain maximum effectiveness, coreligionists within a guild sometimes acted together to the exclusion of others while on other occasions the entire group interdenominationally mobilized as a unit. That is, basic economic and vocational concerns rather than religion were crucial in shaping intra- and intergroup relations. Yet another caution is in order: the actual prevalence of textile-producing (and other) guilds in the Ottoman world is not at all clear. They were readily visible in major cities and towns including Istanbul, Aleppo, Damascus, Cairo, Belgrade, 35 The discussion of guilds in the eighteenth and early nineteenth centuries is derived from Faroqhi, ‘Eighteenth Century Ottoman Craftsmen’, but with a slightly different interpretation. 36 After Engin Deniz Akarlı, ‘Gedik: Implements, Mastership, Shop Usufruct and Monopoly among Istanbul Artisans, 1750–1850’, Wissenschaftskolleg-Jahrbuch (1985–1986): 225–231. 37 By contrast, in nineteenth-century Cairo journeymen worked with the state and curbed the power of the masters. Chalcraft, ‘Striking Cabbies’. 38 Sherry Vatter, ‘Militant Journeymen in Nineteenth-century Damascus: Implications for the Middle Eastern Labor History Agenda’, in Zachary Lockman (ed.), Workers and Working Classes in the Middle East (Albany, NY, 1994), p. 42. 488

The Ottoman Empire, 1650–1922 Sivas and Amasya. Whether or not they existed at all in most urban locales is uncertain. It is unlikely that rural workers operated in guilds and certain that women were not members. Near the end of the Ottoman period, many textile guilds were in extreme disarray because they had failed to effectively counter domestic and foreign competitors. In some locales such as Aleppo, the textile guilds seemed rather vibrant in the early twentieth century but were under merchant domination. Similarly, in early nineteenth century Mosul, textile production was expanding but most artisans were becoming poorer as wealth increasingly concentrated in the hands of a few powerful families.39 Finally, the nature of the guilds where they existed is generally not clear. This is true both during particular moments as well as in their evolution over time. Guilds in some locales were the controlling body, determining members’ access to raw materials and the consumer.40 In other locations merchants determined such access for guild members and employed them as wage or pieceworkers.41 And in yet other urban areas guilds became (or perhaps always were) little more than informal bodies of individuals in the same business.42

Merchant-organized production Putting-out textile networks are visible in every era and in every region – Balkan, Anatolian and Arab – and coexisted both with guild-controlled production and merchant-dominated systems of direct supervision. Putting-out networks were already in the sixteenth century providing some of the west Anatolian carpets from the region around the production centre of Uşak, where guilds and (later) direct merchant supervision also played a role. During the seventeenth century (and quite possibly before), merchants in a putting-out network of undetermined scope moved raw cotton, yarns, undyed and then dyed cloths from worker to worker in various Aegean-area locations of western Anatolia.43 Merchants in the sixteenth century were already buying undyed cloth from weavers in Denizli, a textile centre in twenty-first-century Turkey, for dyeing in Tire, an Aegean town located near Izmir.44 Similar networks functioned at that time in Thrace, in the southern Balkans. Simultaneously, weavers in the Aegean region worked on their own account, selling their output to merchants.45 39 Khoury, State and Provincial Society, pp. 140–141. 40 Akarli, ‘Law and Order’, discussing Istanbul. 41 Abraham Marcus, The Middle East on the Eve of Modernity: Aleppo in the Eighteenth Century (New York, 1989), pp. 162–168 and Peter F. Sugar, Southeastern Europe under Ottoman Rule, 1354–1804 (Seattle, 1977), pp. 81–82 for the mid-eighteenth-century Balkans. 42 Cengiz Kırlı, ‘The Struggle over Space: Coffeehouses of Ottoman Istanbul, 1780–1845’ (unpublished Ph.D. thesis, Binghamton University, State University of New York, 2000). 43 Faroqhi, ‘Crisis and Change’, pp. 458–459. 44 Ibid., pp. 516–517. 45 Ibid., pp. 458–459, 490. 489

The Ashgate Companion to the History of Textile Workers In the eighteenth century, town–country integration in the Balkans was increasing thanks to merchant activity; in textiles and other commodities, puttingout networks were widespread at the village level, notably in the hill districts of present-day Bulgaria.46 During the later eighteenth century, in the Arab provinces around Aleppo merchants bypassed guilds and organized putting-out networks for textile production (but see below for direct merchant supervision in nearby areas).47 In the 1830s, merchants in Kayseri were coordinating a network that embraced much of central Anatolia, from the Mediterranean to the Black Sea. They took raw cotton from the Adana region in the south-east and provided it to spinners in a cluster of north – Zile, Vezir Köprü, Merzifon – and south – Bor – Anatolian towns. They then turned over the yarn to weavers in these same locations or in big urban centres such as Bursa, or to weavers across the Black Sea in the Crimea. By the 1860s, however, this network had succumbed to competition from English machine-spun yarn.48 The death of this putting-out network (and others) at the hands of yarn imports, however, did not spell the end of Ottoman textile production. Rather, other forms of organization emerged. In some locations formerly in the Kayseri network, textile production using imported yarns exceeded that when hand-spuns were used. In Bor, for example, over 2,000 looms were working in 1908, perhaps twice as many as in the 1830s.49 At other locations, new textile industries were born thanks to the new yarns. For example, at Arapkir in central-east Anatolia, British yarns created a new weaving industry in the 1820s–1830s, one that subsequently grew during the nineteenth century, utilizing 1,000 and then 1,200 looms.50 Local merchants and, perhaps, residents who had migrated to Istanbul organized these industries – but which agents were responsible is not known.51 In addition to putting-out and guild forms, there was a third form of textile production organization, that of direct merchant supervision. In Damascus, for example, guild monopolies were badly eroded by the first half of the nineteenth century. There, in 1842, merchants successfully established weaving workshops under their own direct control, despite the guilds’ appeal to traditional rights. In the coming decades the merchants’ control over textile production within the city increased. They set up large workshops and hired a single master weaver who 46 Bruce McGowan, ‘The Age of Ayans, 1699–1812’, in Inalcık with Quataert (eds), Economic and Social History, pp. 637–748, at pp. 698–699. 47 Ibid., pp. 706–707. 48 Quataert, ‘Age of Reforms’, p. 905. 49 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, p. 60. 50 Quataert, ‘Age of Reforms’, pp. 905–906. 51 Cem Behar, A Neighborhood in Ottoman Istanbul: Fruit Vendors and Civil Servants in the Kasap Ilyas Mahalle (Albany, NY, 2003), p. 105 where he says that only local merchants in Arapkir participated. But the author traces in detail the substantial presence of Arapkir residents in Istanbul; it is possible, even likely, that these afforded some connection which helps explain why Arapkir so precociously developed a textile industry around the use of British yarns. 490

The Ottoman Empire, 1650–1922 supervised numbers of journeymen weavers. Moreover, to remain competitive, they lowered piecework wages, which fell over several decades during the mid-nineteenth century. By the 1870s, merchants’ and not guild leaders’ voices were determining wages and work conditions. In Nablus, a remote hill town in the 1850s, some merchants customarily set up weaving workshops and hired both women and men for wages.52 At Mosul, three families were already gaining control over the artisanal population in the early eighteenth century.53 And finally, there are the merchant-supervised carpet knotting workshops that emerged in the later nineteenth century in western and central Anatolia. There, for example, an Izmirbased merchant consortium ran 17 workshops to which women and girls came and knotted carpets, receiving piecework wages.54

The workers Boys, girls, women and men of every religion and ethnicity worked in the making of textiles. Child labour was commonplace in every aspect and locale of textile production. Thus, we find children and women engaged in commercial embroidery, spinning, weaving, silk reeling and carpet knotting in workshops, homes and factories. A few examples should suffice to show the range of child labour. In late eighteenth-century Thessaly, every hand, ‘even those of the children, is employed in the [24] factories’ of Ambelakia, a booming red yarn exporter to Europe. Also, the government, during the 1830s, brought orphaned children from central and east Anatolia to spin yarn in state factories for the fleet. In a similar action taken in the 1860s, the state used orphans to provide uniforms to the gendarmerie in Sofia. In the early twentieth century, children laboured in the merchant-owned cotton mills in Istanbul, Adana, as well as Izmir and the Balkan provinces. At Salonika, boys and girls aged 12 to 18 years worked, while the Macedonian mills employed girls as young as 6 years old. In the silk reeling mills of Salonika during the 1840s, mainly ‘poor children’ worked. Generally, most silk reelers in the steam-powered factories were girls and young women under 18 years of age. At Bursa, girls of 12 to 14 years of age began as apprentices, for half the normal wages. During the 1890s children working at home in the Sivas region helped their mothers commercially weave cotton, both from homespun and imported yarns. At Aleppo, ‘a considerable number of children’ assisted the women who spun cotton and silk for male weavers.55 Gender stereotypes notwithstanding, female labour was an important, indeed crucial, part of the textile industry across the empire. The guilds, it seems, successfully 52 Beshara Doumani, Rediscovering Palestine: Merchants and Peasants in Jabal Nabls, 1700–1900 (Berkeley, 1995), p. 127. 53 Khoury, State and Provincial Society, p. 138. 54 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution, pp. 158–159. 55 Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution is the source for the data on child labour. Also Palairet, Balkan Economies, pp. 80–81. 491

The Ashgate Companion to the History of Textile Workers prevented women and girls from joining their ranks and it is nearly certain that guilds were exclusively male. Thus, it seems likely that guilds did not provide the opportunity for training outside of the home.56 Men therefore often formed the majority of urban cloth weavers in the various areas of the empire – Arab, Anatolian and Balkan. Thus, men commercially wove in twentieth-century Baghdad and in mid-nineteenth-century Aleppo and Diyarbakr. Men wove the famed mohair cloths before that industry collapsed. But the lack of professional training in guilds did not spell female exclusion from cloth weaving for the marketplace. Indeed, Muslim women commercially wove both cotton and linen in a home-industry network in east Anatolia during the 1830s. In the Trabzon area around 1870–1900, Muslim women printed headscarves using British cloth and also wove various silk and cotton cloths at home, likely on a putting-out basis. While most of the evidence shows women weaving for the market within their homes, we also find them at workshops, for example the Muslim women in Nablus workshops during the midnineteenth century. Similarly, at Salonika c.1880, between 8 and 20 females worked in various textile workshops. Thus, females weaving at home may have been the norm but weaving outside was commonplace. Exclusion from guild membership did not mean women were not involved in guild work. In practices that date back centuries, women routinely worked to supply guild members with raw materials and yarns. Thus, for example, women from at least the seventeenth century provided mohair yarn to the Ankara guilds57 while during the 1840s Aleppo-area women spun cotton yarn for the urban male guild weavers. There seems little doubt that women in many locations, while not formally guild members, worked with them on a day-to-day basis. In this respect, guild membership served as one form of gendering in the Ottoman world. Outside of the guilds, women and girls worked in every sector of textile production. Examples abound, ranging from the embroidery network outside Istanbul to weavers in Trabzon and factory girls. For example, women in seventeenth-century Ankara assisted their weaver husbands when the looms, as often was the case, were located at home. And in mid-nineteenth-century Aleppo, women ‘virtually monopolized jobs in silk reeling, processing of wool for felt and wool fabrics, button-and braid-making, embroidery and piecing wool fabric into cloaks’.58 Moreover, female participation in the textile labour force certainly increased during the second half of the nineteenth century. The majority of the new workforce in the silk reeling mills of the Lebanon, west Anatolia, the Salonika region and elsewhere were girls, usually unmarried, who often lived in dormitories. Similarly, mostly young girls worked in the steam-powered yarn factories in Karaferia, Niausta, Wodena in Macedonia, as well as in Salonika. In Aleppo, a significant new cottage industry employed urban women in the mechanized 56 Vatter, ‘Militant Textile Weavers’, p. 39, who suggests, incorrectly I believe, that women did not learn marketable skills at home. 57 Suraiya Faroqhi, Subjects of the Sultan: Culture and Daily Life in the Ottoman Empire (London, 2000), pp. 112–113. 58 Vatter, ‘Militant Textile Weavers’, pp. 36–37. 492

The Ottoman Empire, 1650–1922 knitting of cotton stockings; by the end of the Ottoman era they supplied onefifth of the market in Syria. In the vast carpet-making industry of west and central Anatolia, the bulk of the new workers were females, employed outside the home, in scores of workshops. (Before about 1825, most of those knotting at home were likely to have been female.) Such massive participation of women and girls in the expanding sectors of textile production was no accident, nor was it because of their nimble fingers. After all, except for embroidery work and mechanized silk reeling, men also worked in these various activities. For example, men at the important centre of Kula knotted rugs while in the yarn mills of Salonika they formed about one-quarter of the labour force. The concomitant growth of several textile sectors and of their female labour force was part of the process of lowering labour costs in the struggle for competitive survival. The textile industry already was the lowest paid Ottoman industrial sector and the new jobs in embroidery, mechanized silk reeling and yarn spinning, and in carpet-making, were the lowest in the industry. Thus, the success of the various industries was predicated on the lowering of wages for the female workers. In sum, there was no empire-wide gender division of labour in the textile industry but a host of practices produced by specific circumstances. Similarly, it is an inaccurate stereotype to speak of an ethnic or religious division of labour. In the past, writers freely but inaccurately spoke about the inherent, nearly genetically encoded propensity of certain groups for certain kinds of labour. In this so-called ‘ethnic division of labour’, Muslims were seen to be good soldiers and perhaps farmers while Christians of the different ethnicities were understood to be skilled artisans. Such gross characterizations belied the fascinating reality of the Ottoman workforce and, of interest here, the textile workers. While certain ethnic or religious groups did dominate certain forms of labour in particular regions, they did not do so for any industry in the empire as a whole. Thus, Jewish workers constituted the embroidery networks around Istanbul while in Salonika, with its majority Jewish population, 80 per cent of the girls in the cotton spinning mills were Jewish. But, not far away, ethnic Greek girls staffed the spinning mills at Niausta and Karaferia while Bulgarian girls staffed the mill at Wodena. In the vast carpet-making industry, Muslims who largely were ethnic Turks had long dominated the workforce in and around Uşak. But when Izmir merchants began organizing production in other locations, the workforce largely came to consist of Ottoman Christians. Thus, all of the carpet workers at one location could be Muslims and ethnic Turks while, just a few miles away, they all were ethnic Greek or Armenian Christians. The lack of a particular gender or ethnic profile of the labour force speaks to labour recruitment issues. Overall it seems that some kind of personal familiarity – either through family, acquaintance or place of origin – was a crucial aspect of recruitment. Merchants clearly were major agents of recruitment in both urban and rural areas. The precise mechanisms are unclear, but their use of religious dignitaries – imams, priests or rabbis – is certain and likely commonplace. Contemporaneous histories of both the Bursa and the Lebanon silk reeling industries specifically mention how declarations from religious authorities regarding the propriety of such work helped persuade girls to enter the factories. The religiously homogenous nature of the carpet-making workforce in a particular region suggests appeals made 493

The Ashgate Companion to the History of Textile Workers either through religious authorities or in the name of that religion. In many cases, merchants could persuade workers to enter a workplace because a relative already was present. Particularly crucial for female and male labour, it would seem, was the presence of a friend or relative to introduce the merchant. For girls and women, the relative or friend likely helped to teach the craft as well. In the case of the silk reeling mills, entrepreneurs hired foreign women reelers to come and introduce the new task and offer a role model for women’s work outside the home. Guilds of course frequently solicited the labour of their members’ daughters and wives and the membership of their sons. Hence, the Ankara wives wove with their husbands. The presence of sons was commonplace in guilds in all areas. In some towns and cities, chain migration played an important role in labour recruitment. In one study of an Istanbul neighbourhood, workers in a particular shop most often hailed from the same village in Anatolia.59 More generally, certain occupations in the capital city were the monopoly of migrants from particular towns and villages in Anatolia and the Balkans. Depending on whether it was home, workshop or factory labour, workers’ housing included their own homes, rooms above the shops or dormitories provided by factory owners. When guilds were present, apprentices sometimes lived with the masters while journeymen often maintained their own households, at least in some locations. Weaving was year round, more or less, at many locations, including important centres such as Aleppo and Diyarbakr (with an average of 200–290 work days/year) and less significant ones as well, such as Bitlis and Baghdad. But in many other locations, such as Manisa, weavers wove only in the winter months and took on agricultural tasks in the summertime. This strategy appears repeatedly, often in putting-out weaving at home as well as in workshops. At full employment, Damascus journeymen weavers worked 10-hour days, six days a week, at their looms from just after sunrise until sunset.60 In the Macedonian cotton mills, the workday averaged 15 hours in the summer and 10 in the winter, with a 35-minute break for dinner and none at all for breakfast. In the silk reeling mills at Salonika during the 1840s, the workday averaged 14 hours while at Bursa it averaged 13.5 hours in summer and 7.5 in the winter (including breaks and meals). Was there an ‘industrious revolution’ in which people worked harder in order to buy goods?61 Indirect evidence indicates this to be the case but more research is required. Given the importance of the sector, there are relatively few reports of labour activism among textile workers. Much of the recorded labour action has been that 59 Kırlı, ‘The Struggle over Space’. More than 200 shops were enumerated; of these more than 10 per cent involved textiles. The chain migration of the other workers, including the porters of Istanbul is well known. Donald Quataert, ‘Ottoman Workers and the State, 1826–1914’, in Lockman (ed.), Workers and Working Classes in the Middle East, pp. 21–40. 60 Sherry Vatter, “Journeymen Textile Weavers in Nineteenth Century Damascus: A Collective Biography”, in E. Burke (ed.), Struggle and Survival in the Modern Middle East (Berkeley and Los Angeles, 1993), pp. 75–90, at pp. 78–79. 61 Jan de Vries, ‘The Industrial Revolution and the Industrious Revolution’, Journal of Economic History, June (1994): 249–270. 494

The Ottoman Empire, 1650–1922 among textile workers in or at factories and mills, where labour is concentrated and its actions more visible. State documents frequently record protests around wages and working conditions in state-run factories making headgear and uniforms. Labour unrest in privately owned factories occasionally appears. For example, during March 1908, 1,500 mainly female hand-spinners in the carpet-making centre of Uşak sacked three yarn spinning factories which had deprived them of work.62 The dispersed nature of textile production in the home and in small workshops diminishes the likelihood that labour action at such locations was noticed and recorded. But it can be uncovered. From the 1870s (and perhaps earlier) until after the First World War, journeymen workers in Damascus regularly struck against the masters who employed them. ‘These protests were collective, disciplined actions’ seeking higher wages and better working conditions.63 It does not seem likely that Damascus was unique in this regard. Such class-conscious labour actions likely occurred elsewhere. Textile workers participated in the waves of protest that swept the Ottoman world in the period immediately following the Young Turk Revolution of 1908. Between late July 1908 and the end of the year, some 111 strikes swept every sector of the economy, mostly in the transport sector. Ten strikes during this period involved textile workers, a disproportionately small number given the importance of the sector to the overall economy. In every case, they struck for higher wages or better working conditions and more than half of the recorded textile strikes occurred at large factories. There were very few textile workers in the class-conscious labour federation that emerged shortly thereafter at Salonika, and which later mobilized 7,000 workers for a 1911 May Day celebration. In a general survey of labour unrest between 1872 and 1907, 50 strikes were counted, only nine in the textile sector. In the period 1909–1915, textile workers accounted for 10 of the 38 strikes recorded.64 Textile workers formed a number of new associations very late in the Ottoman period, generally after 1908. Their connections to existing forms of workers’ organizations, notably guilds, are not clear. Thus, in big cities such as Istanbul and Izmir, tailors, weavers and spinners separately formed craft unions. In other sectors, as early as the 1880s, workers in some sectors, mainly transport, had formed mutual aid societies offering limited pension, burial and health benefits. But the involvement of textile workers in such organizational activities is not certain.

62 Quataert, ‘Machine Breaking’. 63 Vatter, ‘Militant Textile Weavers’, pp. 38ff. Vatter is among the few looking for labour activism outside of factories among Ottoman workers. It is quite likely that if such research were carried out in other centres, results similar to the Aleppo experience would be found. 64 Yavuz Selim Karakışla, ‘The 1908 Strike Wave in the Ottoman Empire’, Turkish Studies Association Bulletin 16 (1992): 153–177; Joel Beinin, Workers and Peasants in the Modern Middle East (Cambridge, 2001) pp. 80–81; Şehmus Güzel, ‘Tanzimat’tan Cumhurıyet’e İşçi Hareketi ve Grevler’, Tanzimat’ tan Cumhuriyet’ e Türkiye Ansiklopedisi 3 (1985): 803–828. Of course these statistics do not record all instances of labour unrest and are presented here for impressionistic purposes. The post-1908 strike figures are remarkable given the hostile legislation passed after the 1908 strike wave. 495

This page has been left blank intentionally

19 Turkey, 1922–2003 Lisa A. Seidman

The Ottoman legacy to the Turkish Republic in 1923 was an Anatolian economy battered by war, depopulated of most of its business and artisan class, and laden with the debts of the fallen Empire. A small number of public and privately run textile manufacturing facilities remained in operation. Textiles were a major import item due to the 1923 Treaty of Lausanne, which did not allow duties or tariffs on imports until 1929. Exports were minimal and included only raw fibres, madder root used to make the famous Turkish red dye and a small volume of wool handknotted carpets. Raw cotton output was small, 51,240 tonnes in 1930, and yields were low but adequate for both home-based and factory yarn production. Roughly half of domestic wool production was absorbed into carpet yarn. The Great Depression and the collapse of commodity prices made it unsustainable to rely on exchanging raw materials for imported manufactured goods. The Turkish government, with domestic tax revenues and partial funding from the Soviet Union, put together a Five-Year Economic Plan. The focus of the plan, under the home-grown ideological banner dubbed etatism, was to build industrial capacity to produce consumer goods, especially textiles, and to expand irrigated agriculture including cotton. The state established the Sümerbank in 1933 as a holding company to oversee industrial production; it took over the Ottoman textile plants and built stateowned spinning and weaving enterprises throughout the new nation. Traditional handicrafts produced in small un-mechanized workshops continued to provide  Danielle Leclercq, ‘Migration Flows in 20th Century Europe and Their Impact on School Life’ (Donaueschingen, 1998). By 1927, only 65,000 Armenians and 120,000 Greeks were living in Turkish territory compared with 2.3 million and 2.1 million respectively in 1870.  William Hale, The Political and Economic Development of Modern Turkey (London, 1981), p. 47.  Cumhüriyet Halk Partişi, ‘Tarımda Değişim – Gelişim Süreci’ (Ankara, 2003), section 3.2.4.1, table 2, based on data from the Turkish State Institute of Statistics (SIS).  Hale, Political and Economic Development, p. 56 specifies that the Soviets gave US  $8 million out of the eventual US  $50 million cost of the first five year plan under etatism.

Map 19.1

Turkey

Turkey, 1922–2003 up to 60 per cent of manufacturing value added until the end of the 1930s. Private enterprises in 1939 produced 65 per cent of cotton yarn and cloth, 40 per cent of wool yarn and cloth, and 38 per cent of leather goods. A small number of apparel firms came into being in response to the new Republic’s ‘clothing revolution’ which required men to wear Western suits and hats. For example, Vakko, today a leading ready-to-wear producer, was founded as a hat and scarf manufacturer in 1934. After a brief period in the 1940s and early 1950s in which the state took the lead in textile production, private manufacturers regained their dominance. State-produced cotton cloth output dropped from 75 per cent of total production in 1950 to 28 per cent in 1967. The acceleration of private production was due to post-Second World War investment incentives including: • the 1954 removal of a production tax based on the number of looms in a shop • increased tariffs on textile imports • increased import quotas for textile machinery and • the 1950 establishment of the Turkish Industrial Development Bank (Türkiye Sanayı Kalkınma Bankası) with a programme of foreign exchange credits to help manufacturers buy equipment. These policies were complemented by plentiful cotton growing on land irrigated by the new Seyhan dam. The impact of these incentives is reflected in a rapid increase in weaving capacity and in the proportion of motorized looms after 1953 (76.95 per cent by 1960). Investment in spinning and weaving capacity was so rapid that by the late 1950s there was overcapacity in cotton textiles. Apparel, on the other hand, with  Ibid., p. 59.  Organization for European Economic Co-operation (OECD), The Textile Industry in Europe: Statistical Study (Paris, 1962).  OECD data in the The Textile Industry in Europe series show the increase in domestically produced cotton supplies in response to irrigation from 51,240 tonnes in 1930 to 139,000 tonnes in 1953, and 159,000 tonnes in 1959. Waltina Scheumann, Institutional Arrangements for Controlling Salinity: Experiences from Turkey (Morocco, 2002), and Global Agriculture Information Network (GAIN), Gain Report #TU2026 (Washington, DC, 2002) describe the development of the irrigation systems underlying the expansion of cotton production in Turkey. Over the period 1996/97 to 2000/2001, Turkey produced an average of 3.7 million bales (218 kg per bale) per year and yarn mill use averaged 5 million bales per year. Production continues to grow as irrigation expands in the south-east.  General Agreement on Tariffs and Trade, A Study on Cotton Textiles (Geneva, 1966), pp. 175, 185, 192. The estimate of the proportion of motorized looms is based on 75 per cent of total capacity.  Hale, Political and Economic Development, p. 93 points out the overcapacity in cotton products and OECD, Textile Industry in Europe statistics confirm the complete disappearance of cotton fabric imports into Turkey between 1955 and 1960. However, 499

The Ashgate Companion to the History of Textile Workers few exceptions was still produced by tailors and in private homes up to the 1960s. Industrial production of ready-to-wear clothing began to expand in the 1960s with mass production of suits and outerwear. By 1960 roughly a third of all industrial establishments were producing textiles for domestic markets. However, the economy still heavily depended on imports (except for processed food, textiles, iron and steel).10 An import substitution (IS) policy was adopted to further develop domestic industry. IS increased the state’s burden of foreign exchange payments for imported industrial inputs. Remittances from 910,000 Turkish guest workers in Europe, which peaked at US  $1.4 billion in 1974, formed the main source of foreign exchange revenues during this period.11 The trap of government commitments to both large outlays of foreign exchange and a large public sector wage bill tightened into the 1970s. The government planned for a US  $1 billion expansion of state spinning and weaving enterprises between 1972 and 1977 but it was not implemented due to lack of funds. The oil shocks crises of the 1970s put Turkey’s economy into a deep decline. As worldwide recession reduced demand for Turkish labour in Europe, Turkey lost its main source of foreign exchange.12 In 1979, a US  $1.8 billion loan from the World Bank and IMF was tied to a radical reform programme designed to create an exportoriented economy, improve the balance of payments and combat inflation. On 24 January 1980 the stabilization programme was launched. The second phase included: • the maturation of older apparel firms into the development of their own branded products (an on-going process) • the movement of original equipment manufacturing (OEM) and generic production to the south-eastern region of Turkey where labour is less expensive and • the development of a large home textiles segment, which is extending the growth horizon of the sector as it matures. The beginning of the first phase of textile export expansion was characterized by tremendous growth in generic knit and woven apparel exports with supporting growth in cotton yarns and fabrics. In 1980, exports of apparel (valued at US  $106 OECD trade figures also show that Turkish imports of non-cotton yarn and cloth imports increased from 10.7 per cent of consumption in 1955 to 21.4 per cent in 1959, and then levelled off but continued at 13.2 per cent in 1960. 10 Sinan Akışık, ‘Small-scale Manufacturing Firms in Underdeveloped Countries: The Case of Textile Firms in Bursa Turkey’ (unpublished Ph.D. thesis, University of Pennsylvania, 1985), p. 85. 11 Hale, Political and Economic Development, p. 19 explains that between 1960 and 1975, 910,000 Turks went as guest workers to Europe with their families so the total number of Turkish people in Europe was approximately 1.2 million. 12 Tuğrul Çubukçu ‘The Transformation of the Etatist System Towards a Free Enterprise Economy: The Records and Future Tasks’, in Werner Gumpel (ed.), Turkey and The European Community: An Assessment (Munich, 1992). 500

Turkey, 1922–2003 million or 3.6 per cent of total exports) were much smaller than textiles and yarn (US  $671 million or 23.1 per cent of total exports). Carpets were a more important export than apparel with 3.8 per cent of the total export earnings and 13.9 per cent of the manufacturing export earnings in 1982.13 After six years apparel had caught up with textiles with each valued at US  $1 billion and making up 14 per cent of exports. The maximum gap between apparel and textiles was reached in 1995 with apparel values at US  $6.2 billion or 29 per cent of exports versus textiles at US  $2.1 billion and 10 per cent.14 Germany supplied Turkey with an easy customer base, building on connections between Turks abroad and at home. Export sales soon expanded into the rest of Europe and finally extended to the United States as well as tens of other countries around the world. In addition to formal trade, a large volume of undocumented ‘suitcase trade’ with Russia and Eastern Europe grew quickly. During the 1990s, there was a threefold increase in domestic private investment in the textile and apparel industry. Turkey controlled 3.5–4 per cent of the global clothing market, held the rank of sixth largest clothing exporter in the world and joined the top 10 nations in terms of its capacity in spinning and weaving.15 Turkey’s experienced textile entrepreneurs, many in business since the 1950s, aggressively upgraded technology in both production and marketing during the 1990s.16 Diversification has been the hallmark of the second phase of export growth. According to the Istanbul Textile and Apparel Association (ITKIB), in 2000 the value of apparel exports reached US  $7.2 billion, textiles being US  $2.8 billion. Home textile exports hit the US  $1 billion mark in 2001. Turkey’s 2000 apparel-output was composed of knits (51 per cent), woven apparel (34.6 per cent) and other made-up articles (14 per cent). Within apparel, basic cotton knit T-shirts and other underwear were still the largest category in 2000 (US  $1.3 million) followed by woven women’s wear (US  $1.06 million).17 However, Turkey is increasingly manufacturing clothing for major international labels that require high quality, consistency and speed in production that can only be achieved with sophisticated textile machinery and equipment. The relatively capital-intensive subsectors of branded apparel, as well as home textiles, are important growth areas in the face of low wage competition. Turkey joined the EU Customs Union (CU) in 1996. The expected expansion of sales did not materialize in 1996/97 and on the downside new restrictions were placed on input purchases from countries not party to the EU CU. Turkey’s Free 13 Günseli Berik, Women Carpet Weavers in Rural Turkey: Patterns of Employment, Earnings and Status, ILO: Women, Work and Development 15 (Geneva, 1987), p. 2. 14 The Istanbul Textile and Apparel Association (İstanbul Tekstil ve Konfeksiyon İhracatçı Birlikleri – ITKIB), ‘Turkish Textile and Apparel Industry’ (Istanbul, 2003). 15 US Department of Commerce, U.S. Textile Machinery Industry Overview (Washington, DC, 2000); US Department of Commerce, Turkey: Leading Sectors for U.S. Exports and Investment (Washington, DC, 1998); Interview with Serdar Yavaşoğlu, chairman of the Aegean Clothing Industrialists’ Association, presented on the http://www.exportturkey. com/news website, 2001. 16 Susan Black, ‘Turkey Staking a Claim’, Bobbin Magazine, September (1994). 17 ITKIB, ‘Turkish Textile and Apparel Industry’. 501

The Ashgate Companion to the History of Textile Workers Trade Zones, started in 1985 with 19 open by 2000, have played an important role in keeping input costs down while expanding exports. The biggest single country buyers of Turkish textiles and apparel in 2000 were Germany (36.6 per cent), the USA (16.6 per cent), the UK (11.9 per cent), France (6.8 per cent) and Holland (5.3 per cent).18 Around the turn of the century, exports to the USA were expanding rapidly and markets with Russia were being formalized. A key focus of Turkish textile and apparel firms was to establish a solid US customer base before no-holds barred competition began in 2005 when all quotas were nullified under the GATT.

Labour force overview The population of the early Republic was predominantly rural with only 16.4 per cent of the total 13.6 million people living in urban areas.19 Based on the 1927 Industrial Census, textile enterprises formed 14.3 per cent of the total 65,245 formal manufacturing enterprises. Manufacturing employed 256,855 people with 48,025 (18.7 per cent) working in textiles – not including home-based artisan production. Among formal textile establishments, only 4.6 per cent had 11 workers or more and even fewer, 2.6 per cent, had motorized equipment.20 The 1927 Industrial Census shows that despite the withdrawal of non-Muslim women from the labour market, urban women held a high proportion (52 per cent) of the small number of manufacturing jobs (in enterprises with at least four workers) during this period. This was due in part to the estimated 1 million widows left by the wars preceding independence. Textiles represented 42 per cent of all 38,000 formal manufacturing jobs done by women in 1927 versus 14 per cent of the 108,154 men’s jobs in the sector.21 Despite the high percentage of women within the sector, out of the total population of urban ever-married women, few worked in formal manufacturing (at most 6.4 per cent) or specifically in textile manufacturing (2.7 per cent).22 Due to labour shortages in the first two decades of the Republic, most new entrants to the labour force were needed in an agriculture sector still dominated by manual labour. In fact, census data show a drop in industrial employment from 8.3 per cent in 1935 to 7.4 per cent in 1950.23 Work in the new State Economic 18 Ibid. 19 Urban is defined here as population greater than 10,000. 20 Rüştü Dağlaroğlu, ‘L’Industrie Textile Turque’ (unpublished Ph.D. thesis, University of Neuchatel, 1941), p. 111; Hale, Political and Economic Development, p. 43. 21 Frederic Shorter, ‘The Population of Turkey after the War of Independence’, International Journal of Middle East Studies 17 (4) (1985): 417–441, at p. 433. 22 The author’s calculation based on Shorter’s estimates for this period presented in ‘The Population of Turkey’. 23 Hale, Political and Economic Development, p. 64 comments that the census data may not be entirely accurate because there were methodological changes in the census over the period. 502

Turkey, 1922–2003 Enterprises (SEEs) of the 1940s and 1950s, however, was a welcome source of cash for rural households. They approached such jobs as temporary work for the agricultural off-season and as a result factories found it difficult to remain fully staffed.24 To prepare for the Second World War, the state shifted men from factory work to the army, which exacerbated labour shortages. In addition, a punitive tax (varlık vergisi) was imposed on non-Muslim citizens. The government also forced privately owned textile factories to sell their output to the Sümerbank, which then resold it at high prices driven up by the lack of imports during the war.25 After the war, textile entrepreneurs expressed frustration with etatist meddling, especially the crowding out of private capital. By 1947, the government had begun to shift its investments towards heavy industry. Prior to the Second World War the majority of textile factory owners were either Jewish or dönme.26 However, the varlık vergisi had a chilling effect on non-Muslim textile start-ups after the war. The new textile manufacturing firms that formed, many with only one or two looms, were started primarily by Muslims. The early Istanbul textile manufacturers tended to be from a merchant or artisan background (weavers, bakers and so on) with a minority from the military and/or landed elite (ulema). In contrast, Anatolian manufacturers generally moved into textiles from a pre-Second World War background in industrial processing of raw materials for export.27 The urban labour force grew rapidly during the 1950s as mechanization led to a decline in the demand for farm labour and urban industrial jobs sprang up. Istanbul was inundated with a tide of people; in 1940 the population of Istanbul was less than 800,000 and by 1975 there were 3.9 million people living in the city, including the illegal housing ringing its outskirts.28 One niche of rural textile employment, hand-knotted carpet production, grew substantially from the 1960s to the 1980s. Increased demand for carpets led to threefold increases in employment in carpet weaving from 147,693 in 1958 to 577,800 in 1981. This activity accounted for

24 Hale, Political and Economic Development and Yıldırım Koç, ‘Workers and Trade Unions in Turkey’ (http://www.turkis.org.tr/history.htm, 1999) discuss the situation of industrial labour in the 1950s. Hale shows, for example, that the new Sümerbank textile factory in Kayseri had to hire 3,000 workers to maintain a working staff of 2,000 (p. 65). 25 Hale, Political and Economic Development, p. 60. 26 Edward C. Clark, ‘The Emergence of Textile Manufacturing Entrepreneurs in Turkey: 1804–1968’ (unpublished Ph.D. thesis, Princeton University, 1969) explains that after the Independence War, many private textile factories were owned by dönme. Dönme is the appellation given to the followers of Shabtai Tzvi, a false Jewish messiah who was forced to convert to Islam by the Ottoman authorities in 1666. The dönme continued to follow Shabtai Tzvi and his teachings after his conversion and long after his death. Throughout its history, the dönme community was closed with intermarriage between families and closely linked family businesses. 27 Clark, ‘Emergence of Textile Manufacturing’, pp. 137–140. 28 Hale, Political and Economic Development, p. 26. 503

The Ashgate Companion to the History of Textile Workers 3.5 to 3.9 per cent of the total female labour force, or between 8 and 9 per cent of economically active women in 1981.29 By 1999, according to Ministry of Labour statistics, 505,152 employees worked in the formal textile and apparel sector. Union estimates of both unregistered and registered workers claim that there are 2 million in the sector or roughly 10 per cent of the total labour force.30 It is likely that a large proportion of the unregistered workers are in the apparel rather than the textile branch.

Unions Increased numbers of urban workers in the 1950s, although still by far in the minority,31 began to create pressure for unionization. Trade unions began forming when finally made legal in the post-war liberal atmosphere of 1946, and textile unions played a central role. The main trade union federation, TÜRK-İŞ, was conceived at the 1952 congress of the Textile and Knitting Workers’ Federation (TEKSIF). TÜRK-İŞ was formed as a ‘bread and butter’ style union because the workers of the 1950s were politically and culturally conservative and the state generally suppressed class-based union organizing. A fully developed system of labour industrial relations laws was put into place in 1963 after the coup of 1960. In 1960, 23,380 members of TÜRK-IŞ (20 per cent) were associated with TEKSIF and distributed among 17 regional textile unions focusing on different types of cloth and yarn production. The more radical DISK came about as a 1967 break-off from TÜRK-IŞ. HAK-IŞ was created with an Islamic focus in 1976.32 All three federations contain their own textile unions. Textile workers made up a large portion of manufacturing labour during the 1970s (33 per cent in 1973) with many working for private firms.33 The majority of these workers were not unionized and earned lower wages. Due to union pressure, real wages for workers in the modern industrial sector were increased rapidly and union coverage was extended somewhat. The unions reacted against the 1980 stabilization plan (discussed above) with factory occupations, strikes and protests.

29 Berik, ‘Women Carpet Weavers’. 30 Barış Tan, ‘Overview of the Turkish Textile and Apparel Industry’, Harvard Center for Textile and Apparel Research Working Paper (Cambridge, 2001), p. 34. 31 Aysıt Tansel, Economic Development and Female Labor Force Participation in Turkey: Time Series Evidence and Cross-Province Estimates (Ankara, 2001) reports that in 1960 the majority of employment was still in agriculture, 67 per cent of male and 95 per cent of female employment. 32 Yıldırım Koç, ‘Workers and Trade Unions in Turkey’ and Friedrich Ebert Stiftung, ‘Trade Unions in Turkey’ (Digital Library: http://www.fes.de/fulltext/bueros/istanbul/00253001. htm, 2003) present the history of labour unions in Turkey. 33 World Bank, Turkey: Prospects and Problems of an Expanding Economy (Washington, DC, 1975), pp. 221 and 257. 504

Turkey, 1922–2003 The chaotic period of the 1970s ended with a military coup in September 1980.34 Strict laws dictating union conduct were put into place in 1983 and the power of unions declined until the 1990s. After the 1980 coup, all trade unions, aside from TÜRK-IŞ, were suspended (HAK-IŞ until 1981 and DISK until 1991). In 1999, the union federations threatened a strike involving 70,000 textile workers. The unions, bargaining with the textile employers association,35 demanded higher wages and complained that ‘the employers want to deprive the workers of their existing rights and to tell workers “If there is something to do, you work, otherwise you go home”’.36 Strike action was avoided with pay increases after four months of tense negotiations.37

Wages Minimum wage law has been implemented throughout Turkey since 1974.38 Wages for textile workers are difficult to assess due to the large numbers of undocumented workers and the lack of official wage data broken down by occupation. Roughly speaking Turkey’s manufacturing wages are in the mid-range according to global standards. For example, in 1974 average (skilled or unskilled not specified) manufacturing wages equalled US  $0.63 per hour for Turkey compared to wages for skilled workers of US  $0.61 in Thailand, US  $0.72 in Singapore, US  $0.56 in Indonesia and US  $0.31 in India.39 The textile research group Werner International reports Turkish wages for 1998 at US  $2.48 per hour for textiles (spinning and weaving) and US  $1.84 for apparel compared to US  $0.43 for China, US  $0.39 for India and US  $1.36 for Morocco. Other estimates provide similar figures.40 The ITKIB claims that recent reforms in social 34 Z.Y. Hershlag, The Contemporary Turkish Economy (New York, 1988) points out that, ‘Political and social unrest caused, inter alia, a loss of more than 4,000,000 work days in 1979’ (p. 103). 35 The Turkish Confederation of Employers’ Associations (TISK) was formed in 1961. 36 Turkish Daily News, 6 January 1999. 37 US Department of State, 1999 Country Reports on Human Rights Practices (Washington, DC, 2000) and Turkish Daily News, 9 August 2000. 38 Insan Tunali, Background Study on Labor Market and Employment in Turkey (Ankara, 2003). 39 Folker Fröbel, Jürgen Heinrichs and Otti Kreye, The New International Division of Labor: Structural Unemployment in Industrialised Countries and Industrialisation in Developing Countries (Cambridge, 1980). 40 UBS, Price and Earnings Around the Globe: An International Wage Comparison (Zurich, 2002) provides a range of wage data for female industrial workers in capital cities around the world. The UBS study claims that in ‘Istanbul the wages of female industrial workers who are mostly textile workers are $4,900.00 per year net of taxes and social security contributions (which combine to about 22 per cent of gross pay). The average weekly hours were 42 for this group of workers’ (p. 41). This scenario, US  $2.43/hour, mirrors 505

The Ashgate Companion to the History of Textile Workers security are increasing costs to the extent that company-financed workers’ benefits and taxes are almost equal to the cost of salaries. The Werner estimate starkly contrasts with claims made by TEKSIF during the 1999 strike threat that a union textile employee with 20 years experience earns US  $153 per month or US  $0.80 an hour (less than the 1999 minimum wage). Further, it should be emphasized that workers not covered by collective bargaining agreements earn approximately 53 per cent of covered workers’ wages.41 Therefore a range of textile wages is the most appropriate estimate: in 1998/99, 500,000 textile workers earned between US  $0.80 and 2.48 per hour and 1.5 million received between US  $0.42 and 1.31 per hour. Males are likely to receive a gender premium, estimated in 1988 as 15 per cent.42

Child workers Working children in Turkey include both unpaid family labour and wage-earners. In 1994 children between the ages of 6 and 14 made up 19 per cent of the population. Young children (6–11) are unlikely to work in Turkey.43 SIS reports a 1.3 per cent employment rate for 6–11-year-olds in 1994.44 Older children are more likely to participate. Research on 12–17-year-olds indicates that 35 per cent of Turkish teenage girls who work outside the family home and/or farm choose textile-manufacturing employment compared to 14 per cent of boys.45 The reported proportion of children in the workforce has declined since Turkey joined the International Programme on the Elimination of Child labour (IPEC) in 1991.46 Between 1994 and 1998 there was a dramatic drop in the number of working 6–14-year-old children from 8.5 per cent

the Werner estimate for textile sector workers. Finally, The Washington Times, ‘Textile Industry Market Still Faces Barriers in the United States’ (Washington, DC, 2000) states in a special report on Turkey that ‘average monthly wages for Turkish textile workers are about $500 compared to $100 in China’. 41 İŞKUR, ‘Background Study on Labor Market and Employment in Turkey’. 42 Ibid., p. 6 reports that economic crises led to a steep decline in real minimum wages from the peak of US  $200 per month in 1998/99 to US  $120 per month in 2001. 43 Ibid. 44 State Institute of Statistics (SIS), ‘Child Labor in Turkey’ (Ankara, 1995). 45 Meltem Dayıoğlu and Ragui Assaad, The Determinants of Child Labor in Urban Turkey (n.p., 2002). 46 İŞKUR, ‘Background Study on Labor Market and Employment in Turkey’ provides details on child labour policy during the 1990s. In 1991 Turkey joined the International Programme on the Elimination of Child Labour (IPEC) organized by the International Labour Organization. In 1998, following an extension of mandatory years of schooling from five to eight, Turkey signed the ILO Convention 138, which established 15 as the minimum age of employment and called for progressive increases in the minimum wage. 506

Turkey, 1922–2003 (974,000) to 4.2 per cent (511,000).47 Outside of larger enterprises enforcement of labour laws in the textile sector is inconsistent. However, these figures do include ‘homework’ such as carpet weaving.

Organization of production Etatist and later import substitution policy which included high tariff protection for textiles (109 per cent in the early 70s) led to inefficiencies in plant size and balance between units that reduced productivity. During the post-1980 period many SEEs have been privatized.48 The privatization of Sümerbank began in 1995 and is ongoing.49 Foreign direct investment (FDI) increased in the 1990s in Turkey in response to the Customs Union and the South Anatolian development project (GAP).50 In addition, the attractive combination of Turkish Free Trade Zones in conjunction with skilled low-cost workers and proximity to Europe has brought textile companies, such as Hugo Boss in 1999, to set up operations in Turkey. However, FDI presence in the sector is small with 137 apparel and 49 textile firms out of 1,028 foreign-owned companies in manufacturing in 1998. Networking between firms at all levels (from local to international) has made it possible to forge a cohesive export industry out of a vast number of small privately owned firms.51 Groups of textile firms have relocated to, or started up in, industrial parks and Free Trade Zones.52 During the 1990s average firm size increased.53 Textile firms use complex subcontracting; nearly 30 per cent of textile firms (with more than 25 workers) both take in and put out work. This system helps to overcome capacity constraints as well as allocate production more efficiently when some parts of a garment may be produced more efficiently by small shops or pieceworkers.54 47 Dayıoğlu and Assaad, ‘Determinants of Child Labor in Urban Turkey’ report that the fraction of working male 6–14-year-olds declined from 9.9 per cent to 4.9 per cent and from 7.1 per cent to 3.6 per cent for females. 48 World Bank, Turkey, p. 258. 49 Tan, ‘Overview’, p. 33. 50 Ibid. 51 ITKIB, ‘Turkish Textile and Apparel Industry’ reports that 2,360 new Turkish companies started exporting clothing between 1996 and 1998 alone. Of these firms, 82 per cent were members of ITKIB located in Istanbul. During the first nine months of 1997 these new ITKIB members produced exports valued at 45 per cent of all ITKIB members’ production for that year with a value of US  $4.4 billion. 52 The Free Trade Zones Act was passed in 1985 and by 2000 there were 19 Free Trade Zones spanning the country with two more being developed. 53 Tan, ‘Overview’. 54 Erol Taymaz and Yılmaz Kılıçaslan, ‘Subcontracting Dynamics and Economic Development: A Study of Textile and Engineering Industries’, Middle East Technical University Economic Research Center Working Papers (Ankara, 2002) conducted a study using 1990 national data on textile firms with at least 25 workers. They report 507

The Ashgate Companion to the History of Textile Workers Home-based work is prevalent in artisanal production such as carpets or embroidery, as well as in the garment and footwear industries. Home-based work is generally conducted through privately run putting-out networks. For example, in the case of carpets, state carpet cooperatives attached to workshops were set up to promote rural employment in the 1960s. However, as world demand for Turkish carpets increased in the 1970s and 1980s, weavers preferred working with private merchants, who provided inputs on time, distributed output and adjusted prices quickly to market conditions.55 Another form of networking that developed after 1980 is the sector foreign trade company (SFTC) first started in 1994. SFTCs are formed with permission from the Trade Ministry and consist of multiple small export firms in the same industry. In 2002 there were 32 SFTCs with 11 in the textile and apparel sector and six in home textiles.56 SFTC members aim to achieve economies of scale and share technology and marketing experience. They also spread investment risk among the group – making unusual projects feasible. Networking in textiles and apparel is also facilitated by regional and national umbrella organizations.57

International environment Turkish workers in Europe created relationships and gained cultural and technical knowledge that contributed to the success of textile exports in the post-1980 period. Three cases provide insight into these international connections. Firstly, during the 1980s, an informal clothing production enclave was formed in Amsterdam by Turks who shifted into entrepreneurial textile activities after a stint as ‘guest workers’.58 Secondly, in the Paris garment district, 44 per cent of the sewing machine operators are Turkish men who set up shop with severance pay from auto or construction

that, on average, 75 per cent of the firms are involved in subcontracting either as a parent firm (31.1 per cent), a subcontractor (15.7 per cent) or both (28.7 per cent). Clearly the proportion that strictly subcontract, for instance take in work from another firm (a practice called fasoncuk in Turkish), would be higher if smaller firms were included in the survey. 55 Berik, ‘Women Carpet Weavers’, p. 14. 56 Export Promotion Center of Turkey (IGEME), Home Textiles in Turkey (Ankara, 2002). 57 The Istanbul Textile and Apparel Association (ITKIB) formed in 1986 and now has more than 17,000 members. ITKIB manages the distribution of apparel quotas among firms, keeps track of data in the sector, organizes training and so on. The Turkish Clothing Manufacturers Association (TCMA) has 400 members including the most influential in the sector. TCMA was formed in 1976 and works to improve quality in the industry and create international institutional alliances. 58 Joop Hartog and Aslan Zorlu, ‘Turkish Clothing in Amsterdam: The Rise and Fall of a Perfectly Competitive Labour Market’, De Economist 147 (2) (1999): 151–182. 508

Turkey, 1922–2003 jobs.59 Thirdly, there are 279,000 non-German self-employed workers in Germany – the majority of whom are Turks including some textile entrepreneurs.60 Turks involved in the textile sector abroad may transfer information, technology and/or capital back to Turkey directly or through family connections. For example, when the Turkish apparel enclave in Amsterdam was shut down by the Dutch government in 1992, at least 150 enterprises (more than half the sector) moved operations to Turkey. Many moved into the Free Trade Zones where they have flourished.61 The Turkish entrepreneurial group in Germany has grown quickly from less than 2 per cent of the Turkish labour force in Germany to 8.8 per cent in 1998 compared to 10.1 per cent for ethnic Germans. An interesting development is that ‘some German-Turkish firms are now established in both countries’ and ‘such activities are often embedded in familial and social networks that span the two countries and make transnational business activities possible’.62 The extent and impact of these links is an area that requires research.

59 Nancy Green, ‘Women and Immigrants in the Sweatshop: Categories of Labor Segmentation Revisited’, Comparative Studies in Society and History 38 (3) (1996): 411–433, at p. 421. 60 Antoine Pécoud, ‘Unemployment, Self-employment and Multiculturalism Among German-Turks in Berlin’, paper prepared for the 13th SASE Annual Meeting on SocioEconomics, University of Amsterdam, 28 June – 1 July 2001, consulted at http://www. sase.org/index.php?option=com­_wrapper&Itemid=53 (30 September 2009). 61 Hartog and Zorlu, ‘Turkish Clothing in Amsterdam’, p. 171. 62 Pécoud, ‘Unemployment’, p. 6. 509

This page has been left blank intentionally

20 The evolution of the Uruguayan textile industry María Magdalena Camou and Silvana Maubrigades

The first attempts to create a textile industry in Uruguay date back to the last quarter of the nineteenth century. Only in the beginning of the twentieth century did they start to meet with some success. Traditionally agricultural products had been Uruguay’s main export product. Until 1930 the food industry, which was responsible for nearly 50 per cent of total value added, dominated Uruguayan manufacturing industry. Within the food sector the meat industry, which produced the country’s main export product, was central, followed by the beverage and textile industries. These had risen and expanded since the end of the nineteenth and beginning of the twentieth century. The textile industry had developed late and moderately compared to other branches until the end of the First World War. By the early 1920s, it had received a major boost, becoming the second sector in manufacturing industry measured in value added. By then, the Uruguayan economy, based on the export of agricultural products, was able to retain, albeit with fluctuations, a dynamism that allowed an improvement in living standards and population growth, mainly fed by immigration. Although the industry was still weak, it offered a substitute for textile imports. The textile industry included the wool sector, which worked with domestic raw materials, and the cotton sector, which relied on imported raw materials. Both industries imported other raw materials as well as capital goods. The 1929 crisis deeply affected Uruguayan industry as a whole. The textile industry was particularly hit, due to reduced imports of capital goods as a consequence of the fall of exchange rates, and shrinking of the domestic market, the main destination of textile production. The textile industry recovered by the mid-1930s and continued its expansion process, enlarging its relative weight in manufacturing as a whole during the whole period of import substitution. From the early 1950s, the textile branch stagnated. Due to the crisis of the import substitution model, by the mid-1950s Uruguayan manufacturing industry faced a persistent stagnation. The crisis was triggered by the decrease in export income as a consequence of a fall of exchange rates. This mainly affected the textile industry, which suffered a reduction in state protection

Map 20.1

Textile regions in Uruguay

The evolution of the Uruguayan textile industry           Figure 20.1

Textiles as percentage of total value added in manufacturing, 1900–2000. Source: M. Camou and S. Maubrigades, El desafío de la

productividad en la industria “tradicional” uruguaya (Montevideo, 2006).

margins. The branch started to decline in comparison to other industrial branches. In a context of falling domestic and foreign demand, weak protection and increasing domestic prices for raw materials, the branch also lost competitiveness in foreign markets. The industry’s ability to compete relied on the low costs of its raw materials, and when this factor disappeared the restrictions on access to technology and high production costs made most of the industry unattractive. On a global scale, the textile industry underwent a relocation process after Second World War, by concentrating in lower income and lower wage countries. The result was that the greater the per capita income of a country, the smaller the relative weight of the textile sector in manufacturing industry became. Table 20.1 Percentage share of the textile industry in manufacturing, Uruguay, 1960–1982 1960–1971

14.1

1972–1976

12.1

1977–1980

8.7

1981–1982

6.5

Source: L. Sierra and R. Irigoyen, Uruguay: competitividad externa del complejo textil lanero (Montevideo, 1986), p. 84.

513

The Ashgate Companion to the History of Textile Workers In the 1970s the branch had a new boost, growing at a rate of 50 per cent over the rest of industry. The textile sector grew in an absolute sense, but its relative weight in manufacturing industry value added decreased, because of the high wage component in production costs. Wages fell drastically during the period of military dictatorship and dissolution of unions. Nevertheless, the Uruguayan textile industry still continued to be very important in the 1980s, when it was the third largest sector in Uruguayan industry, measured in value added. However, from 1981–1982 the economic recession hit the textile industry harder than any other industrial sector.

Background With European industrial development in the nineteenth century, the demand for textile fibres grew, stimulating wool production in peripheral countries. In the 1860s the interruption in cotton export from the United States during the Civil War added to this demand. With the resumption of US exports, other countries supplying fibres faced an overproduction crisis and a lack of demand. In this context, in Uruguay, private entrepreneurs proposed to create a textile industry capable of domestically processing the raw material produced. In the 1870s, passion was aroused during the debate on the need to industrialize the abundant wool production of Uruguay, exported mostly dirty to Europe. This production came to Uruguay somewhat later than to the rest of southern Latin America. Wool refining only started in the 1860s, in reply to international demand. Previously there had only been some handicrafts and individual enterprises, as for example a family workshop for wool knitting founded in 1837 with European immigrant capital. Due to civil wars in the countryside, these had only a short existence, which frustrated attempts towards intensive sheep breeding. Others tried to base a textile industry on imported raw material, or on other commodities such as silk. All these enterprises were also frustrated by barriers set by the government which profited from taxes on the imports of finished textile goods.

Early textile industry In 1897, with the help of protectionist laws that had been introduced, the idea of a domestic textile industry was realized with the establishment of the textile factory ‘La Victoria’ by Lorenzo and José Salvo. With the foundation of the first wool knitting factories, and the boost that the establishment of the first spinning mill in 1905 gave to the sector, the textile industry experienced a production growth of 39 per cent between 1900 and 1913. This growth was reflected in the number of workers, which had reached a total of 1,408 workers by 1913, 6 per cent of manufacturing industry as a whole. 514

The evolution of the Uruguayan textile industry The First World War and the first post-war years meant a great boost to the branch, in spite of problems with the supply of capital goods. European production stopped and the war brought into existence an important export market for the country’s textile products, which supplied countries in both America and Europe. Additionally, the domestic market became more important in this period, through substitution of textiles that previously had mainly come from Europe. The second half of the 1920s was a conflictive period for the textile industry. With peace restored, Europe, having reconverted its industry, went after the recovery of its previous markets, thus putting pressure on Uruguay’s local market. The import of woollen and cashmere cloths that amounted to 769,424 kg in the period 1915–1920, rose to 1,108,165 kg in the period 1921–1925. In addition, Uruguay’s early social laws on the working day, wages and workers’ insurance made labour more expensive than on the European Continent. The cost of labour was above the Italian or German level, with a cost of living higher than the European and a working week of 48 hours. At the same time the costs of buildings and machinery were 40 per cent higher and credit was scarce. By the second half of the 1930s, the textile industry had entered a period of rapid growth. In 1930 the textile sector had 212 workshops against a total of 6,570 for the whole manufacturing industry, employed 5 per cent of industrial workers and generated 4.3 per cent of manufacturing value added. By that time the textile branch added 45 per cent in value to raw materials. Import substitution of wool textile products became an alternative market for wool producers. In the period 1929–1933 5 million kilograms of washed wool was processed, and between 1934 and 1938 this number jumped to more than 20 million. The main buyers of textile products were Belgium (36 per cent of the total) Germany (24 per cent) and Italy (20 per cent).

The period of import substitution While the earlier period been characterized by rising exports, the quarter century between 1930 and 1955 was a dynamic period for the domestic industry in the framework of the process of import substitution. Within this phase of continued growth of the textile industry, three periods can be distinguished: the world economic crisis of 1930–1938, the period of the Second World War and the immediate post-war years, and the last one, which saw the beginning of the crisis of the sector and lasted up to 1955. The first period was marked by trade barriers as a consequence of the international crisis. Economic measures were taken in order to mitigate the effects of the crisis 

The eight hours law was promulgated in 1915 and the labour accidents insurance law in 1914.  Eduardo Acevedo, Anales históricos del Uruguay (Montevideo, 1934), pp. 342 and 343.  Raúl Jacob, Breve historia de la industria en Uruguay (Montevideo, 1981), p. 112. 515

The Ashgate Companion to the History of Textile Workers and to protect domestic industry. The annual growth in manufacturing growth in those years was around 2 per cent. The textile industry showed an exceptional expansion, with an average rate of 17.3 per cent between 1930 and 1938 and an increase in industrial product share to double that registered in 1930. Within the textile sector, wool was the main sub-branch due to the comparative advantages of locally produced raw wool. As there was no spinning industry, the cotton industry had to rely on imported yarn, which led to a much slower growth. It should be pointed out that even if much of the production was intended for the domestic market, industrial exports also grew, reaching 4 per cent by 1930 and 20 per cent in 1938. Growth also took the form of the foundation of new enterprises, like the cotton firms S.A. Fabril Uruguaya in 1930 and Textil Uruguaya S.A. in 1932, and in 1933 the wool firm ILDU. In 1936 Campomar y Soulas, the country’s largest textile enterprise employed 2,000 workers in its plants. The second largest textile firm was ‘La Aurora’, owned by Martinez Reina, which employed 522 workers. The sector as a whole counted 24 textile factories with more than 50 labourers, with a total of 4,664 employed workers. As a hundred textile factories and 5,835 workers were counted in the Industrial Census of the same year, we can see that the 24 per cent largest factories occupied 80 per cent of the workers. The second period to be highlighted is the Second World War and the immediate post-war years, in which manufacturing production slowed down to an average annual growth of 2.2 per cent, due both to the problems created by the war and the domestic economic situation. To the protection already enjoyed by the domestic industry, protection resulting from the military conflict was added. After having overcome supply problems at the beginning of the war (inputs, raw materials, machinery, spare parts), the most important expansion in textile history took place. The growth rate in those years was double that of manufacturing industry as a whole. Sales in the foreign market of washed wool, yarn, fabrics and, on a smaller scale, tops (rolls of washed and carded wool) increased and domestically the import of cotton fabrics was substituted by domestic production. Between 1936 and 1948 production increased at an average annual rate of 14 per cent. Important investments in machinery were realized (19 per cent annual increase), which exceeded the growth of the amount of labour employed (8 per cent). Cotton textiles production had been later in establishing itself than the wool sector, but experienced a stronger growth in the years between 1936 and 1948, with an average annual rate of 21 per cent. Domestic cotton products replaced imports, both in yarns and cloth. A strong investment in machinery (26 per cent annually) mainly took the form of new weaving and spinning mills. Lastly, the period from 1947 to 1955 showed the largest growth in domestic manufacturing industry, with an annual growth rate of 6.3 per cent supported by the implementation of a redistributive income policy and ample possibilities for import substitution. The textile industry maintained an annual growth of 6.4 per cent, with an increase of 35 per cent in the number of factories, 60 per cent in the number of workers occupied and 75 per cent in the gross value of production.

516

The evolution of the Uruguayan textile industry Market liberalization was promoted after the war and this provoked an increase in competence on the part of the suppliers to the domestic market, which had increasingly to compete with imported fabrics. Imports doubled between 1944/46 and 1947/49. This development was reversed in 1951 with a new protectionist boost, partly fostered by a devaluation of the Uruguayan currency, favouring the cotton as well as the non-export wool sector. The exporting wool sector, once protection of exports was re-established in 1947, showed a steady growth between 1949 and 1953. The main item was tops, supported by a differential exchange rate set in 1949, which in five years achieved an increase of production by a factor of 22.

The crisis of the textile sector and its transformations The period between 1955 and 1971 showed stagnation in industrial production as a whole, with an annual growth rate of only 1 per cent. The textile industry, which had led development in manufacturing in previous times, now registered a negative annual growth rate of nearly 2 per cent. In spite of this fall, it remained important within manufacturing industry as a whole, occupying third place after food and beverages. The main cause of this decline may be found in the stagnation of wool production in the second half of the 1950s and its further decline in the following decade. While in the period 1950–1955 wool exports amounted to 78,000 tons, in the following five years they reached only 64,000 tons. But, under the protection of export refunding policies applied since the mid-1960s, major changes took place in the structure of these textile exports, favouring the export of products with larger value added. Thus the relative importance of tops, woven goods and yarn among the total volume of wool exports went up from nearly 22 per cent in 1950/1959 to 36 per cent in 1970. However, policies in support of exports were not sustained. This was even more so by the end of the 1980s when this policy was abandoned and a reduction in tariffs for the textile sector was applied, conforming to the fashion for more open economies. Textile production was one of the leading branches of industrial development in the 1970s, showing a growth of 50 per cent above the figure for manufacturing industry as a whole. Until 1976 it continued to have a sustained growth (3.2 per cent annually), relying mainly on tops production. The following period also showed growth but it was based mainly on wool and cotton spinning and weaving, with an annual growth of 18 per cent and 9 per cent respectively. In spite of this rapid economic growth, textile production became relatively less important in manufacturing production as a consequence of the fall in value added. This was a consequence of the labour-intensive production in this sector and the relative decline in wages in those years. Besides, there was a 26 per cent decline in the number of workers employed between 1968 and 1978.

 Luis Bértola, La industria manufacturera uruguaya, 1913–1961 (Montevideo, 1991), p. 213. 517

The Ashgate Companion to the History of Textile Workers By the mid-1980s, the textile branch was still in third place (9 per cent) within manufacturing industry as a whole, now behind food (20 per cent) and oil by-products (14 per cent). This is explained, by the high share of the gross value of textile production. But in the textile industry value added per product unit (50 per cent) was above the average in manufacturing industry (43 per cent). Important technological innovation took place in the textile industry, at least in some items, after a long period of stagnation in investments. Thus, according to research carried out in the 1980s, an 80 to 100 per cent renewal of machines was reached in workshops manufacturing tops. The textile industry also maintained its importance in the labour market, occupying the second place in industries measured by the number of workers employed, after the food industry. Of the direct employment generated by the manufacturing industry, 12 per cent was in textiles. Exports showed a greater relative weight of textile products with a smaller industrial value added. Thus textile exports were mainly (85 per cent) composed of washed and carded wool and for the rest by yarns and fabrics. Due to the high value of wool compared to other textile fibres, it was exported mainly to richer countries, especially in the European Economic Community.

Organization of production The organization of the labour process in the Uruguayan textile industry was Taylorist from the very start, in the sense that it was marked by a high degree of work division, a repetitive character and low qualification of tasks and the use of technologies intended to save time and make work more efficient. At the same time, methods of social discipline and constraint, of settlement of labour and training and replacement were used with the aim of achieving increases in productivity. In Uruguay, little historical research is done on the evolution and organization of labour and it is not easy to find information on this subject from the available sources. Therefore, company archives are important sources, especially for the very early periods. In the two main textile companies in the country in the first half of the twentieth century, Campomar & Soulas and La Aurora, we can trace through the payroll the increasing complexity of sections and the increase in the number of occupations. In both companies the mills had been specially built for textile production. They had separate rooms for each manufacturing phase, distributed according to the sequence in work process. In the La Industrial plant belonging to Campomar & Soulas, inaugurated in 1907, the different sections were located around a central courtyard, which was closed with gates. These gates were only opened during the change of shift, thus forming a spatial separation between work and everyday life. The plant of 60,000 square  Graciela Sapriza, Los caminos de una ilusión (Montevideo, 1993), p. 55. 518

The evolution of the Uruguayan textile industry metres, located in a small town, some 200 km from the capital city, was the main local source of jobs. It developed a strong link with the local population, offering social services to the local community. Campomar, as well as most Uruguayan textile companies founded in the first stages of industrial development, worked with a system of full vertical integration. All wool processing from its raw state up to the manufactured fabric ready to be sold was done inside the plant. This method of production implied huge enterprises with an important concentration of personnel and diversity of tasks within the factory. Campomar started with 170 looms. Some years later, in 1939 and 1945, two publications by the company speak of 200 looms in this same plant. In 1960 the total number of looms had grown to 276. In 1948, an important renewal of looms took place. The most important innovation, for machinery bought during the period, seemed to be the progressive substitution of mechanic by automatic looms (even if in 1960 mechanic looms still formed the majority) and the acquisition of carding machines for tops. Both innovations made it possible to increase qualified personnel productivity. The maintenance of the industrial plant was in the hands of mechanical and electrical workshops and departments of bricklayers and painters, water purification, steam and electric energy generation, and machinery repair and construction. We have compared the occupation structure of Campomar in three moments of its history: at the beginning (1918), in the time of the first expansion (1934) and in the phase of greater incorporation of technology (1951). Our division over the different groups of occupations was based on the tasks of the worker and the required qualification, controlling for the income range of each group. Among the qualified workers we find occupations such as carpenter, electrician, bricklayer, industrial fitter, maintenance workers and also qualified workers who were part of production process in tasks which required qualification such as warper, stoker, dry cleaner, machine operator and milling machine operator. When we observe the distribution of workers’ activities, we note that as time progressed there was a trend towards more skilled personnel. While in 1918 workers, including all non-skilled workers, constituted 62 per cent of the total, in 1951 the figure was 26 per cent. The flip side of this was a significant increase in percentage of skilled workers and labourers and apprentices. Labourers and apprentices are a low-income group with great social mobility. If we study their path in the company we see that most skilled workers were hired first as labourers or apprentices, and later on obtained their training within the company. The group

 Raúl Jacob, Breve historia de la industria, p. 74.  Newspaper El Día, Extraordinary edition of industrial advertising, 25 August 1939 and Empresa Campomar, Dejemos hablar a los hechos (n.p., n.d.).  Forms from the regional survey on textile industry by CEPAL in the Campomar Archive.  In the workshops of La Industrial parts of the machinery, such as washing ‘trains’, were constructed. 519

The Ashgate Companion to the History of Textile Workers of weavers slowly diminished in size during the period, due to the introduction of automatic looms, which increased productivity per worker.

  











  

:RUNHU

/DERXUHUDQG DSSUHQWLFH

:HDYHU

6NLOOHG ZRUNHU

Figure 20.2 Labour force structure in La Industrial, 1918–1951. Source: Company Archive Campomar & Soulas. There may be several causes for this process of specialization. More and more modern machinery used in the production process meant that activities became more specific per worker, some workers only weaving, others maintaining the looms and so on. The spread of Taylor’s ‘scientific organization of work’ meant a hierarchic and bureaucratic, pyramidal and stratified organization. The decisions were made at director level, but were transmitted step by step through the different levels to the whole personnel. Edwards points out that while in the early stages of the Taylorization process the trend is to workers’ homogeneity, as the process goes on, big companies tend to adopt a greater differentiation between workers in hierarchic scale, thus constituting a more impersonal control mechanism.10 The average number of years of work in the company (seniority) is distributed in a different way according to activity. Skilled workers and workers with supervisory positions were clearly over-represented among workers with seniority in the company. This trend is clearly visible when we look at the work history of the workers with seniority in the company. Table 20.2 shows the previous activity of workers who occupied the position of foreman, weaver or skilled worker in 1949. Thus it may be noted that there were what we could call ‘entrance posts’ to the higher ranks in the factory. The position of foreman is occupied by workers who have been promoted to that rank from other positions, such as weaver, skilled workers and labourers. Weaver in 42 per cent of the cases is a position maintained at least since 1930, but 50 per cent of them have been promoted to that category.

10 Richard Edwards, ‘Conflicto y control en el lugar de trabajo’, in Luis Toharia (ed.), El mercado de trabajo: teorías y aplicaciones (Madrid, 1983), pp. 141–157, at p. 153. 520

The evolution of the Uruguayan textile industry Table 20.2 Work histories of skilled workers, Campomar Company, 1949 Occupation 1949 Occupation 1930

Foremen

Apprentice

Weavers

Skilled workers

4%

Worker

35%

30% 58%

Labourer

60%

11%

Weaver

20%

42%

Skilled workers

20%

8%

12%

100%

100%

100%

Total

Note: Only workers who were employed in 1930 and 1949. Total cases: 135. Source: Company Archive Campomar & Soulas.

In the case of skilled workers, a great percentage belonged to the category ‘labourer’, which stood a better chance of being promoted. This pattern of occupation and promotion in working positions shows a strongly developed internal labour market. The isolation of the place where the factory was located and the fact that it was the main source of jobs in the region added to the strength of the internal labour market. According to the requirements of a more bureaucratic control, which divides more precisely the spheres of conception and programming from execution, and requires more intermediation between these two fields, the proportion of administrative and managing personnel increases. Among the managing and administrative personnel, foremen had more seniority, namely on average 15.4 years in our sample. There existed a previous category called ‘apprentice foreman’. A worker had to have fulfilled a period of at least 1,500 working days in this position before he could be promoted to foreman. The average number of foremen per worker fluctuated. In 1930 there was one foreman per 22 workers. In the early 1930s, when the number of workers increased greatly, this number doubled. By 1951, it had decreased again to one per 21 workers. Important changes took place in the relationship between entrepreneurs and workers. In 1943 the Salary Councils (Consejos de Salarios) were established. These were a collective bargaining forum, in which entrepreneurs, workers and the state took part. The Salary Councils acted also as a conflict conciliation mechanism. They were established in a period of increasing union power and social legislation. The Textile Workers Union (Unión Obrera Textil), created in 1940, and in particular the textile salary council of La Industrial obtained increases above wages common

521

The Ashgate Companion to the History of Textile Workers in other branches. At the Campomar company real salaries increased and wage differences between workers were reduced.11 However, an increase in the wages of supervising personnel enlarged the gap between that group and the workers. In 1930, foremen earned 37 per cent more than workers and in 1949 this difference had gone up to 74 per cent. The strengthening of controlling positions coincided with the moment – in the early 1950s – when the company and the branch as a whole entered a period of stagnation and crisis. In Campomar, the minutes of the Board of Directors illustrate the increase in the costs of salaries and the complications arising from the workers’ social struggle. The economic crisis of the mid-1950s and the end of the industrialization period of import substitution marked the beginning of a progressive dismantling of organized labour relations with central bargaining between organizations of workers and employers and of a deterioration of working conditions and of workers’ quality of life. The military dictatorship (1973–1984) brought a general fall in real salary and an increase in unemployment. This led to a reappearance of neo-artisanal ways of textile work. These consisted of networks of informal homeworkers who did piecework for companies that sold their production in the domestic as well as foreign markets. This sector employed up to approximately 7,000 women, an important number considering the size of the Uruguayan labour market.12 At the same time, by the mid-1970s new machinery had been acquired and existing machinery renewed in nearly all Uruguayan textile companies. The aim of these technical changes was to increase production capacity to face foreign demands. The sector that was more modernized was not production, but management and sales, through computation. The work process and the type of products manufactured did not undergo essential changes. The Taylorized work organization as well as the training required from workers did not change in most companies. A survey carried out among textile firm owners in 1991 showed that in 54 per cent of the companies studied, technical change had had a direct impact on employment, usually diminishing it.13 At the end of 1988 a new collective agreement was signed. For the textile sector it formally implied a reinstatement of collective wage negotiations. In fact, it created a legal framework for de-regulation of collective agreements in the textile sector. The new rules increased the freedom of the employers to modify production processes and labour requirements just as the unions were weakening. 11 María M. Camou, ‘Industrialización y trabajo: un enfoque de la relación salarial desde una empresa textil, 1922–1949’ (Master thesis, Universidad de la República, 2001). A thorough analysis is made of the evolution of wages per category and gender. 12 Marcos Supervielle, Nelson Argones, Ana López and Graciela Pratt. ‘Algunas hipotesis acerca de las transformaciones en los procesos de trabajo en el capitalismo’, Revista de Ciencias Sociales 4 (1990): 50–60, at p. 56. 13 Marcos Supervielle and Francisco Pucci. ‘Políticas de relaciones laborales e innovaciones tecnológicas en Uruguay. El caso del sector textil’, in Gisela Argenti (ed.), Uruguay: debate sobre la modernización posible (Montevideo, 1991), pp. 177–223. 522

The evolution of the Uruguayan textile industry

The workers Before describing the workers in the textile industry it should be mentioned that, at a national level, research work on the industrial sector has been devoted to the analysis and characterization of the growth stages, fluctuations and crises in the different branches that compose it and their evolution as a whole. In those studies, labour appears as a factor of production. They inform us on the development of the number of workers and the influence this had on production. The existing literature and sources seriously limit any description of labour market behaviour in the long term. Thus we have to present two successive views, which even if complementary, correspond to different levels of aggregation. We start with a discussion of the number of workers in the textile branch since the first census of the twentieth century, carried out in 1908. Then an analysis of labour relations at Campomar & Soulas is presented. We add this case study to compensate for the lack of information at the national level.

Textile workers in manufacturing industry Over the first half of the twentieth century, the share of textile workers in the labour market in manufacturing industry as a whole continued to grow. This was mainly due to the labour intensive character of textiles. The data at aggregate level shows that in the first decades of the twentieth century, an increase of the number of workers accompanied the consolidation of this sector in manufacturing industry as a whole. When both sales in the domestic market and the share of textiles in total export sales increased in the 1930s, the share of workers employed in textile rose further. Even though the textile industry clearly showed signs of crisis from 1955, the absolute number of workers occupied in the sector continued to grow for some time, perhaps as a consequence of the strictly regulated labour market. After the profound economic depression of the 1960s, the share of the textile industry in total manufacturing employment gradually declined to about 12 per cent. The continued fall of the absolute numbers employed notwithstanding, this figure was maintained until the end of the century. De-industrialization halved the number of workers in the textile industry in the 1990s, but hit other branches as hard.14 Small enterprises, with fewer than 20 people, where 65 per cent of employment disappeared, were most affected by the process of labour expulsion. Larger companies, of 100 people or more, were able to maintain employment at the level of 1968. Unfortunately, there is hardly any information available on the division of the labour force according to sex. Information is limited to the period up to 1937. In this period the participation of women in textiles was much higher than in manufacturing on average. 14 From 1982, Uruguay, as well as most other Latin American countries, suffered from the so-called foreign debt crisis, characterized by inflation, productive stagnation, fiscal deficit and so on. 523

The Ashgate Companion to the History of Textile Workers Table 20.3 Workers employed in textile industry and manufacturing, 1908–1997 Manufacturing Men

Women

Textile industry

Total

Men

Women

Total

%

1908

17,979

4,245

22,224

848

560

1,408

6

1919

32,167

4,705

36,872

572

350

922

3

2,263

5

6,487

11

1930 1937

46,204 43,987

13,759

57,746

2,479

4,008

1948

91,960

10,885

12

1955

135,311

21,854

16

1968

166,575

22,905

14

1978

130,068

16,746

13

1988

124,317

15,518

12

1997

51,712

6,314

12

Source: Banco de datos de Economía e Historia Económica de la Facultad de Ciencias Sociales, http://www.fcs.edu.uy/bd/banec/bancode.htm

In 1919, men occupied in the textile industry formed 2 per cent of the workers in industry as a whole. By 1937, that figure had risen to 6 per cent. Male participation thus tripled between 1919 and 1937. The share of the textile industry in the total of female labour force participation was 7 per cent in 1919, a figure already above the level men would reach by 1937. This number had grown by 1937 to 29 per cent. Women in textiles thus increased their share of the total female labour force fourfold. The Campomar & Soulas files allow us to deduce that a great deal of this increase of female occupation may be due to the crisis that the sector suffered in the mid-1920s. The companies’ strategy in this period seems to have been to reduce costs by using female workers, who earned only 45 per cent of male wages.

Textile trade unionism In the first decades of the twentieth century the state assumed an important role in economic life, communications and the expansion of the educational system. An advanced labour legislation, the secularization of private and public life and special protection for woman workers were introduced in this first phase. This social state favoured the urban sectors: a rising middle class and an industrial 524

The evolution of the Uruguayan textile industry proletariat, both swollen by a strong wave of European immigration. The deep economic crisis that began in 1913 resulted in an increase in consumer goods prices and unemployment. At that time the first textile strikes, for the eight-hour law, began. In Montevideo and Juan Lacaze, women took the lead. Within the labour movement at this time, an important debate took place between anarchists, who proposed direct action and a general strike, and socialists, who argued against this extreme proposal in the context of growing unemployment. Under pressure from the state and entrepreneurs, the first textiles unions proved not to be very strong. But although the mobilization in favour of it was not as large as expected, the eighthour law was approved unanimously. Between 1930 and 1951, with the creation of new unions, the first stable organization of textile workers, the Textile Workers Congress (COT), was formed. In this new phase of a social state, higher demands for better living and working conditions were voiced. The textile union was from the beginning a hierarchical structure that sought the protection and support of the state. Relations between workers and owners were peaceful at the time. Beyond the direct achievements reached by the union in this period – wage increases, and legal services and health insurance for members – it made a good general impression. An example of this was the favourable wage negotiations of the unions in the Salary Council. After the Second World War, besides the strong unions, many social, sports and cultural organizations were founded, which created a separate place for the working class. Towards the end of the 1950s, with the growth of the economic crisis and a strong increase in unemployment, the unions began to have a decisive role in the struggle with the entrepreneurs. The strike became a decisive tool against layoffs and lockouts. A radical change in the economic policy of the government, with the state now favouring the entrepreneurs, influenced the struggle between workers and employers in the Salary Council. Another difficult period in industry and especially in the textile sector started in 1960. Under these circumstances, in 1964, textiles unions from all over the country merged into a unified organization. This increased workers’ negotiation power, which did not radically change their economic situation but allowed for important achievements like the reduction of the gender gap and improvement of the wages of the lowest income groups. Towards the end of the 1970s social conflicts flared up. Strikes and factory occupations were the most common strategies of the union movement. The government assumed a repressive role, while taking drastic economic decisions, including freezing prices and salaries in 1968. In that year there were 446 day strikes, 264 strikes and 87 factory occupations organized by the union movement, while the textile workers’ leaders were strongly represented at the top of the National Workers Union (CNT). The sharpening of social and economic conflicts finished in a coup d’état in 1973. In answer to this, the working class called a general strike. The outcome was a fragmented and defeated labour movement. After the dictatorship and even up to this day the union movement has failed to reach the level of mobilization of the 1960s. In the 1990s the government imposed a liberal policy with increased income inequality, price liberalization and free circulation of capital. The country 525

The Ashgate Companion to the History of Textile Workers de-industrialized, wages were reduced and labour relations deregulated. The textile industry was one of the most affected sectors, with lockouts at the most important factories and mass dismissals of workers.

The workers of Campomar & Soulas The workers of La Industrial plant were indigenous, originating from the countryside or small settlements near the factory. This factory was characterized by its location in a small town, with a small number of inhabitants. The plant, which opened in 1907, constituted a pole of attraction for workers from different parts of the country. To this internal migration we may add immigrants coming mainly from Italy and Catalonia, some of them directly hired by the company in Buenos Aires or Montevideo. After the arrival of immigrants from some Italian regions, migratory chains to these localities came into being. Contemporary sources indicate that many of these immigrants were qualified workers. The first wage books show a majority of Italians and Spaniards holding the most qualified and best paid jobs in the factory. The personnel registers show also the existence of entire families working in the plant, as well as several generations working for the company. It has been possible to collect information on the participation of men and women in this company since 1918, when women formed 48 per cent of the labour force. Until the early 1940s this proportion was maintained with little change, later decreasing to 40 per cent in more prosperous times. The presence of children under 14 years old was notoriously high at La Industrial, where they formed 30 per cent of the labour force in 1930. Both women and minors of both sexes earned lower wages than men. Among minors, girls formed the majority: in 1933 they represented 54 per cent and in 1940 71 per cent of the workforce. The continuity of several generations of the same families working for the company shows a policy of the firm that privileged the hiring of relatives. This is shown in the printed personnel forms that workers had to fill in when they entered the factory. They were asked about women and children belonging to the household. For the whole of the textile branch, interviews held with former women workers emphasize that they transmitted the skills needed in the factory from mother to daughter. The real wages of workers in the company were maintained with small changes in the 1930s, which was a period when unions exercised little influence and production stagnated. From the early 1940s, before the introduction of the Salary Councils, workers gained strong wage increases. Both the gross wage and a number of wage supplements rose. The increase in real salary happened within a context of labour and social law advances, and was a consequence of the growing role of the state in the economy. The regulation of working relations continued to be intensified during this period, for instance through compensation for work accidents (1941), a compensation for dismissal (1944) and the establishment of labour exchanges in the different branches (1947). To all this, we may add the 526

The evolution of the Uruguayan textile industry introduction of a family allowance regime and the extension of social security to all of the occupied population. The average salary of women in the Campomar company was kept below the level of average wages of men. This difference, which had been 50 per cent in 1918 diminished to 20 per cent by 1950. In the early 1930s, a temporary increase in women’s wages took place. By the 1940s, and especially after the institutionalization of wage regulation, the gap between female and male wages decreased. The greatest wage equality between the sexes corresponds to a period of lesser female participation. This process suggests two non-exclusive explanations: when the wage difference between the sexes is less, the company prefers to employ more men and/or in a period of real salary increase women show a trend to withdraw from the labour market, leaving men to work outside of the household. The claim for wage equality between men and women, put forward by groups of workers since the beginning of the twentieth century in the different companies, saw its first success in the textile branch. This success may be explained by the union strength but also by the importance of female labour force participation in the textile branch and women’s activism. This conquest, a first in Latin America, took place in 1960 after a strike with mass participation. However, just as with other achievements of this period, it would be overturned after 1967 when collective wage negotiation was abandoned.

Relationship with the state Protectionist state policies towards industry in Uruguay, and in general in all underdeveloped countries, basically used two kinds of instruments: customs tariffs and exchange rate differences. The pre-1930 development of industry has been an object of research and debate in Latin American historiography since the 1970s. One of the axes of the discussion was the relationship between industrial development and protectionist state policies. Protectionist measures in Uruguay during this period were generally applied at critical moments with the aim of increasing the tax revenues of the state. The tool used, taxes based on appraised values, had a stabilizing effect on both industry and state finances.15 This policy favoured industries with a domestic market, but it increased the prices of exports. The first measures protecting the textile industry were taken in the last quarter of the nineteenth century. In 1875, the first decree was passed which aimed to facilitate and promote the establishment of textile factories in the country. In 1887 an act exempting wool cloth factories from taxes was passed. The 1888 Customs Act, which exempted machinery for the industry in general and certain kinds of raw materials required for the textile industry from import taxes, was another measure 15 As imported commodities had a fixed value the degree of protection had an effect inverse to the price. 527

The Ashgate Companion to the History of Textile Workers favourable to the establishment of the industry. Nevertheless, only in 1898, were the first tariff grants for the manufacture of wool threads and fabrics, washing and dry cleaning given to Salvo Hnos. The tax facilities dating from 1875, 1887 and 1888 were projects that still had not been realized by 1898. The many state attempts to further a domestic textile industry in Uruguay failed because of the lack of capital, skilled workers, entrepreneurs and markets. At the beginning of the twentieth century the Uruguayan economy was growing steadily, based on the export of agricultural products. This made it possible to pay for imports for a growing population. In 1912 an act was passed to reduce import tariffs for raw materials and other materials needed for production. The effects of this new provision would not be felt during the years of the First World War and world crisis. Prices of import products rose, but in the 1920s, this promoted the growth of new industrial sectors. Among these, as noted before, was a cotton industry based on imported yarns. The protectionist policy had contradictory effects on the textile sector. Higher import rights made it easier for Uruguayan products to compete in the home markets, but the better exchange rate of the Uruguayan peso with the British pound meant reduced prices for fabrics and wool, of which Great Britain was the main supplier. After the 1929 crisis economies became closed and production was restructured for the domestic market. The industrial sector was favoured by several economic measures that promoted manufacturing production. Among these were regulation of foreign trade, exchange control and the creation of a state petrol refinery. At the same time, devaluation of the peso acted as an additional means of ‘natural protection’ of domestic products against imported ones. During this period, Uruguay followed the same trend as many other Latin American countries. It used the exchange rate as a way of protection and promotion of the industrial sector. Preferential exchange rates were established for the import of raw materials and supplies for the domestic industry and export were stimulated, preferentially of products with more value added. These measures contributed to the growth of the sector in first instance but at the same time they jeopardized its long-term development, as these forms of protection might not be continued forever and Uruguay did not produce capital goods. The strong growth of the textile sector during this period was fostered by several protectionist measures introduced in 1931: temporary import prohibitions on a list of articles which included suits and other textile articles, a rise in tariffs to 48 per cent on merchandises competing with domestic production, preference for domestic products in state acquisitions and tax exemptions on raw materials and other requisites for the industry. The first measure was later changed into an additional increase of import rights of 50 per cent. After a decade of abundant protectionist legislation, in 1941 a specific organism was created the main task of which was the centralized distribution of foreign currency. During this period, economic policies aimed at stimulating import substitution were strengthened and redistributive policies intended to strengthen the domestic market were implemented. The most important of these measures was the regulation of salaries and consumption prices. Together with import 528

The evolution of the Uruguayan textile industry control mechanisms and the promotion of investment in machinery for the textile industry, several policies to promote export diversity were implemented. Even if this happened in a discontinuous way and with little success, textile exports still found a favourable environment to develop in. Textile exports started during the Second World War, the main exported items being yarns, wool fabrics and tops. During the war, textile exports were favoured by a special exchange rate. The dollar value for manufactured products was above that for dirty wool, on average 10 per cent higher for tops and 17 per cent higher for finished textile products. Textile manufacture exports started to have a significant weight, but their evolution would always be conditional on state policies to promote the sector. The immediate post-war period saw a transition in which international markets were unstable, affecting industrial export sectors, among them the wool textile sector. Production intended for the domestic market was also threatened by the renewed activity of the international market. By the early 1950s, a new set of protective measures for the industry gave new strength to the textile sector. This new expansion stage was almost exclusively based on the export of tops. Uruguayan tariff protection met both domestic opposition, from wool exporters, and foreign opposition, from English and North American importers who complained of the high degree of protection which distorted the world market. This allowed Uruguayan tops to be sold at prices even lower than dirty wool. As a consequence of this pressure, the percentage of free exchange assigned to tops was again reduced in the early 1950s. The industrial policy of the early 1950s – even if it had more discretionary orientation than in the 1930s – included export subsidies to products with low value added but did not affect investment decisions. In this context, some industrial sectors – among them tops exporters – realized a short-term growth but could not avoid the structural crisis that had started by the middle of the decade.16 The industry, during this period, grew under state protection. When by the mid-1950s a strong fall in the export value of the main agricultural and cattle products and terms of exchange took place, state resources were rapidly reduced and industry protection policies began to be dismantled. The economic crisis was also evident in smaller demand in the domestic market. By the end of the 1950s state policies had started to change from import substitution towards a liberalization of the economy. This process was not linear, but had several advances and backlashes, as a consequence of the conflicts of interests among the actors: several industrial groups, agricultural and cattle breeding exporters, workers, the state and so on. Uruguay stands out within the Latin American context as one of the countries that started this process earlier and also continued longer to search for alternatives.17 In 1959 multiple exchange rates and import controls were eliminated, but charges and tariffs were continued. By the mid-1960s, a regime of promoting industrialization and wool exports had again 16 Bértola, La industria manufacturera, p. 234. 17 Henry Finch, 1974–1997: Towards a New Economic Model (Liverpool, 1996). 529

The Ashgate Companion to the History of Textile Workers been created and a system of reimbursements to textile exports was implemented, which would benefit the textile industry. State reimbursements were promoted and the aim was to give priority to export products that had to some degree been processed rather than to exports of dirty wool. Later on, during the first years of the dictatorship that ruled the country from 1973 to 1984, a model of economic openness and export substitution was advocated. This strategy aimed to promote the development of non-traditional industries to replace, for the first time in the country’s history, refrigerated meat and raw wool as the main exports.18 In the framework of this model, the textile industry specifically benefited from the extension of exemptions to machinery imports, bans on the export of dirty wool for several seasons in order to ensure supply to the industry and reimbursements ranging from 33 per cent to 48 per cent for products with greater value added.19 The competitiveness of Uruguayan textile products was based on this reimbursement regime and other protective measures. In the case of products with more value added, such as suit fabrics, competitiveness was assured only through reimbursement, for when the cost of industrial processing was added, the comparative advantage of cheap raw material disappeared. The policy of trial and error continued and by the end of the 1960s a new turn in favour of traditional cattle breeding had started. Measures to promote nontraditional exports were progressively de-activated. More significantly, the state started to work in this period on the mechanism of the exchange lag which made domestic costs higher. At the same time, the 1981 protectionist policies of the European Economic Community, the main market for tops and Uruguayan plane wool fabrics, had a negative impact on the textile sector. In the following years, in the context of a state policy primarily geared to paying off foreign debts, a return to traditional exports took place and protectionist measures for the textile sector continued to be dismantled. In summary, state policies towards the sector, if looked at in the long term, were characterized by a great instability, up to a lack of consistency in the tools used. To this may be added the lack of interaction between domestic policies and foreign market conditions, which placed the sector in situations where export promotion combined with restrictions in foreign demand and vice versa.

18 Inés Moraes, La producción de lanas en el Uruguay contemporáneo: una visión de largo plazo (Montevideo, 2003), p. 28. 19 Reimbursement to domestic tax returns is a practice accepted by GATT. In the case of Uruguay, the reimbursement regime does not require a previous assessment of the amount of taxes paid. These are therefore, in effect, subsidies. 530

21 USA: shifting landscapes of class, culture, gender, race and protest in the American Northeast and South Mary H. Blewett

During the period of American colonization (1620–1760), domestic production of linen and woollen textiles, dependent on the development of flax-growing and sheep-herding in the colonies clustered along the Atlantic coast, complemented the importation of fabrics of all kinds from Britain. Maritime traders could exchange New England lumber and fish for raw cotton grown by African slaves in the West Indies. Mercantilist policies by the British Empire in the seventeenth and eighteenth centuries discouraged the development of competitive industries in the colonies. Male-dominated artisan trades in textiles faded away in the New World, while spinning wheels, handlooms and carding tools operated by women and girls crowded colonial households. British efforts in the 1760s to impose taxes on its North American colonies provoked boycotts of English fabrics. The making and wearing of handmade cloth became an act of patriotic resistance both before and after the War of the American Revolution (1776–1783) and the War of 1812 between the New United States and Great Britain.



The author is greatly indebted to scholars working in social history and the new labour history since 1970.  Laurel Thatcher Ulrich, The Age of Homespun: Objects and Stories in the Creation of an American Myth (New York, 2001).

Map 21.1

Textile regions of the USA

USA: shifting landscapes

Origins of textile manufacturing in New England The British government at the urging of textile manufacturers forbade by law the emigration of skilled machinists and technologically advanced machinery to discourage competitors throughout the developing world. Still in the 1790s, English immigrant Samuel Slater copied the Arkwright water frame for spinning. American textile industrialization began at Slater’s spinning mill, supported by the crafts and metal firms along the Blackstone River in Pawtucket, Rhode Island. The ‘Rhode Island system’ of cotton manufacturing used young rural children under the closest supervision to tend water-powered throstle spinning frames. Rhode Island spinning mills ‘put out’ the yarn to be woven by women on handlooms in rural households then marketed the cloth in the region. The invention of an efficient cotton gin by Eli Whitney of Connecticut in 1793 greatly encouraged these developments. By 1813 Paul Moody of Waltham, Massachusetts, had successfully copied the English power-loom in the Boston Manufacturing Company’s machine shops located on the shallow, slow-moving Charles River. Backed by profits from shipping, real estate and commerce, a small clique of venture capitalists, called the ‘Boston Associates’, who first integrated production in Waltham, saw much bigger opportunities in the 32 foot waterfall on the Merrimack River at East Chelmsford, Massachusetts. There in 1820 they built the textile centre of Lowell. The ‘Waltham or Lowell system’ completely integrated the processes of cotton textile production based on water-powered machinery and technical advances in water turbines. By 1830, Lowell had become the big success story with the largest corporations in the United States, capitalized between US  $600,000 and a million dollars. New England mill treasurers quickly obtained political backing in Washington, DC, overcame ideological resistance to Old World models of industry and acquired a protective tariff. As an internal market opened in the Ohio River Valley and cottongrowing spread into the rich interior of the Southern states, the Lowell model of mass production of cheap, staple cotton goods inspired replication by many mills, large and small, located on rivers and streams throughout the Northeast. Within a decade, integrated production of cotton textiles faced serious competition and  Gary Kulik, ‘Pawtucket Village and the Strike of 1824: The Origins of Class Conflict in Rhode Island’, Radical History Review 17 (1978): 5–37; Jonathan Prude, The Coming of the New Industrial Order: Town and Factory Life in Rural Massachusetts, 1810–1860 (Cambridge, 1983).  For a critique of the use by historians since 1935 of the term ‘Boston Associates’ as if it were a corporate organization responsible for American industrialization, see Francois Weil, ‘Capitalism and Industrialization in New England, 1815–1845’, Journal of American History 84 (1998): 1334–55.  Robert E. Dalzell, Enterprising Elite: The Boston Associates and the World They Made (New York, 1993).  Alfred Chandler, The Visible Hand: The Managerial Revolution in American Business (Cambridge, Mass., 1977), pp. 20, 67–72; Charles Sellers, The Market Revolution: Jacksonian America, 1815–1846 (New York, 1991). 533

The Ashgate Companion to the History of Textile Workers falling profits. After 1820, investment capital from Providence, Rhode Island began to develop the village of Fall River in south-eastern Massachusetts, later the national centre of cotton print cloth production. Profits from the whaling industry promoted New Bedford, Massachusetts, as a fine cotton goods centre. New England lacked an immediate supply of labour to run textile factories. The Lowell mills recruited a rural workforce from the five states of the region, consisting of small numbers of skilled male mechanics and large numbers of female factory operatives. They travelled along an interconnecting system of rural wagons, stagecoaches, packet boats and the quickly developing early railroads of New England. By 1836, eight major Lowell firms with a total investment of US  $6.2 million employed 6,000 workers. Until supplanted by rail, transportation canals delivered bales of raw cotton shipped to the piers of Boston from the expanding Southern cotton plantations dependent on male and female slave labour. The Lowell system of integrated production increased the demand for cotton resulting in the rising price for African slaves, just as the English anti-slavery movement encouraged the British Navy to halt cross-Atlantic trading. Raw cotton became a major export commodity; by 1850 the cotton-growing South provided 75 per cent of Britain’s needs. In Lowell and Waltham, young, unmarried women operatives were housed in paternalist boarding houses and their behaviour regulated and controlled. Mill agents did not expect them to become a permanent factory proletariat but to receive good wages and treatment, spend a few years away from their rural homes, perhaps save some wages, and eventually marry. Their cultural backgrounds of rural life, New England womanhood, Protestant religion, public education and republican sentiments based on patriotism from the War of the Revolution made them a socially cohesive group. These cultural and political values bound female operatives together as the ‘daughters of freemen’ in labour protest, when they objected to cuts in piece-rates in the face of falling profits. The men and women in the rapidly industrializing Northeast (Maine, New Hampshire, Massachusetts, Rhode Island, Connecticut, New York, New Jersey and Pennsylvania) shared many grievances over wage reductions and a 14-hour working day as well as Protestant religion, artisan training, manhood suffrage and the republican heritage of the American Revolution. Female textile workers in Waltham protested as early as 1821 in ‘turnouts’, or work stoppages, often spontaneous and hard to sustain. The Lowell Female Labour Reform Association, the first American women’s labour union led by operative Sarah Bagley, continued strike activity into the early 1830s and lobbied into the 1850s for legal limits on the working day.10 But the divisions Thomas Dublin, Women at Work: The Transformation of Work and Community in Lowell, Massachusetts, 1820–1860 (New York, 1979) p. 20.  Chandler, Visible Hand, p. 20.  Norman Ware, The Industrial Worker, 1840–1860 (Chicago, 1924). 10 Helena Wright, ‘Sarah Bagley: A Biographical Note’, Labor History 20 (1979): 398–413; Terry Murphy, Ten Hours’ Labor: Religion, Reform, and Gender in Early New England (Ithaca, 1992). 

534

USA: shifting landscapes among men and women workers, native and immigrant, white and ‘coloured’, factory operative and skilled artisan, made unity difficult. The strikes failed, and many operatives simply returned to their rural homes. Mill agents began to consider a more permanent, docile workforce among the Irish immigrant families who had dug the waterpower canals and laid the granite foundations for the Lowell factories. By 1860, Irish immigrants represented 62 per cent of the workforce. A strike in 1859 pitted the Yankee ‘mill girls’ of Lowell against the rapidly growing numbers of Irish women operatives in the cotton mills. Cultural differences and religious bigotry undermined class unity as rural Yankee Protestants, clustered in the best jobs, broke the strike initiated by Catholic female throstle spinners.11 The exploitation of ethnic, religious and racial divisions by textile capitalists weakened labour protest. Yet Yankee female operatives continued to work in post-Civil War production but in fewer numbers and from more distant areas. In 1860 New England dominated American cotton textile production with 52 per cent of the largest mills and 75 per cent of the spindles. Massachusetts and Rhode Island remained the centres of regional production, especially in Lowell, Fall River and the Blackstone River Valley. Manchester, New Hampshire, developed rapidly, while Philadelphia became the largest cotton-manufacturing centre outside of New England. Of an estimated total of 122,028 operatives in 1860, 61 per cent were female and 38 per cent male.12 Woollen production also flourished in New England, but cotton centres usually had a diversified product as in Lowell and Lawrence, founded by the Boston investors in 1848. After the Civil War, Lawrence became the New England centre of worsted production.

Textile industrialization and cultural change in the Northeast Other early textile mills arose in Pennsylvania and New Jersey but few were as distinctive in organization as in New England. Paterson, New Jersey, became a centre of silk production in large factories with shifting populations of immigrant labour and no unions.13 The exception was Philadelphia’s proprietary capitalism, a factory system based on small firms initially owned individually or by partnerships that primarily employed British immigrant skilled workmen, produced various fine and specialty goods, and dominated local textile production well into the twentieth century. Philadelphia mills disdained bulk production of cheap textiles as unstable and unprofitable. Flexible batch production of fine goods (frequently 11 Dublin, Women at Work, pp. 145–164, 198–199. 12 Melvin T. Copeland, Cotton Manufacturing Industry of the United States (Cambridge, Mass., 1917), pp. 8, 11. 13 Herbert G. Gutman, ‘Class, Status, and Community Power in Nineteenth-Century Industrial Cities: Paterson, New Jersey’, in Work, Culture, and Society in Industrializing America: Essay in American Working-class and Social History (New York, 1977), pp. 234– 292. 535

The Ashgate Companion to the History of Textile Workers divided into specific processes) made by well-paid skilled union men under paternalistic managers offered both industrial peace and sustained prosperity if not market dominance.14 By 1880, 250,000 workers of English, Irish, German and native-born backgrounds crowded this most diversified industrial centre in the United States.15 By the 1850s, immigrant labour from England, Scotland and Ireland had transformed the workforce in many north-eastern cotton factories. Lancashire immigrants in Fall River and New Bedford, Massachusetts, ‘abounded as nowhere else’, while other British workers crowded textile mills in Rhode Island, New York, New Jersey and Pennsylvania.16 Men and women workers from different ethnic cultures with competing ideas about union organization, gender relations and political ideology – and who faced fierce opposition from textile capitalists – shaped production relations in northern industry. Weavers and mule spinners in the Northeast hammered out textile unionism in the cotton industry during multiple strikes between 1848 and 1894. English immigrants brought with them, depending on their skills and the timing of their emigration between the 1820s and the 1890s, various strategies and tactics that included Old World rituals of mass demonstrations. English-led activism emerged first in the early textile mills of Philadelphia. In 1848 textile workers organized and led by English immigrants struck in Fall River. During this strike, the weavers and other operatives joined the skilled mule spinners and mill mechanics to create a coalition of native-born and foreign-born workers to protest a surprise wage cut between 5 and 17 per cent. Introducing English customs of confronting overbearing employers in the streets of Fall River created a recurrent pattern of strike activity, which prevailed for the rest of the nineteenth century. The mill owners, who decided to crush this uprising by eliminating its leaders, also established policies followed consistently thereafter.17 Native-born textile workers invited British immigrants to lead the 1848 strike, but tensions between the two groups undermined their unity. English immigrant Thomas Norris emerged as the main leader, drawing on his experience in ‘old England’ where he claimed textile workers had successfully fought for their rights as industrial workers. Strike-breakers were to be labelled ‘knobsticks’, an English term quickly applied to all who refused to join the strike, while all knobsticks received in derision ‘three British cheers’: Hip, Hip, Hurray! The strikers kicked the doors of the mills’ counting houses and knocked on the windows. This 14 Cynthia Shelton, The Mills of Manayunk: Industrialization and Social Conflict in the Philadelphia Region, 1781–1837 (Baltimore, 1986); Philip Scranton, Proprietary Capitalism: The Textile Manufacture of Philadelphia, 1800–1885 (Cambridge, 1983) and Figured Tapestry: Production, Markets, and Power in Philadelphia Textiles, 1885–1941 (Cambridge, 1989). 15 Scranton, Proprietary Capitalism, p. 421. 16 Roland Berthoff, British Immigrants in Industrial America (Cambridge, Mass., 1953), p. 31. 17 Mary Blewett, Constant Turmoil: The Politics of Industrial Life in Nineteenth-Century New England (Amherst, 2000). 536

USA: shifting landscapes was unheard of: the sanctity of the feared counting house where the managers congregated had been violated. But many Yankee operatives strongly disapproved of these actions and confrontations with strike-breakers as they left the factories. After Norris was arrested, the prosecution charged him with leading an ‘infatuated rabble’ and creating riot in the streets of the city. The presiding judge insisted that while striking operatives did have the right to meet and combine, they possessed no right to obstruct other operatives when leaving or going to work. Forcible obstruction constituted riot. Endorsing a legal concept of conspiracy to riot and collective guilt for which the strikers were held responsible, the judge insisted that the act of each bound all. The jury found Norris guilty and jailed him. Many of the strikers, who testified for the defence, feared the blacklist (being on a formal list denying employment) and left town. Divisions among native-born and immigrant workers and the repressive tactics of mill owners continued to divide workers in the nineteenth-century textile industry. Other Fall River strikes in 1850 and 1870 led by English mule spinners demonstrated both the limitations and the possibilities in coalitions with native workers. As experienced textile workers from urban centres in Lancashire and Yorkshire, the mule spinners bitterly objected to the tyrannical overseers to whom they referred as ‘masters’, a Lancashire term for employers that indicated expectations of mutual obligations between the master and his ‘hands’: the operatives. No white American in pre-Civil War New England would voluntarily use the word master, although many complained about their working conditions as wage slavery and their overseers as slave drivers. The word ‘master’ applied to an American overseer would place white workers on a level with black slaves. Furthermore, the New England term ‘help’ for operatives offended English workers as servile and demeaning.18 The spinners’ organization challenged the New England mills’ quotations on the price of raw cotton and their profit margins in the cloth market, asking mill agents to open their account books to seek reconciliation as in Lancashire. This request to open the accounts was a far more potent act than pounding on the counting house door in 1848. The employers absolutely refused and disclosed nothing to the public, thereby controlling information to negotiate a fair wage for skilled male workers. No longer a sacred but intimidating centre of industrial growth, for the mule spinners the counting house had become the home of ‘unchristian despots’ and ‘unmanly sycophants’.19 The Fall River mills had copied the English self-acting mule to replace throstle spinning frames. But the agents claimed that one mule spinner could do his work without the assistance of a nearly adult ‘big piecer’ or apprentice spinner and a ‘little piecer’ as in Lancashire. In New England the mule spinner, often called a ‘barefoot aristocrat’ in England, would in effect do the work of two. Inexperienced ‘back boys’ aged between 8 and 12 years old became common in all New England mills as helpers to ‘piece up’ or repair the multiple strands of yarn spun from cotton roving 18 Blewett, Constant Turmoil, pp. 90–92. 19 ‘An Operative’, Fall River News, Feb. 20, 1851. 537

The Ashgate Companion to the History of Textile Workers during the back and forth motions of the huge frames. This change undermined both the mule spinners’ authority as men who supervised other adult males and weakened their claim that spinning was fit work only for strong, experienced men. Furthermore, the Fall River machinery ran faster and faster.20 Mill agents insisted that the capitalists who bought new machinery should reap the benefits of increased productivity. In response, the striking spinners angrily attacked their employers’ arrogant use of power. Their open, contemptuous hostility towards the mill owners and overseers challenged the general acceptance by Americans that industrial development would advance the mutual interests of employers and their workmen. This hatred inspired by the mistreatment of adult males touched both English and Yankee textile workers. Skilled immigrant spinners knew they were essential to mill operations, while open contempt for abuse by agents with political power and wealth undermined native-born belief in the rights of free-born men and manhood suffrage. Nonetheless, the manufacturers insisted on submission to their terms and used bribes, threats, evictions and dismissals even for those sympathetic mule spinners in other mill centres who donated to the Fall River strike fund. Skilled white men, native-born or immigrant, encouraged by their religious and political convictions insisted on decent treatment and fair wages, but the agents were undermining the meanings of working-class manliness. The 1850 strike represented a broad coalition of native Yankees and Lancashire mule spinners with connections in other textile centres. Still, the defeats in 1848 and 1850 meant New England textile workers remained unorganized. Immigrant workers sent donations to the 1853–1854 Preston, Lancashire strike fund, thus maintaining transatlantic ties and hoping some day to create throughout New England trade unions strong enough to counter strikebreaking and blacklisting.21

The impact and aftermath of the American Civil War, 1861– 1865 The United States and Britain came close to naval warfare during the political crisis over the American blockade of the South’s cotton supply. For four years, few cotton bales arrived in New England or in Lancashire. Meanwhile woollen mills profited from contracts to make Union Army uniforms. Most cotton mills suspended production, dismissed their workforces and warehoused their raw cotton or sold it off at high prices. The chief buyers were the Fall River mill agents who decided to gamble on a staple product for the post-war years: cheap, cotton print cloth. They anticipated the completion of the transcontinental railroad in 1869 and the 20 William Lazonick, ‘Industrial Relations and Technical Change: The Case of the SelfActing Mule’, Cambridge Journal of Economics 3 (1979): 221–262. 21 Blewett, Constant Turmoil, pp. 94–97. 538

USA: shifting landscapes opening up of the domestic market for cheap goods. Tremendous post-war urban growth sparked by massive industrial expansion and a resulting flood of European immigrants created demand for cheap fabrics. The freed people of the South as well as poor white families preferred inexpensive light cotton clothing to the mixtures of woollen and cotton or linen trash used for pre-war clothing. Demand leapt in 1865 and 1866, and profits in the print cloth soared. Competitors in Massachusetts, Rhode Island and Connecticut began to threaten Fall River’s control of this lucrative market. The cotton textile workforce in Fall River quadrupled between 1865 and 1875. By 1880, Fall River boasted 33 corporations running 150 mills in comparison with 11 corporations in Lowell. Fall River investors dominated this expansion (92 per cent) in contrast with the dependence of Lowell and Lawrence on Boston capital. Growth during the 1870s stimulated rapid mill construction in Rhode Island and Connecticut as well as south-eastern Massachusetts. By 1880, 57 per cent of the printed cotton cloth produced in the United States was made in south-eastern New England textile centres, well over half made in Fall River mills. Any unprinted textiles or ‘grey goods’ were shipped to New York and Philadelphia printers. Mills from Maine to Pennsylvania produced the remaining 43 per cent, but Fall River’s massive productive capacity dominated post-Civil War production, setting the price of print cloth on the national market and the regional wages paid to operatives. Lowell had been eclipsed as the leading textile centre in New England, while Philadelphia managers had no interest in the unpredictability and labour unrest of the emerging ‘Fall River system’. No city in New England grew so fast as Fall River, more than doubling its population between 1860 and 1870. By 1875 the city’s residents had nearly doubled again to 45,260 with 38 per cent foreign-born. The workforce in Fall River in the mid-1870s consisted of one-quarter native-born (25 per cent), one-third English immigrants (34 per cent), one-fifth Irish immigrants (21 per cent) and less than one-fifth French-Canadian immigrants (17 per cent).22 Fall River had the highest percentage of foreign-born workers in Massachusetts, 53 per cent; Lawrence was third with 45 per cent, followed by Lowell with 36 per cent foreign-born and New Bedford with 23 per cent. Among the female workforce, more wives worked in the Fall River mills than in Lowell or Lawrence. In 1869 the Lancashire mill owners cut wages by 10 per cent. To thin the labour supply, unions representing English weavers and mule spinners paid for the immigration of thousands of their members to the United States. In addition, by 1872 there were 3,646 French-Canadians living in Fall River; two years later 4,000 worked in the mills, but experienced mule spinners from Montreal could not get jobs in New England.23 Mule spinners opposed French-Canadians in their trade, and many others feared a post-Civil War tide of non-English speaking, Catholic, 22 Census of Massachusetts: 1875 (Boston, 1877), vol. 3, p. 49. 23 Philip T. Silvia, Jr., ‘Neighbors From the North: French-Canadian Immigrants vs. Trade Unionism in Fall River, Massachusetts’, in Claire Quintal (ed.), The Little Canadas of New England (Worcester, Mass., 1983), pp. 44–65. 539

The Ashgate Companion to the History of Textile Workers rural Quebecois who, as a ‘race’, would oppose a 10-hour day: man, woman and child. Thus a racialized Quebec people became a pawn to be used by employers against operatives. Table 21.1 Persons employed in cotton goods, Fall River, Lowell and Lawrence, 1875 Fall River Total employed

%

Lowell

%

Lawrence

%

14,216

100

9,960

100

8,585

100

Males

6,644

47

2,661

27

3,448

40

Females

7,572

53

7,209

73

5,137

60

882

12

315

4

169

3

Married females

Source: Massachusetts State Census, 1875, vol. 3, pp. 442–443.

During an 1870 strike, Fall River mule spinners, dominated by English and Irish-born, Lancashire-trained immigrants, tried to use deferential Lancashire methods including their status as respectable, reasonable men of skill to manoeuvre their employers into wage negotiations. The mill agents refused to negotiate and hired strike-breakers. Male and female weavers confronted the knobsticks with violence: yelling, jeering, kicking and pummelling strike-breakers with their fists and throwing stones and dirt clods. Peaceable, manly mule spinners controlled events, while weavers’ rowdiness served class politics. The mill agents challenged the manliness of mule spinners, claiming that in other textile centres young women ran mules easily. The spinners scoffed that they worked harder at their machines than anywhere else, walking 25 miles per day over 11 hours with only one young boy to help them: ‘a pretty good day’s work for any man’. Indeed few had the stamina to work out the full month of six-day weeks without staying home as ‘sick’ to regain their strength. A list of ‘sick spinners’ routinely filled the jobs of exhausted men. For the employer to define a spinner’s work as just beyond the extent of an adult male’s physical powers, an abuse the operatives called the ‘grind’ or ‘lashing the help’, undermined manly pride in strength and skill.24 Mill agents justified the presence of strike-breakers by citing market demand that determined the supply of labour and in turn set wages. If wages were too low, spinners should go to other mill towns, which demanded their skills. For the agents, any opposition to market forces represented interference with individual rights and ‘natural’ economic laws. But strikers blamed wage cuts on decisions made in Fall River counting houses – not by an impersonal market – decisions 24 Fall River News, July 29, Aug. 8, 18, 20, 1870; Massachusetts Bureau of Labor Statistics, First Annual Report (1871), p. 486 and Thirteenth Annual Report (1882), pp. 348–354. 540

USA: shifting landscapes that could be changed by negotiations. The managers refused to talk, sold their inventories, starved out the strikers and blacklisted the leaders.25 Other New England agents adopted similar policies. Protest strategies based on respectability and deference might have worked in Lancashire but not in Fall River or in any other New England mill centre. Between 1860 and 1910, total employment in American cotton manufacturing tripled, while value of product increased sixfold. The largest centres of New England production in terms of value of product were Fall River, followed by Lowell and Manchester, New Hampshire.26 During the late nineteenth century, British immigrant mule spinners and weavers in the Northeast struggled over the best form of union organization, strategy, ideology and objectives. Many historians view English immigrants as ‘invisible’ or easily acculturated into American society and regard the career of mule spinner Robert Howard, the leader of the National Mule Spinners’ Union in the American Federation of Labor (AFL), as typical.27 But the majority of weavers and many other workers opposed Howard’s exclusive trade unionism. They defended their heritage of Lancashire popular radicalism, including the ideal of a moral economy unbound by market forces and the goal of an amalgamation of all skill groups.28 The two Fall River strikes in 1875 and their regional impact provide examples. During the 1870s as Fall River capitalists attempted to dominate the American domestic print cloth market in the West, South and urban East, they wished to purge their English workers of ‘their chronic insubordination’. From the moment that Lancashire workers landed, the disdainful mill agents systematically challenged their customary measures of skill and strength. Fall River mills used the cheapest raw cotton and the best machinery, paid the lowest wages in the region and demanded ever more intense physical exertions from their operatives – especially from mule spinners – to produce massive quantities of inferior fabric, relying on the cloth printing process to conceal defects. They controlled the domestic market for print cloth by having the capacity to glut it with the cheapest possible goods. In turn, English workers held the ‘Fall River system’ in contempt for making ‘shoddy’ (flimsy cloth) using ‘shoddyite morality’.29 The Massachusetts legislature established a 10-hour law in 1874 and in the 1890s specified the length of a weaver’s cut on which piece-rates were calculated and regulated the fining of weavers by the mills. However, mill agents ignored these state regulations, targeted female weavers for fines, paid the small penalties when occasionally prosecuted and successfully defied labour reform. 25 Blewett, Constant Turmoil, pp. 184–189. 26 Copeland, Cotton Manufacturing, pp. 17, 28. 27 Charlotte Erickson, Invisible Immigrants: The Adaptation of English and Scottish Immigrants in 19th Century America (Ithaca, 1990); Isaac Cohen, American Management and British Labor: A Comparative Study of the Cotton Spinning Industry (Westport, 1990). 28 Neville Kirk, Labour and Society in Britain and the USA (2 vols, Aldershot, 1994); Clifton K. Yearley, Britons in American Labor (Baltimore, 1957). 29 Providence, [RI] Sun, Feb. 27, 1875; Boston Herald, Feb. 24, 1875. 541

The Ashgate Companion to the History of Textile Workers

Amalgamations or trade unions? In January 1875 the Fall River mills cut wages by 10 per cent and flooded the cloth market to undersell their competitors. The mule spinners and male weavers reluctantly accepted the reduction but faced a revolt among female weavers and some Quebec-born operatives. These rebels organized all-female meetings to shame their male co-workers into the only successful weavers’ strike in nineteenth-century New England. During two regional strikes in 1875, the first in January and the second in August, New England labour politics focused on the differences among mule spinners and weavers over the nature and direction of worker organization. The successful challenge of activist women to male leadership and the differing priorities given amalgamations of all operatives by the weavers and local craft unions by mule spinners produced intra-class conflict over the meaning of workingclass manhood and the desirability of female activism. The transnational culture of Lancashire provided the framework. In 1875 there were approximately 8,000 weavers and 2,000 mule spinners in the Fall River workforce. Women represented one-third of the striking weavers and half of the weaving workforce. Female operatives organized themselves across skill and ethnic lines, insisting that ‘all national differences’ end to preserve the general rights of operatives. Lancashire immigrants provided the leadership, but the women’s meetings included native-born Americans and Irish and FrenchCanadian immigrants. Their leaders demanded action to prevent recurrent wage cuts, denouncing conciliation and deference as cowardly and unmanly. Using their heritage of nineteenth-century popular radicalism, the Lancashire women leaders cited examples of the effectiveness of resistance – win, lose or draw – on the relationship between labour and capital. Women strikers directly challenged deferential manhood as inadequate, insisting on bold action and imaginative strategy.30 The initial issue of supporting the women’s strike in January, thereby tolerating their independent activism, created dissension among Lancashire men who labelled them ‘babbling Amazons’.31 Careful union men soberly weighed the possibilities of strike action, not misguided by emotional, rebellious Eves. But other Lancashire men agreed with the women strikers on the lessons of historic resistance. Before 1875, male weavers deferred to the leadership of the mule spinners, but they quickly joined the women’s strike, followed by the mule spinners, and other operatives. The strike spread to New Bedford and to the English immigrants in Rhode Island’s Blackstone Valley: Valley Falls, Central Falls, Lonsdale, Berkeley, Ashton, Pawtucket and Woonsocket. Support for the strike included Lawrence, Newburyport, Lowell and Taunton, Massachusetts. Prospects for a settlement brightened. The mule spinners of Lowell struck in March, while the print cloth mills of Lawrence promised their workers no wage cut. Operatives in Lonsdale, a centre of print cloth operations, were notified of an actual wage increase in April, 30 Fall River News, Jan. 14, 1875. 31 Fall River News, Jan. 18, 1875; Boston Globe, Feb. 22, 23, 1875. 542

USA: shifting landscapes while the agents in New Bedford and in Newburyport reversed wage cuts. Mill agents throughout the region stopped or rescinded pay reductions. The activities by the women weavers during the 1875 strikes significantly expanded their participation in labour protest. The wives working in the Fall River mills provided mature leadership for the women’s meetings as well as connections among working-class families. Single women were also prominent. Women strikers appeared regularly on public platforms, agitating at strike meetings and travelling to other textile centres to raise money. They urged the mule spinners of both Fall River and Lowell to organize young women being recruited to operate the first, very primitive ring spinning frames. This new technology would begin to replace mules by 1879. Weaver activism also encouraged 400 male and female FrenchCanadian operatives to support the strike in Fall River and in Taftville, Connecticut. The reputation of French-Canadian immigrants as docile, submissive and faithful only to their own proved false. The impressive unity in 1875 yielded results. In April, all strikers were rehired and wages raised. Despite the misgivings about female activism, once the battle was over the weavers controlled local labour politics. Their aggressiveness and unexpected victory had defeated the arrogant employers. Only the weavers’ leader who had advocated a region-wide amalgamation was blacklisted. As a result of the strike, Fall River wages became the standard for print cloth production in the Northeast. Quickly the weavers began to organize a regional association of textile operatives for a standard list of wages and to agitate for additional 10-hour laws, making New England the Lancashire of America. They announced their goal: ‘to establish a union in every village where the spindles revolve or the click of the busy shuttle is heard’.32 In July, the mill owners challenged the weavers’ ambitions by cutting wages to glut the depressed print market with even cheaper cloth. Angry weavers and mule spinners agreed to challenge their employers’ power to set wages based on the prospects of the cloth market. In doing so, they resurrected the Lancashire tradition of a moral economy that rejected their employers’ definitions of marketplace rules and the ideology of supply and demand.33 Instead, textile operatives would influence the price of cloth by withholding their labour and taking an Englishstyle late summer ‘long vacation’. Many New England mills had already closed down waiting for the market to rise. The operatives seized control of the timing of the work stoppage for late summer when their living costs were low. This defiant act denied the validity of a morally neutral market run by natural economic laws, which masked the dominant position of the mill agents. By early August, 34 mills and nearly 15,000 operatives squared off. The mill owners became absolutely determined to crush this unprecedented threat to their power at all costs. The fierce struggles during the ‘vacation strike’ in 1875 exposed the mule spinners and the male weavers to charges of recklessness and irresponsibility. 32 Blewett, Constant Turmoil, pp. 193–208, quote on p. 208. 33 John Bohstedt, ‘The Moral Economy and the Discipline of Historical Context’, Journal of Social History 26 (1992): 265–284. 543

The Ashgate Companion to the History of Textile Workers Hundreds of French operatives chose to return by rail to Montreal or left town for other New England mill centres. This became the pattern for French-Canadian workers during late nineteenth-century strikes. After four weeks of no work and no price increase in the cloth market, the operatives abandoned their ‘vacation’ and accepted the wage cut. They were beaten for the moment. But the Fall River agents were eager to demolish any hopes for an American Lancashire. Fearing that secret unions would flourish despite the anti-union contracts operatives were forced to sign, mill owners particularly wished to destroy Lancashire-style union discipline: ‘Up goes a hand and out goes the help!’34 They decided to lock out their employees for another month, threaten more wage cuts and starve them into giving up their unions. On 27 September when the mills reopened, the operatives flocked to the mill yards, willing to work but refusing to sign the anti-union contracts. Hunger and the knowledge that the mills would employ only the utterly defeated produced a response by angry operatives that recalled late eighteenth-century bread riots. These traditions shared only by Lancashire people threatened to alienate other uncomprehending ethnic communities. Thousands of disappointed workers marched to a nearby city park. A delegation, asking for relief from the mayor, returned unsuccessful. Hundreds of men and women strikers marched to City Hall, cheering and yelling ‘Bread!’ ‘Tyranny!’ Boys held symbolic poles on which were impaled loaves of bread, while an American flag upside down as a distress signal preceded a sign that read: ‘15,000 white slaves for auction’, topped with a loaf of bread. Lancashire workers were incensed at American pretensions to individual freedom in contrast with the servility thrust on them. The historic significance of the food riot rituals that directed the angry turmoil was clear to Lancashire people, but baffling to many others. A few recognized the revolutionary implications of Manchester-style bread riots in New England, while labour reformers appreciated the ‘extraordinary … English cool-headed control’ that disciplined crowd action. But others saw only mayhem, as showers of stones, bread and brickbats fell on arresting police. The violent spectacle of the demonstration split the textile operatives into confused, hostile camps. These customs appeared as ‘hideous’ and ‘incendiary’ conduct by a riotous ‘mob’, and the employers exploited the situation to demean the mule spinners and weavers as Lancashire brutes. These divisions and the crushing defeat of the weavers’ strategy convinced the mule spinners that the weavers’ union with its contingent of female rebels had led the strikers into disaster. Begging city authorities for bread backed with threats of violence was no manly way to deal with their employers.35 The struggle shattered the fledgling amalgamated union and its regional vision. The mule spinners signed away their union memberships, while the local weavers’ organization disintegrated. With the Lancashire past rendered

34 Fall River News, Aug. 17, Oct. 22, 1875; Massachusetts Bureau of Labor Statistic, Thirteenth Annual Report (1882), p. 341; New York Herald, Sept. 29, 1875. 35 Blewett, Constant Turmoil, pp. 213–222. 544

USA: shifting landscapes too passionate and perilous in 1875, the mule spinners recaptured the leadership of union organization in the New England print cloth industry. Other Fall River strikes in the late nineteenth century demonstrated the deep divisions between the mule spinners organized in trade unions and weavers seeking an amalgamation of all textile workers in the Northeast, including the print cloth mills of Cohoes, New York, and the silk mills of Paterson, New Jersey.36 Secretary Robert Howard of the Fall River Mule spinners’ Association saw no future for Lancashire-style protest in American trade unionism or for a coalition between spinners and weavers. He believed that trade unions with ample strike funds and region-wide locals offered a better chance to deal with mill agents. He won acceptance as a union leader who made his members ‘as obedient and docile and harmonious as the parts of a mule frame’.37 Howard counted on mill agents’ stubbornness to contrast with his strategy of patience, forbearance and moderation. One thousand spinners out of a total Fall River workforce of 14,000 adopted exclusive policies that attempted to restore their proper sense of respectable manhood: recognition of their union by the managers and higher wages based on calculations of costs and profits. But these efforts could not counter their employers’ power to refuse to negotiate, to hire strike-breakers or to set wages and working conditions. Even as the technology of ring spinning of all yarns undermined their craft in the 1890s, the mule spinners refused to organize young female and male ring spinners. Instead the mule spinners wanted a family wage paid to skilled men. The late 1880s and early 1890s brought great prosperity to the Fall River print cloth mills and high dividends to their shareholders. Wages lagged far behind, especially for the 7,000 weavers in 1888. With women and boys crowding the ring spinning rooms, the number of family men in weaving grew to half the workforce. Tensions developed among the weavers, reflecting the wishes of male weavers to imitate mule spinners’ trade unionism and demand for a family wage and exclude women, who were increasingly viewed as irrelevant. Abandoning their Lancashire traditions, the weavers’ organizations union voted in 1905 to join the mule spinners in the United Textile Workers (UTW), AFL. Lancashire immigrants dominated the first decades of conservative UTW leadership.38

36 Daniel Walkowitz, Worker City, Company Town (Urbana, 1978); Herbert G. Gutman, Work, Culture and Society in Industrializing America (New York, 1977) and Power and Culture: Essays on the American Working Class (New York, 1986); John Cumbler, WorkingClass Community in Industrial America: Work Leisure, and Struggle in Two Industrial Cities, 1880–1930 (Westport, 1979). 37 Boston Globe, Aug. 18, 1879. 38 Herbert Lahne, The Cotton Textile Worker (New York, 1944). 545

The Ashgate Companion to the History of Textile Workers

Countering Southern competition By 1910, the total number of spindles used by the American cotton industry ranked second only to the British, followed by Germany, Russia, France and India. Total exports, especially from developing mills in South Carolina, grew from US  $7.5 million in 1860 to US  $35.1 million in 1910, largely to Latin America.39 Northern mill agents successfully undermined regional labour protest. They divided the increasingly diverse nationalities and skills and exploited gender antagonisms through preferential treatment, unequal wages, denunciations of labour radicals or demonstrations of arbitrary power through wage cuts. Simple assertions by agents of secret, superior knowledge deprived the operatives of indisputable information on prices or wages. The announcement of unspecified wage cuts demonstrated their power to manipulate through fear and uncertainty, while dangling wage advances if the market improved. During strikes or lockouts, some mills unloaded huge, secret inventories for considerable profit. An informal network of agents and owners of cotton mills from Maine to Connecticut, organized into the Arkwright Club of Boston, worked together to crush regional labour protest by 1898, tolerating only unions of skilled men.40 But the rise of cotton textile manufacturing during the post-Civil War period in the Piedmont region of the Carolinas and in Georgia threatened the very foundations of New England cotton manufacture. By 1905 three Southern states together produced more in total value of cotton goods (29.27 per cent) than Massachusetts (28.68 per cent). In 1917 only seven cotton textile corporations out of hundreds of other industrial firms listed over US  $20 million in capital assets.41 The American South would dominate cotton textile production in the twentieth century as New England and the Northeast slid into decline, accelerating in the 1920s. Between 1923 and 1936, New England cotton operatives decreased from 209,000 to 74,226, and by 1976, the Southern mills possessed 18 million spindles compared with 280,000 in New England.42

39 Copeland, Cotton Manufacturing, pp. 275, 220–225. 40 Blewett, Constant Turmoil, pp. 310–313, 340–344. 41 Chandler, Visible Hand, p. 505. 42 Royal Little, How to Lose $100,000, 000 and Other Valuable Advice (Boston, 1979), pp. 90, 94. 546

USA: shifting landscapes Table 21.2 Southern competition and the New England cotton industry, 1900–1905 State

Rank

Percentage of US goods

Percentage increase

1905

1900

1905

1900–1905

Massachusetts

1

32.57

28.68

16.92

South Carolina

2

8.76

10.97

66.32

North Carolina

3

8.36

10.49

66.53

Georgia

4

5.44

7.81

90.57

Rhode Island

5

7.09

6.80

27.32

New Hampshire

6

6.78

6.56

28.45

Note: United States: total value of cotton goods: 1900, US  $339,200,320; 1905, US  $450,467,704. (Percentage increase in total value of goods: 32.80). Connecticut ranked eighth behind Pennsylvania and Maine tenth behind Georgia. Source: Massachusetts State Census, 1905 (Boston, 1905), vol. 3, pp. xix–xx.

To counter Southern competition, Northern cotton mills, such as the Boott Mills of Lowell, adopted several strategies: firstly, downward pressure on the wages of Northern operatives and the hiring of new, unskilled immigrant groups to use advanced technology, such as the Draper loom with automatic bobbin changing; secondly, the development of a more diverse product for new markets, such as towels, velvets, corduroys and plush for automobile upholstery; thirdly, the establishment of branch operations in the South to produce coarser goods at lower cost. New England textile operatives faced the familiar patterns of regional wage reductions and deteriorating working conditions. Scientific management although tempting to some firms was ‘rare’ in New England mills.43 By the turn of the twentieth century, the Boott Mills decided to invest no more capital in maintenance from its depreciation reserve – much less upgrading the process of production – but simply to run its mill operations ‘into the ground’.44 According to Boott mill worker Lancashire immigrant Harry Dickinson, ignorant foremen, recruited from ‘a junk pile’, did nothing to control heat, humidity and drafts.45 New England mill workers faced miserable conditions, worked harder and lived with the fear of increasing

43 Email, Laurence Gross, November 8, 2004. 44 Laurence F. Gross, The Course of Industrial Decline: The Boott Cotton Mills of Lowell, Massachusetts, 1835–1955 (Baltimore, 1993). 45 Ibid., pp. 236–238. 547

The Ashgate Companion to the History of Textile Workers workloads, plant closings, poor health and no future.46 A brief period of wartime prosperity in the 1940s encouraged quick unionization at the two remaining Lowell textile firms. Working conditions improved just before the collapse. Northern textile machinery companies briskly sold the latest technology to the Southern mills. Textron of Providence, Rhode Island, incorporated in 1943, demanded more productivity from New England operatives to meet the lower costs of its Southern mills and return higher profits. When resisted, Textron executives ruthlessly closed New England mills (1943–1955), abandoning communities and workers and taking huge tax losses to finance diversified acquisitions, thus creating the modern ‘conglomerate’.47

Upheavals in the woollen and silk industries In the late nineteenth century the New England woollen and worsted industries, protected by high tariffs against foreign competition, became increasingly dominated by powerful groups of industrial capitalists, sometimes organized into trusts: outright ownership of multiple mills in various locations.48 Table 21.3 Impact of rising tariff protection on US wool manufacturing, 1859–1919 Census year

Value of domestic production (US  $)

Percentage of imports to total

1859

65,596,000

33.8

1869

177,496,000

18.9

1879

194,157,000

13.7

1889

212,075,000

22.8

1899

229,572,000

7.9

1909

392,976,000

6.6

1919

985,296,000

1.9

Source: Adapted from Cole, American Woolen Manufacture, p. 39. 46 Mary Blewett, The Last Generation: Work and Life in the Textile Mills of Lowell, Massachusetts, 1910–1960 (Amherst, 1990). 47 Little, How to Lose, pp. 71–73, 84–100, 104–106, 113; New York Times, Obituary, 14 January, 1986. 48 Arthur H. Cole, American Woolen Manufacture (Cambridge, Mass., 1926), p. 225. 548

USA: shifting landscapes By 1917, the American Woollen Company and two additional corporations in Lawrence controlled a total of US  $189.8 million in capital assets in comparison with the US  $29.8 million of the American Thread Company.49 Between 1905 and 1911, the manufacturing dominance of the American Woollen Company in the New England worsted industry set regional wage rates. During those years, American Woollen owned 50 woollen and worsted mills in New England, employing 30,000, the largest industrial employer in the region. The New England workforce in the woollen and worsted industry represented experienced industrial workers from Belgian, France and Yorkshire, England, shut out of the American market by US tariffs in the 1890s.50 Other immigrant workers, especially from southern Italy, began to crowd into worsted mills in Massachusetts and Rhode Island, bringing with them the transnational politics of syndicalism, which contributed to the development of the Industrial Workers of the World (IWW), organized in 1905. Like French-Canadians, Italian workers became racialized, this time as ‘naturally’ violent, radical and emotional. The diverse workforce of 12,000 in Lawrence, swollen by 1910 with thousands of immigrants including women and children from southern and Eastern Europe and the Ottoman Empire, spoke 40 different languages and seemed impossible to organize into labour protest.51 Massachusetts had continued to shorten the legal working day and prohibited children under 14 from factory work. Factory managers accepted shorter hours without cutting piece-rates, but as the 54-hour law went into effect in January 1912, William Wood, president of the American Woollen Company, also cut piece-rates without notifying the operatives. Many textile workers, who earned low wages and faced bullying overseers and speeded-up work, discovered this rumour to be a fact as they opened their pay envelopes. UTW-AFL, representing about 2,800 skilled workers, accepted the situation, but the tiny, local IWW sprang into life over the surprise wage cut for recent immigrants. The ‘Bread and Roses’ strike in 1912 reflected great anger and deep alienation among the masses of textile operatives. Women workers, wives and daughters, from the Polish homelands and Italy walked off the job, sparking a spontaneous march of thousands of strikers through the Lawrence streets. It was they who demanded roses as well as bread. By mid-January, 25,000 textile workers from 40 nations were on strike, unprecedented in American industrial life. Cooperative, cross-ethnic networks of streets, tenement steps and kitchens, forged by women in ethnic neighbourhoods and based on the needs of new immigrants for housing, decent food and the ordinary daily rituals

49 Chandler, Visible Hand, p. 505. 50 Gary Gerstle, Working-Class Americanism: The Politics of Labor in a Textile City, 1914–1960 (Cambridge, 1989); Mary Blewett, The Yankee Yorkshireman: Migration Lived and Imaged (Urbana, 2009). 51 James Barrett and David Roediger, ‘Inbetween Peoples: Race, Nationality and the “New Immigrant” Working Class’, Journal of American Ethnic History (1997): 3–44; Ardis Cameron, Radicals of the Worst Sort: Laboring Women in Lawrence, Massachusetts, 1860–1912 (Urbana, 1993), pp. 74–83. 549

The Ashgate Companion to the History of Textile Workers of life, coordinated and sustained the strike.52 IWW strategies held together a multi-ethnic working-class revolt by creating locals to encourage the members of each ethnic group to speak for themselves, while multi-lingual IWW organizers provided overall direction to prevent violent acts and encourage the radical vision of One Big Union shared by many strikers.53 After the Lawrence settlement granted a 10 per cent wage increase, additional successful strikes broke out in the Lowell cotton mills and in the worsted mills of Rhode Island and Passaic, New Jersey. A crucial 1913 strike in the multi-ethnic silk weaving and dyeing industry of Paterson, New Jersey, met bitter hostility from the textile industry to the imminent threat of a successful IWW amalgamation sweeping the Northeast. Paterson’s silk mills, employing 38,000 with a high concentration of experienced, radical silk operatives from northern Italy, dominated the city, which represented New Jersey’s major industry. The strike of six months by weavers over a speed-up from two to four looms resulted in a costly defeat, virtually ending IWW activity in textiles.54 Federal government suppression of the organization as anti-American during US participation in the First World War insured its demise as an alternative to trade unions. Some woollen mills, such as J.P. Stevens, established Southern branches, but the American Woollen company continued to produce woollen and worsted especially for wartime demand until – a shell of its former magnificence – Textron acquired its remaining 11 small mills and sold them as tax losses in 1954.55

Paternalism and protest Between 1875 and 1905, cotton mills industrialized an agrarian South. At first Southern textile production, dependent on the rural white people of the Piedmont region of the Carolinas, made cheap cotton goods.56 By the turn of the twentieth century, Southern textiles posed both threats to the better grades of cloth made in the Northeast and the export trade.57 The rural backgrounds of poor, illiterate Southern operatives, the paternalist but authoritarian organization of mill centres which provided housing and services, the fear of black and foreign workers as competitors, and the long hours and low wages made labour protest difficult. Yet both the Knights of Labor which had been active in Massachusetts and Rhode Island and the early UTW-AFL textile unions successfully organized resistance in 52 Cameron, Radicals, pp. 117–169. 53 Melvyn Dubofsky, We Shall Be All: A History of the Industrial Workers of the World (New York, 1969). 54 David Goldberg, Tale of Three Cities: Labor Organization and Protest in Paterson, Passaic and Lawrence, 1916–1921 (New Brunswick, 1989), pp. 19–45. 55 Little, How to Lose, pp. 132–133, 136–146. 56 Melton Alonzo McLaurin, Paternalism and Protest: Southern Cotton Mill Workers and Organized Labor, 1875–1905 (Westport, 1971). 57 Patrick J. Hearden, Independence and Empire: The New South’s Cotton Mill Campaign, 1865–1901 (DeKalb, 1982), pp. 53–68, 125–144. 550

USA: shifting landscapes urban areas before 1905.58 Mill agents used time-honoured anti-labour methods: the blacklist, the lockout, evictions from mill housing and racial fears to undermine solidarity. Table 21.4 Growth of the Southern cotton industry, 1880–1900 State

Mills

Capital

Workers

North Carolina

1880 1890 1900

49 91 177

2,855,800 10,775,134 33,011,516

3,343 8,515 30, 273

South Carolina

1880 1890 1900

14 84 80

2,776,100 11,141,833 39,258,946

2,053 8,071 30,201

Georgia

1880 1890 1900

40 53 68

6,348,657 17,664,675 24,222,169

4,493 10,459 19,398

Source: McLaurin, Paternalism and Protest, p. 10.

The Southern mill worker was neither docile nor submissive to working conditions and low wages but accepted the cotton South’s class structure and rural values. Operatives in North Carolina mill villages had expectations of mutual rights, obligations and responsibilities, similar to Lancashire textile workers but based on yeoman or freeholder status in the antebellum slave South. When operatives resisted violations of reciprocity, protest became subtle but direct, based on their belief in a rural moral economy of individual independence, self-respect and family security. Women as well as men shared these views as well as their status of ‘whiteness’.59

Like a white family: cotton and rayon yarn mills Cotton mills in the Carolinas tapped a rural labour market with a cultural heritage of self-esteem and sociability. Mill hands were taught to spin and weave and tolerate factory discipline, cotton lint and accidents but also learned how to manoeuvre 58 McLaurin, Paternalism and Protest, pp. xiii–xviii; Jacquelyn Dowd Hall et al., Like a Family: The Making of a Southern Cotton Mill World (Chapel Hill, 1987), pp. 100–106. 59 Michael Shirley, ‘Yeoman Culture and Millworker Protest in Antebellum Salem, North Carolina’, Journal of Southern History 67 (1991): 427–452; Michelle Brattain, The Politics of Whiteness: Race, Workers, and Culture in the Modern South (Princeton, 2001). 551

The Ashgate Companion to the History of Textile Workers in a workplace that seldom allowed the expression of grievances. Between 1880 and 1930, Southern cotton workers earned about 60 per cent of wages paid in the North for a 12-hour day. The sexual division of labour changed only with the total employment of females on ring spinning. Mill hands needed cash wages but did not allow the dissolution of their social world. If personal negotiations with the boss failed, they took some days off, quit or occasionally organized a work stoppage: a pattern reminiscent of the Lowell factory women. In 1907 the annual turnover rate reached 176 per cent, while mill men estimated that 20–40 per cent of their mill hands ‘floated’ from village to village. The mill villages provided company housing and schools, company groceries with easy credit and ensuing debts, and even raised wages to tame the independence of the mill hands. But a separate mill hands’ world in the village revolved around family and kin, intermarriage and a common culture of sharing and sociability, dependent on female networks and church-going.60 Farm labour was always difficult and demanding, but as the seasons changed the variety of tasks contrasted with the boring sameness of mill routines. Southern men resisted mill discipline as long as possible. Still, cotton mill wages seemed steadier than agricultural commodity markets. Recruiting a workforce from rural families who expected their children to work long hours at an early age resulted in a high rate of child labour in the mills, often at the behest of the father. Between 1880 and 1910, one quarter of the mill hands were children under 12 years of age doing the simplest jobs assisting the ring spinners. When production broke down as steam engines stalled, children could take off to play, swim and fish. Blacks were denied production jobs. ‘Coloured’ men did the dirtiest, most dangerous work in the mill, such as dyeing and opening bales, while a few coloured women cleaned up the toilets and the floors, maintaining the social world of white dominance. Wartime profits and textile depression in the 1920s introduced ‘hard rules’ into the Piedmont mills. New machinery ran faster and worker productivity rose while wages were cut. The second generation of Southern mill hands refused to accept this stretch-out of work, and labour unrest broke out in the Piedmont on an unprecedented scale.61 Rayon yarn mills built by German firms in the 1920s for the US market processed local wood pulp. By 1940, the South, especially in Tennessee and Virginia, produced 70 per cent of US rayon. Rural towns offered free land, no taxes and cheap docile female operatives. But in 1929 white girls as young as 14 from small farm families went on strike in East Tennessee over the low wages and poor working conditions: for beginners a 56-hour week and weekly wages of US  $8.16. Depending on networks of friends and acquaintances, the strikers, called ‘disorderly women’, held out for better treatment, while openly taunting and tantalizing the National Guardsmen sent to control them. The UTW-AFL officials proved uninterested in the strike, and

60 Hall et al., Like a Family, pp. 107, 129–180. 61 Ibid., pp. 44–45, 53, 66–67, 77, 80, 183. 552

USA: shifting landscapes many operatives drifted away, returning as wages rose during the Second World War.62 Still, gender and a common rural culture produced class unity in 1929.

No depression at Burlington Mills The Burlington Mills of North Carolina, begun in 1923 by J. Spencer Love, built the biggest rayon operations in the South by integrating production processes, diversifying product, linking production and sales, and gaining tariff protection. Replacing silk weaving with rayon in 1929, the company responded to consumer demand for inexpensive slinky dress goods and stockings. As the cotton industry collapsed in the North and struggled in the South, rayon weaving and finishing operations expanded. The workforce at Burlington leaped from 200 in 1924 to 6,900 in 1933 and 11,000 in 1939, while the number of factories rose from one to 19 by 1937. Leaving the costly process of producing rayon yarn for weaving to the chemists at Du Pont, Celanese and American Viscose, Burlington managers financed expansion from local or Southern investors and credits extended by machinery companies. While competitors in New Jersey, Massachusetts and Pennsylvania shifted their silk mills to rayon goods, Burlington held the edge in modern machinery. Ninety per cent of North Carolina mills, followed closely by South Carolina, had high-speed automatic looms, compared with only 9 per cent in New Jersey and 20 per cent in Pennsylvania. Low wages and little labour unrest solidified the advantage, especially as compared with Paterson. Productivity soared; profits more than tripled those of Northern mills. When the rayon craze lagged in the 1940s, Burlington Mills wove nylon, then acrylics and polyesters relying on synthetic laboratories. Production diversified into brand-name hosiery and carpets, polyester-woollen men’s suiting and even returned to cotton production in response to market demand. Textron and J.P. Stevens also diversified, but Burlington managers proved more flexible, making elastics, fake fur and plastic-coated fabrics. In 1962 when Love died, Burlington Mills moved into denim.63 Silk, rayon and most woollen production had ended in the Northeast.

The 1934 General Textile Strike The industrial depression of the 1930s hit textiles very hard, North and South. The UTW-AFL faced oblivion, most mills overproduced at lower costs and operatives endured an aggravated stretch-out of their workload at lower wages. The violent 62 Jacquelyn Dowd Hall, ‘Disorderly Women: Gender and Labor Militancy in the Appalachian South’, Journal of American History 73 (1986): 354–382. 63 Annette C. Wright, ‘Strategy and Structure in the Textile Industry: Spencer Love and Burlington Mills, 1923–1962’, Business History Review 69 (1995): 42–80. 553

The Ashgate Companion to the History of Textile Workers Gastonia, North Carolina, strike of 1929 was lost, but the national strike of 1934 united hundreds of thousands of cotton, silk, woollen and rayon workers from Maine to Alabama against worsening conditions under the early New Deal programmes of President Franklin D. Roosevelt.64 The strike began in the cotton mills of Huntsville, Alabama, pushed by a hard-driving UTW leader from New York. By September, nearly a million operatives had joined together to demand change. ‘Flying squadrons’, fleets of cars and pickup trucks driven by activists, flew from one mill centre to the next across the Piedmont, in New England and Pennsylvania, calling out the operatives to picket duty. Strikers in Fall River and New Bedford, Massachusetts, joined others in Rhode Island, New Hampshire and Maine or wherever textile mills continued operations. The cotton mills of the Carolina Piedmont became the focus of activity. Violence grew as deputy sheriffs and National Guardsmen shot and killed squadron members and pickets. Still, the strike spread, impelled by desperation. A woman striker in Greenville, South Carolina, wrote the President that every textile worker in the South would join the strike if ‘they were not afraid of starvation’ or retribution. Young women strikers became especially active on picket lines and in union halls.65 The key issue was the exhausting stretch-out. Violently repressive state governments in Rhode Island, Georgia and the Piedmont disrupted picketing and intimidated strikers, while recalcitrant mill owners gladly sold off their inventories. The isolation of the many scattered cotton mills, the immediacy of extreme hunger and the uneven response to the strike by skilled textile workers in New England and Philadelphia, whose wages rose under New Deal programmes, undermined the one month general strike. By 1939 as war demand rode for textiles, the Textile Workers Union of America (TWUA), an industrial union in the Congress of Industrial Organizations (CIO), replaced the UTW. Still, the involvement of hundreds of workers at the Burlington Mills in the 1934 strike, including incidents of dynamiting company property, forced the mills to reinvent paternalism as personnel management to stave off post-war unionization.66

Operation Dixie: the post-war economy, race relations and amalgamation War in 1939 created demand for US cotton and woollen textiles. Big orders, higher profits and worker unrest over lagging wages and workload speed-ups prompted 64 John A. Salmond, Gastonia, 1929: The Story of the Loray Mill Strike (Chapel Hill, 1995); John A. Salmond, The General Textile Strike of 1934: From Maine to Alabama (Columbia, 2002). 65 Salmond, 1934 Strike, pp. 55, 241. 66 Annette C. Wright, ‘The Aftermath of the General Textile Strike: Managers and the Workplace at Burlington Mills’, Journal of Southern History 50 (1994): 81–112. 554

USA: shifting landscapes the TWUA-CIO to organize what was left of the industry in the Northeast. At the Boott Mills in Lowell, bolstered by Navy contracts for cotton duck, managers accepted TWUA unionization in 1942.67 Still in 1955, the Boott agents closed their doors. The TWUA’s future lay in the South with 80 per cent of the nation’s mills. The campaign to organize Southern mills, dubbed ‘Operation Dixie’, began in 1946. But the Second World War had greatly altered the South with huge federal military installations, steady post-war Defense Department contracts and rising standards of living. Wages for both union and non-union workers increased steadily. As consumers, Southern wage-earners bought automobiles and consumer durables on instalment. All of these changes made strikes much more difficult, adding to employer and state government hostility and racial tensions.68 The TWUA general strike in 1951, based on established locals, sought to make textile wages the envy of all Southern workers and unmatchable by other employers. However, the strike did not attract as much support as in 1934. By 1955, Operation Dixie had failed, but textile operatives earning good wages still faced exhausting speed-ups and higher workloads. Union grievance procedures, whatever the outcome, diminished their fears of intimidation by foremen.69 Unions supported segregated workplaces and separate unions for blacks, who began to enter production jobs in 1964 after the US Civil Rights Act challenged occupational segregation. By 1967 8 per cent of cotton workers in the Piedmont area were blacks, who left low-paid farm and domestic work and responded well to factory discipline. Meanwhile white workers and their unions resisted integration or left to seek work in the South’s booming cities. By 1970 North Carolina represented the premier state, producing near one-third of American textiles and employing 280,000 workers.70 But after 1980, many textile mills began to close, giving up 500,000 jobs, a decline accelerated by the North American Free Trade Agreement and the elimination of tariff protection against Mexican and Canadian textiles. The trend in production jobs by 1995 seemed to be all-black production workers and white supervisors.71 As textile mills and garment factories disappeared, the textile workers’ unions amalgamated first with the garment workers in the Amalgamated Clothing and Textile Workers’ Union and most recently with the hotel workers’ union. Hard-hit mill villages in the Piedmont area returned to their rural culture of family and kin, subsistence agriculture and

67 Gross, Industrial Decline, pp. 195–211. 68 Timothy Minchin, What Do We Need a Union For? The TWUA in the South, 1945–1955 (Chapel Hill, 1997). 69 Daniel Clark, Like Night and Day: Unionization in a Southern Mill Town (Chapel Hill, 1997). 70 John Gaventa and Barbara Ellen Smith, ‘The Deindustrialization of the Textile South: A Case Study’, in Jeffrey Leiter et al. (eds), Hanging By a Thread: Social Change in Southern Textiles (Ithaca, 1991), pp. 181–196, at p. 182. 71 Timothy Minchin, Hiring the Black Worker: The Racial Integration of the Southern Textile Industry, 1960–1980 (Chapel Hill, 1999), pp. 265–271. 555

The Ashgate Companion to the History of Textile Workers odd jobs to survive the new era of layoffs and plant-closings but faced sharply declining incomes and heavy medical expenses.72

Reconstructing during globalization Textile capitalists restructured to meet late twentieth-century foreign competition worsened by a strong dollar. In 2001, Chinese fabrics alone represented just over a quarter of all American textile imports and 12 per cent of apparel imports. Together Pakistan, Canada and Mexico totalled 31 per cent of textile imports, while Korea and India represented 11 per cent. The combined Asian imports at 47 per cent of the total or 12.4 per cent of US textile output represented a racialized ‘menace’, according to textile manufacturers.73 But American consumers bought cheap, foreign-made garments, merchandized in large discount stores or through mail order and Internet catalogues, at a far higher rate: 50.5 per cent import penetration ratio to US production. As the US economy shifted away from industrial production towards financial, technology and service jobs, many women workers and consumers lost their interest in or ability to turn fabric into clothing. Southern mills adapted to globalization by reshaping the ‘race to the bottom’ of cutting wages or moving offshore. Economic sociologists use the concept of embeddedness in local communities founded on racially structured labour markets to demonstrate this. Merged corporations, such as Pillowtex-FieldcrestCannon in 1997, constructed new ‘secure’ (close political ties to federal, state and local government) locations of production in impoverished rural areas away from established mill centres. One example is the move of the Allied Signal Seat Belt Company from Knoxville, Tennessee, to rural Greenwood, Alabama.74 Using the latest automation and computerized technologies, these firms reduced the number of employees and hired non-union black females or Asian and Hispanic immigrants. Like nineteenth-century Philadelphia, Southern mills concentrated on flexible batch production for specialized customers, such as designers Calvin Klein and Tommy Hilfiger.75 The numbers of textile workers in Southern mills continued to decline as textile firms tried to remain competitive in the 1990s. Multinational corporations such as Burlington Industries produced denim in Mexico as Burlmex. Shipping American fabric, already finished and cut, to Central American nations for assembly and reimportation with special tariff protection provided by the Caribbean Basin Initiative (1986) helped textile mills avoid the fate of US garment factories. Still foreign competitors forced Burlington Mills to experiment in costly new technologies and 72 Gaventa and Smith, ‘Deindustrialization’, pp. 192–193. 73 American Textile Manufacturers’ Institute, 8 March, 2004 at www.atmi.org 74 Gaventa and Smith, ‘Deindustrialization’, pp. 188–190. 75 Cynthia Anderson, Michael Schulman and Phillip Wood, ‘Globalization and Uncertainty: The Restructuring of Southern Textiles’, Social Problems 48 (2001): 478–498 at p. 481. 556

USA: shifting landscapes downsize, selling its carpet operations in 2003 to emerge from bankruptcy under a new name. The merger in March 2004 of the Cone Mills Corporation and Burlington Industries, both of North Carolina, combined their denim manufacturing companies in Mexico with jacquard weaving in the USA and other operations in India. This new International Textile Group, however, faced large debts and the pressing need to modernize machinery.76 In January 2005, all protection of American textiles (and clothing) production in the form of import quotas under a 1995 World Trade Organization agreement ceased. An estimated 600,000 additional jobs would be lost to the American economy, while Chinese textiles especially benefited, confirming fears of its complete global domination.77

76 New York Times, March 18, 2004. 77 New York Times, September 2, 2004. 557

This page has been left blank intentionally

PaRt II

INtERNatIONaL COMPaRISONS

This page has been left blank intentionally

22 Global trade and textile workers Prasannan Parthasarathi

Because of their high value and low weight, textiles have long been a major commodity in local, regional and, from the sixteenth century, global trade. Indeed, fragments of Indian cotton cloth have been found in Roman archaeological sites in the eastern Mediterranean. This trade, of course, would have taken place some two thousand years ago. The same axis of exchange, from the Indian Ocean to the Mediterranean Sea, was active in medieval times, which has been documented recently by Ruth Barnes’ detailed analyses of cotton cloth fragments from the Cairo bazaar, now deposited in Oxford’s Ashmolean Museum. It should come as no surprise, then, that global flows of textiles shaped the lives of textile workers in the period from 1650 to the present. As these two examples of exports from the Indian subcontinent suggest, cotton has long been the pre-eminent textile fibre when it comes to long-distance trade. It holds this position because it is a cloth that can be comfortably worn in a wide variety of climates: it suffices as the primary garment in the heat and humidity of the tropics and serves as underclothing in the colder temperatures of temperate zones. Cotton also takes and holds dyes well, which allows for the manufacture of colourful and elaborately patterned cloths that are easily washed. Finally, the prices of cotton cloths have historically been very low in comparison with its competitors. Due to these qualities, since at least the nineteenth century cotton has been the most important textile fibre in the world. By 1913 cotton accounted for 80 per cent of global fibre consumption: cotton consumption for textile uses was about 4 billion kilograms, while that of wool, the second most important fibre, was about 800 million kilograms. In 1990, even after the rise of synthetic fibres, which accounted for 38 per cent of global fibre use in textiles, cotton accounted for 48 per cent of world fibre share and wool had slumped to the low share of 4.9 per cent. Therefore, this chapter on global trade will focus on cotton, but not neglect  Ruth Barnes, Indian Block-Printed Textiles in Egypt: The Newberry Collection in the Ashmolean Museum (Oxford, 1997).  Donald Coleman, ‘Man-Made Fibres before 1945’, in David Jenkins (ed.), The Cambridge History of Western Textiles (Cambridge, 2003), vol. 2, pp. 933–947, at p. 939.

The Ashgate Companion to the History of Textile Workers other fibres, most critically wool, silk and eventually synthetic materials such as polyester and rayon. In this chapter, the long period of roughly 350 years, from 1650 to 2000, will be divided into three periods. The first runs from 1650 to 1780. In this period cotton manufacturing was largely located in the vicinity (within a few hundred miles) of areas where cotton was grown and in these decades the major exporting region of the world was the Indian subcontinent. The second period runs from the 1770s to the eve of the First World War. From the eighteenth century, new centres of cotton manufacturing began to develop, initially in Europe, but eventually spreading to the Americas, Australia and East Asia. Some of these were located at long distances from cotton cultivation regions and many of them had no previous history of cotton manufacturing. For the next roughly century and a quarter, these new centres of cotton began to grow rapidly, initially concentrated in the British Isles, particularly in the area around Manchester, but elsewhere as well. These new centres established themselves as world leaders in cotton manufacturing and in the global textile trade. Old centres of cotton manufacturing, in India and elsewhere, did not disappear, but diminished in importance, in both absolute and relative terms. In the final period, from the First World War to 2000, many of the nineteenth-century centres of cotton manufacturing entered protracted decline and in the last few decades there has been a new regional pattern of global textile manufacturing, with the revival of Asian production, and, therefore, of global textile trade.

1650–1780 Almost all of the world’s cotton manufacturing in the mid-seventeenth century was found in an arc of relatively warm weather climate regions, starting in the Americas and encompassing both Mesoamerica and the Andean region, through Africa, including both north and sub-Saharan as well as the areas of the eastern Mediterranean, into Asia, in particular South and South East Asia, and major regions of China and Japan. Running parallel to this arc of manufacturing was an arc of cotton growing. Cotton can be grown in a variety of soils, from the heavy and clayey to the light and sandy. It requires, however, a long growing season for the pods to achieve full maturity, which places severe restrictions on where it may be grown successfully. In contrast to these regions, the more temperate areas of North America, Europe and northern Asia drew upon wool, linen, hemp and silk as the raw material for their cloth production. In these areas as well, these raw materials were drawn from networks of trade that stretched at most for only a few hundred miles. In Western Europe, for instance, there was a sizable trade in raw wool from across the continent. The one exception to the generally local level of trade in the raw material for cloth production was the case of silk. Raw silk, because of its very high value, had for a long time been transported long distances. In the middle of the

562

Global trade and textile workers seventeenth century there was a sizable trade in raw silk from Persia and China to Europe. Although cotton manufacturing was diffused quite widely, the Indian subcontinent was the major source of cotton cloth exports in the mid-seventeenth century, and this was to remain the case until the late eighteenth century. As already indicated, India had long exported cotton textiles to the eastern Mediterranean. In the mid-seventeenth century this trade was booming, partly because of the explosion in supplies of monetary media to the global economy after the discovery of silver in the Americas. Cairo, as well as other major trading centres in the territories of the Ottoman Empire, became a significant destination for Indian cotton goods. There was also a booming trade to the ports of the Persian Gulf from where cloth was taken to markets in Ottoman territories as well as to Persia. Although this trade has not been studied in the detail that it richly deserves, several of its features are known. Firstly, the trade was most definitely not purely in luxury textiles, but was very much a mass trade. Rudiger Klein has argued this in his University of London Ph.D. dissertation from textual sources. His conclusions have received material confirmation from the work of Ruth Barnes, who analysed very carefully several hundred textile fragments found in the Cairo bazaar. She has dated these fragments to the period between the eleventh and fifteenth centuries and the vast bulk of them were of coarse weave and decorated with crude blockprinted patterns. Both of these findings indicate that most of her samples were cheap cloth for lower end, not luxury, consumption. Much of the cotton cloth that was exported to this area came from western India, in particular from Gujarat. Jean-Baptiste Tavernier, while travelling through the region in the mid-seventeenth century, remarked upon large numbers of weaving villages that were devoted almost exclusively to production for the Persian Gulf market. Two other long-standing destinations for Indian cotton exports were South East Asia and East Africa. In the former, Indian cotton cloth was exchanged for spices, which were consumed in the subcontinent itself as well as shipped from there to west Asia and the eastern Mediterranean. In the latter, cloth was an essential item in the slave trade of the Indian Ocean as textiles moved from east to west and cargoes of humans moved in the opposite direction. The demand for Indian cottons in Africa made possible the slave armies that Islamic kingdoms in the subcontinent fielded in medieval and early-modern times. From the sixteenth century, with the entry of European traders into the Indian Ocean, three new markets were opened for Indian cloth. The first was in Europe itself. In the sixteenth century the Portuguese began to import small quantities of Indian painted cloth. These tended to be of very high quality, but diffused in a small scale to other parts of Europe, most critically, for future developments, to the economically vibrant nations of England and Holland. From these early luxury beginnings, in the seventeenth century Dutch and English traders, who had been selling Indian cloth in South East Asia to finance their purchases of spices, experimented with shipping cargoes of cloth back to Amsterdam and London. These goods quickly found mass markets in Europe, setting off what contemporaries labelled as the ‘calico craze’. 563

The Ashgate Companion to the History of Textile Workers Indian cotton cloths found ready buyers in Europe because they were inexpensive, light, colourful and colourfast, and easily washed. These textiles were found in a vast range of styles and qualities. Fine cloth that was elaborately painted with floral and other ‘nature’ patterns found buyers among the upper classes in Europe. At the same time, imports of Indian cottons made it possible for the poorer classes to have access to similar cloth, albeit of a coarser variety and with block printed, rather than hand-painted, floral patterns, but an exciting contrast to the drab browns and greys of woollen cloth. And to the casual observer, both rich and poor were dressed alike, producing a kind of social levelling through consumption. For the poorer classes of Europe, Indian goods also made it possible for them to wear a lighter cloth that was less expensive than linen or silk and far more comfortable than even the lightest woollens. By the end of the seventeenth century, the appetites of many Europeans for cotton cloth had been whetted. The second major new market for Indian cloth that arose in the seventeenth century was in West Africa. As already stated, East Africa had long been buying Indian goods and it is likely that some of these cloths made their way to the Atlantic coast of Africa via transcontinental trading networks. European slave traders, however, established the direct sale to the Atlantic coast of Africa and Indian textiles became a linchpin in the Atlantic slave trade. Experts on that trade have estimated that in some decades of the eighteenth century Indian cloth accounted for about 35 per cent of the goods that were exchanged for slaves. It must be recalled that the eighteenth century was also the peak of the Atlantic slave trade, with half of the slaves that were carried to the Americas between 1500 and 1900 taken in that century. The final new market for Indian cloth was in the Americas, for which we have the least information. Indian cottons were exported to the Americas via two channels. The first was an export trade from north-western Europe to the islands of the Caribbean and North and South America. The plantation complex was a major market, especially for the clothing of house slaves for whom the relative lack of durability of cotton was less of an issue. The second route went east from the subcontinent to Spanish America through Manila. About this trade almost nothing is known, but there is certainly scattered evidence for the sizable impact that Indian cloth had in the Spanish colonies of the Americas. To put the foregoing material together, by the eighteenth century Indian textiles found buyers in South East Asia, East and West Africa, West Asia and the eastern Mediterranean, Europe and across the Americas. In the seventeenth and eighteenth centuries it was the only textile that can be said to have been integral to the global trading system. Indian cloth was central in the slave trade. The demand for Indian cottons was also a prime reason for the flow of silver from America, Europe and West Asia to the Indian subcontinent: Europeans and others had no choice but to bring silver and gold, as well as other monetary media, to purchase Indian cloth because there was limited demand for European goods in the Indian subcontinent. Since the seventeenth century it has been assumed that the low prices for Indian cloth were due to low wages for the manufacturers. In contrast to the sturdy and well-paid artisans of Europe, Indian weavers were portrayed as impoverished and 564

Global trade and textile workers oppressed. In recent years this picture has been radically revised as evidence has been provided for similar standards of living for weavers and spinners in Britain and regions of India in the mid-eighteenth century. The competitiveness of Indian cloth, it is now believed, had its origins in the lower prices for grain, which were half or quarter of prices in Britain. Therefore, comparable standards of living could be achieved with far lower money wages in the Indian subcontinent, which translated into lower prices for cloth. In most markets around the world, Indian cloth was imported freely. The one exception was in Europe, where from the late seventeenth century restrictions began to be placed on the sale of Indian cloth. These ranged from high import duties on the cloth to the outright ban on their importation. In England, the actions of weavers, especially those in silk, were instrumental in the erections of barriers to Indian goods. In London there were reports of the silk workers of Spitalfields attacking women who wore Indian calicoes and muslins, and even reports of gowns made of Indian material being ripped off the wearer. These systems of protection were to assist European manufacturers in their attempts to imitate Indian cotton cloth, which is explored in more detail in the next section. In the eastern Mediterranean, where there were no barriers to the importation of Indian cloth, there may have been a decline in textile manufacturing in some areas in the eighteenth century. This is suggested by price information for Cairo, where the price of raw cotton rose enormously in the eighteenth century, most likely driven up by the tremendous expansion in exports of raw cotton from Egypt to Marseilles. The price of cloth, however, did not change in the same period. The stability of cloth prices suggests that Indian imports were setting those prices, making it difficult for local producers to pass on higher cotton costs. The outcome would have been the squeezing of Egyptian textile producers and most likely a decline in manufacturing capacity as Indian imports sold for less. Indian imports also most likely made up for the decline in local cloth production due to growing exports of raw cotton. This competition with India also helps to explain the declining value of silk weavers’ estates in Cairo between the late seventeenth century and the late eighteenth century. In the seventeenth and eighteenth centuries, no textiles except those from cotton were traded on a large scale globally. Of course, wool cloth from Europe entered into circuits of trade in the Atlantic and was sold in West Africa and North and South America. Some wool cloth also made its way to markets in the eastern Mediterranean  Prasannan Parthasarathi, ‘Rethinking Wages and Competitiveness in the Eighteenth Century: Britain and South India’, Past and Present 158 (1998): 79–109 and ‘Agriculture, Labour, and the Standard of Living in Eighteenth-Century India’, in Robert C. Allen, Tommy Bengtsson and Martin Dribe (eds), Living Standards in the Past: New Perspectives on Well-Being in Asia and Europe (Oxford, 2005), pp. 99–103.  For further details, see Prasannan Parthasarathi, ‘Indian Cloth Exports to the Eastern Mediterranean in the Eighteenth Century’, paper presented at the International Economic History Association 13th Congress, Buenos Aires, July 2002.  Beinin, Chapter 7, this volume, ‘Egypt’. 565

The Ashgate Companion to the History of Textile Workers and there was a trade in woollens, especially fine cloth made from mohair, in the reverse direction. European woollens, however, found a limited market in Asia, and this was not due to any lack of effort on the part of European traders. The English East India Company, for instance, was under constant pressure to vend English broadcloth in the Indian subcontinent, but these textiles never found many buyers. There was also a trade in silk cloth from China in many markets, including South and West Asia and Europe, but this remained a small, luxury trade.

1780–1913 Before the seventeenth century, cotton cloth was manufactured in a few areas of Europe, most notably in the Ottoman provinces of south-eastern Europe, in the Italian peninsula, in Switzerland and in southern Germany. With the exception of the Ottoman territories, these were fairly small-scale manufactures and were dwarfed by the cotton manufacturing centres of Asia and Africa. From the seventeenth century, the craze for Indian calicoes led growing numbers of Europeans to experiment with the manufacture of cotton goods. The most famous of these centres was, of course, Manchester and its surrounding towns, which became the cockpit of the world’s cotton industry in the nineteenth century. In the eighteenth century, however, as the national and regional overviews in this volume indicate, cotton spinning and weaving came to be found quite widely in Europe. In Barcelona, printing on imported cotton cloth was established between 1728 and 1742 and in the 1760s spinning of yarn and the weaving of cotton cloth commenced. In the Habsburg territories, cotton manufacturing was concentrated in Bohemia, where in 1770 the entrepreneur Johann Josef Leitenberger founded a central workshop with 100 workers relying upon the labours of 2,000 dispersed spinners. In the Netherlands cotton cloth production began in the eighteenth century in North Brabant and Twente. Other regions where cotton began to be used in textile manufacturing include Central Europe, Russia and France, where a sizable cotton manufacturing industry arose in the eighteenth century utilizing cotton imported from the eastern Mediterranean. The push to imitate Indian stuff even reached Brazil, where the Portuguese debated bringing weavers and printers from South India to instruct local manufacturers in how to manufacture good cotton cloth. Cotton cloth production in these centres was limited due to a combination of technical limitations and problems of price. The most important technical limitation  Smith et al., Chapter 17, this volume, ‘Spain’. Also see, J.K.J. Thomson, A Distinctive Industrialization: Cotton in Barcelona, 1728–1832 (Cambridge, 1992).  Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’.  Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’.  Delson, Chapter 4, this volume, ‘Brazil’. 566

Global trade and textile workers was the inability of Europeans to spin cotton yarn that was strong enough to serve as the warp. Because of this, in many of these new cotton centres, cotton wefts were combined with linen warps to weave a variety of mixed fabrics that were often referred to as fustians. This was not an accurate name for these cloths as they were much lighter than the traditional fustians, which were heavier and thicker fabrics. From the early eighteenth century these mixed textiles were made in imitation of India all-cotton cloth and marketed in West Africa under Indian names. Although in appearance they may have matched Indian goods, the texture of these mixed cloths could not compare to the softness and suppleness of all-cotton cloth. Nor could they take dyes as well. The second problem that these early cotton manufacturers faced had to do with the prices of their cloth, which were far higher than all-cotton goods from the Indian subcontinent. This severely restricted demand for them outside protected home markets. These problems of price and quality were solved with the invention of three machines for the spinning of cotton in the 1770s. The first, the spinning jenny, dramatically reduced the price of cotton yarn. The second, the water frame, made it possible to spin yarns that were strong enough for the warp. And the final, the mule, enabled Europeans to spin very high quality yarns for the manufacture of muslins and other fine textiles. Yarn spun on these new machines was far cheaper than handspun and effectively nullified the price advantage that Indian manufacturers had long held due to cheaper grain prices. The decline in prices boggles the mind, with the price of yarn of 40 hanks to the pound (spun on the water frame) plummeting from 16s. in 1779 to 2s. 6d. in 1812. For yarn of 100 hanks to the pound (spun on mules) prices went from 38s. in 1786 to 5s. 2d. in 1812.10 With these inventions, the centre of gravity in global textile manufacturing and export moved from the Indian subcontinent to north-western Europe. And for the first time, cotton manufacturing began to be undertaken on a massive scale in regions where the fibre was not grown. Both of these developments had important consequences for global trade. With the application of machinery to cotton spinning in Britain, cotton manufacturing grew rapidly. Cotton imports to the British Isles reached 400 million pounds by 1840 as compared to a figure of 5 to 6 million pounds in 1780. The quantity of yarn and cloth that was spun and woven far exceeded domestic demand and British manufacturers from the earliest days of machine spinning turned to overseas markets for the sale of their products. By 1840 more than half of British cotton production was exported and by the 1870s this figure was around two-thirds, rising to nearly three-quarters in 1913.11 The earliest markets were in Europe itself and in 1820 these nearby markets took nearly two-thirds of British cotton goods exports. By 1850, this figure had fallen to about one-third and by 1896, to less than 20 per cent. The declining importance of Europe was made up by expansion in Latin America, the eastern Mediterranean 10 S.D. Chapman, The Cotton Industry in the Industrial Revolution (London, 1972), p. 44. 11 D.A. Farnie, The English Cotton Industry and the World Market 1815–1896 (Oxford, 1979), p. 7. 567

The Ashgate Companion to the History of Textile Workers and Asia. By 1850 Latin America, chiefly Brazil, absorbed nearly a fifth of British cotton textile exports and the eastern Mediterranean almost a tenth. In the latter, as Halil İnalcık has shown, much of this cloth was marketed with Indian names, which is a sign of the enormous popularity of Indian textiles in that market in previous centuries.12 From the 1820s Asian markets began to also be penetrated by British yarn and cloth, but this trade to the east reached sizable proportions only after mid-century; by 1896 Asian markets took more than 40 per cent of British cotton exports. The expansion of this trade benefited from the opening of the Suez Canal and the diffusion of steam ships, which lowered the costs of transport to Asia. From the late eighteenth century, the response to British cotton manufactures took one of two forms. The first was to erect barriers to the imports of British goods, which gave protection to local textile manufacturing. By 1820, British yarn and cloth faced tariffs in much of Europe and in the United States. From the middle of the nineteenth century, British exports paid import duties in major markets in Latin America, including Mexico and Brazil. These protected areas of the world tended to be the quickest to import the new British technologies and with the assistance of British technical experts to develop indigenous cotton industries. In most places, as was the case in Britain, these manufacturers relied upon imported supplies of cotton. These emulations of the British cotton industry were most successful in the United States, located in the New England towns of Fall River, Lowell and Manchester, France, the Netherlands, in Central Europe, both in the Habsburg Empire and in the regions that were to comprise Germany, and Spain. Of these, the United States was to reach the greatest heights, possessing by the early twentieth century the second largest textile industry in the world. Cotton manufacturing was first undertaken in New England in the late eighteenth century, but after the Civil War these old centres came under competition from the South where labour was both cheaper and more quiescent. To maintain their competitiveness, the mills of New England modernized technology and put pressure on workers to cut wages and intensify work, but this was a losing battle by the twentieth century. 13 Cotton manufacturing in the USA also took advantage of export markets, most critically in Latin America, where before 1840 US textile manufacturers were able to find buyers during times of slack domestic demand. From the mid-nineteenth century, after the opening of its treaty ports, China was to be an increasingly important destination for US exports.14 From the 1870s, Japan also sought to develop a domestic cotton industry to replace imports of yarn and cloth, which were draining Japan’s foreign exchange reserves. State support contributed to the rise of this industry, but it did not take the 12 Halil İnalcık, ‘When and How British Cotton Goods Invaded the Levant Markets’, in Huri İslamoğlu-İnan (ed.), The Ottoman Empire and the World Economy (Cambridge, 1987), pp. 374–455. 13 Blewett, Chapter 21, this volume, ‘USA’. 14 Caroline F. Ware, The Early New England Cotton Manufacture: A Study in Industrial Beginnings (Boston, 1931, repr. New York, 1966) pp. 189–195. 568

Global trade and textile workers form of protection from import competition as in many other parts of the world. The Japanese state provided capital at below-market rates and subsidies to the industry. From the 1890s, the state also aided the export push of Japanese spinners in a variety of ways, including subsidies and cheap credit. From the 1920s, however, the Japanese textile industry was to receive protection.15 The profusion in cotton manufacturing gave rise to an international trade in the raw material and the global trade in cotton became a major part of the nineteenthcentury world economy. Cotton cultivation for export became a significant agrarian activity on four continents, North and South America, Africa and Asia, and different centres of manufacturing in the world came to depend upon different sources of cotton. The most important event in the nineteenth-century cotton trade was the American Civil War, which cut off supplies of raw material to many of the world’s cotton industries. The scramble to find new sources of cotton had a global impact as European cotton traders fanned across the Indian subcontinent, Africa, central Asia, Brazil and elsewhere to encourage the cultivation of cotton and to purchase it. Although in many places the switch to cotton growing was to be temporary, in others cotton became an essential crop for agriculturalists. The second response to British yarn and cloth imports was to give them free access without any restrictions on their importation. In these countries or areas, the adoption of new technology was far more fitful in the nineteenth century. These places experienced significant deindustrialization since the creation of a ‘modern’ textile industry emerged with some delay, during which time employment in textile manufacturing fell. Such a pattern has been identified in the nineteenth-century Ottoman Empire. Imports of cheap British yarn, which increased fiftyfold between 1810 and 1910, put a severe dent on value added and employment in local manufacturing, and, in 1900, hand-spun accounted for only one-fourth of cotton yarn consumed in the empire. Donald Quataert has argued, however, that employment did not plummet in textiles because ‘the import of yarns thus meant vast savings for the Ottoman textile industry and an enhancement of its ability to compete’. Yet he provides no evidence in support of this assertion and the decline in spinning must have meant a sizable loss of employment. A notable feature of the Ottoman textile industry was the limited adoption of machinery for spinning and weaving in the nineteenth century.16 In Mexico, according to Jeffrey Bortz, because of cheap imports from Britain, textile workers went from comprising half of Mexico City’s population in 1788 to one-third in 1842. It was only from the 1830s that British machinery was imported into the country and several mills spinning cotton yarn were founded. By 1898, the number of factories had grown to 118 with nearly half a million spindles and 22,000 looms. In the second half of the nineteenth century, the emerging Mexican 15 E. Sydney Crawcour, ‘Industrialization and Technological Change, 1885–1920’, in Kozo Yamamura, The Economic Emergence of Modern Japan (Cambridge, 1997), pp. 50–115, at pp. 89–91. 16 Quataert, Chapter 18, this volume, ‘Ottoman Empire’. 569

The Ashgate Companion to the History of Textile Workers cotton industry benefited from tariffs on imported yarn and cloth. Nevertheless, the import of cloth was not fully cut off. While local factories produced lower quality cloths, high quality goods for the wealthy were imported.17 Similarly in China, from the mid-nineteenth century, imported yarn and cloth found buyers in much of the country. British manufacturers dominated in the more expensive and lighter categories, while American and Japanese prevailed in the coarser and heavier. Imports of machine-spun yarn made major inroads in Chinese markets, increasing twenty-four times between 1871 and the early years of the twentieth century. In the same period, imports of cotton cloth doubled. By 1900, China was the largest importer of cotton yarn and the second largest importer of cotton cloth in the world. Robert Cliver reports that the effect of these imports on Chinese hand-spinners was ‘devastating’, leading to massive protests in Guangzhou in the 1860s. By the 1930s, wages for spinners who continued their craft despite competition with machine-spun imports had plummeted to pennies a day. Because of foreign and domestic opposition, early efforts to establish cotton manufacturing on a mechanized basis failed. Chinese officials opposed them out of fear that they would lead to riots by handloom weavers. Foreign interests feared the loss of lucrative markets for the output of their factories. The first mechanized cotton mill in China was erected only in 1890 in the city of Shanghai and by 1915 there were about a million spindles in the country.18 The Indian subcontinent was the region of the world that was probably the hardest hit in the nineteenth century by the rise of the British cotton industry. Indian textile manufacturers first lost export markets all over the globe to British competitors who were now able to use the productivity of machinery coupled with the discipline of the factory to overtake the competitive edge that cheap food had given Indians in the eighteenth century. The process by which British cloth supplanted Indian stuff has been described in striking terms for the eastern Mediterranean. Halil İnalcık found that British manufacturers very deliberately undercut Indian cloth with lower priced imitations. This strategy was also followed in West Africa and South East Asia. Therefore, the Indian export trade of the eighteenth century blazed the trail for British cotton manufacturers in the nineteenth century. After competing successfully in Indian export markets, British yarn and cloth began to be sold in the subcontinent itself. Replicating a pattern that has been identified for other parts of the world, British competition first made itself felt as spinning and yarn began to be imported in substantial quantities from at least the 1820s. Cloth imports lagged behind, but reached sizable quantities by the midnineteenth century. Despite the new fashion in some quarters to criticize the thesis of nineteenth-century de-industrialization in India, it is my opinion that there was a substantial decline in textile manufacturing in the century. By the late nineteenth century employment in textiles may have approached the levels of one hundred years earlier (but this cannot be confirmed due to lack of data), but textile manufacturing was a far smaller proportion of total employment in the 17 Bortz, Chapter 13, this volume, ‘Mexico’. 18 Cliver, Chapter 5, this volume, ‘China’. 570

Global trade and textile workers Indian economy and represented a far smaller contribution to total income. By any definition of the term, this must be seen as deindustrialization. In the nineteenth century, because of its colonial status, India was subjected to free trade and became a captive market for British manufactures. And by century’s end, Indian consumers had become essential to the survival of Lancashire. As Thomas Stuttard of Burnley (in Lancashire) stated, ‘On India we rely, and if we lose India, Lancashire is practically ruined.’19 In 1900–1901, nearly two billion yards of cotton cloth were exported from Britain to India, a quantity which represented about 60 per cent of annual cotton cloth consumption in India. In the 1890s the Indian market absorbed nearly 30 per cent of British exports of cotton textiles, while some 70 per cent of the total cotton textile output of Great Britain was sent to foreign markets in those years. Combining these two figures, in the 1890s India consumed 20 per cent of the cotton manufacturing output of Britain!20 As was noted by Eric Hobsbawm nearly 40 years ago, global trade, especially to the protected markets of its empire, propped up the British economy, postponing Britain’s economic decline, and perhaps economic modernization.21 Other forces, however, were also working to topple Britain’s preeminence in the global textile trade. The nineteenth century was the century of cotton and this fibre dominated the global textile trade. Nevertheless, wool and, to a far lesser extent, silk continued to be exchanged across the world. In the case of wool, Australia arose as a major sheep-rearing region and began to supply Europe with raw material. And there was regional competition among wool manufacturers, especially in the Atlantic world where European producers competed against each for markets. And export markets for wool manufacturers were to be necessary for survival. In the region of Aachen, for instance, exports were about 70 per cent of total wool fabric production in the 1830s.22 Similarly, in the case of silk there was intense competition in East Asia between Chinese and Japanese silk producers. Although Chinese silk was held in high esteem in global markets in the nineteenth century, from the first decade of the twentieth Japanese silk began to dominate the world market. By 1909, Japanese exports surpassed those of China and in the 1920s accounted for 90 per cent of exports to the United States, the single largest market for silk. Government support was crucial to the success of Japanese silk producers, who improved sericulture techniques and established mechanized filatures. As a consequence, Japanese silk developed a reputation for superior quality, which aided it enormously in its competition against the Chinese manufacture.23

19 Quoted in Farnie, English Cotton Industry, p. 81. 20 Calculated from data in Farnie, English Cotton Industry, pp. 7 and 118. 21 Eric Hobsbawm, Industry and Empire: The Birth of the Industrial Revolution (New York, 1999), ch. 7. 22 Ebeling et al., Chapter 8, this volume, ‘Germany’. 23 Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’ and Cliver, ‘China’. 571

The Ashgate Companion to the History of Textile Workers

1913–2000 Britain came to dominate the global textile trade of the nineteenth century through a combination of machinery that dramatically reduced the cost of production, the intensified exploitation of labour through the factory system, and global political power, which allowed it to have preferential access to major markets. From the late nineteenth century, however, Britain’s cost advantage from machinery was challenged as low-wage countries, most critically India and Japan, adopted British technology and British labour regimes. As a result, from the late nineteenth century, the global textile industry came to have excess productive capacity, a problem that was to dog producers throughout the twentieth century. Excess productive capacity, in turn, was to have significant consequences for textile workers. A ‘modern’ mill industry, financed largely by local capital, began in India in the 1850s and was able to compete effectively against British cloth only to a very limited extent before the early twentieth century. Some of the handicaps it faced were lack of protection from the government of India, the unavailability of bleaching and dyeing facilities for the finishing of cloth, and the inability of Indian manufacturers to break into the long-established commercial networks that cloth importers had established.24 By the late nineteenth century, however, Indian manufacturers were able to compete against British yarn both at home and abroad. Before the First World War, Indian mills had their greatest successes in the lower counts of yarn and it was these coarser yarns that were profitably exported to China from the 1880s, displacing British yarn imports in that country. According to one estimate, in 1893–1894, 40 per cent of Indian machine-spun yarn was exported, much of this to China. From the early twentieth century, Japan became a major player in the global textile trade and the Japanese cotton industry came to rely upon foreign markets as a source of demand. From the cotton city of Osaka, Japanese yarn began to undercut Indian stuff in China and by the 1920s Japanese producers had substantially displaced Indian yarn in China and had begun to export to South East Asia, India itself and the Middle East, Africa and Latin America. By the 1930s Japan had become the world’s second largest producer of cotton yarn and cloth.25 At the same time, from the beginning of the First World War, there was a tremendous expansion of textile manufacturing in a number of regions. In part, the disruption of world trade and shipping during the war, which was followed shortly after by global economic depression, made it possible for many countries to commence spinning and weaving at home. In addition, there was protection of domestic industries in many parts of the world with the erection of tariffs on imported yarn and cloth. Even in India, tariffs were enacted on cotton products. As a result, in this period textile capacity and output grew rapidly in many countries, including Argentina, Uruguay, India, China and Turkey. In Argentina, between 24 Amiya Kumar Bagchi, Private Investment in India, 1900–1939 (Cambridge, 1972), pp. 229–37. 25 Hunter and Macnaughtan, ‘Japan’. 572

Global trade and textile workers 1920 and 1939 cotton yarn output increased from 645 tonnes to more than 29,000.26 In the case of India, cloth production nearly tripled, while imports fell by twothirds in those two decades. In China, the number of spindles increased from 3.6 million in 1922 to more than 5.6 million in 1936. From 1914 to 1945, therefore, the proportion of total world textile output that was entering global trade was most likely far lower than in the nineteenth and early twentieth centuries as production and consumption began to be centred within ‘national’ economies. This international situation was to continue in the first few decades after the Second World War, especially because in the 1950s and 1960s industrial development became the goal of economic policy in the underdeveloped world. To achieve this, barriers to imports of textiles were widely erected. As a consequence, the cotton industry in Britain went into decline from the mid-1950s, having lost export markets as well as facing competition from cheaper imports at home. Lancashire interests pressured both Labour and Conservative governments to enact protection to save the cotton industry, but neither party was willing to enact such policies.27 By the late 1960s, as a result, Lancashire was transformed from the workshop of the nineteenth-century world to a post-industrial wasteland. Decline was not universally the case in the post-war period, however. Japan rebuilt its cotton textile industry and by the late 1950s was back to the output levels of the 1920s. Cotton manufacturing in the United States found a lifeline in cheap non-union labour in the South and cotton plants took advantage of businessfriendly policies in states such as North Carolina, South Carolina and Georgia. Less expensive and less militant labour was effectively combined with investment in modern machinery to maintain US textile competitiveness until the 1970s. From the 1980s, however, even these surviving vestiges of past industrial greatness gave way to the pressures of global competition. In the case of Japan, cotton textile mills shifted to China where cheap labour and modern technology forged a formidable competitor. Between 1980 and 2000 Japanese textile production was halved.28 In the United States, the cheap labour of Mexico and other countries beckoned many US firms and there was a steady decline in cotton textile production in the 1970s and 1980s. Paradoxically the North American Free Trade Agreement (NAFTA) reversed this decline as it gave US manufacturers free access to the markets of other NAFTA members. In 1997, these countries received 45 per cent of US textile exports. As a consequence, in the late 1990s cotton yarn output had returned to the levels of the late 1960s and at century’s end US cotton yarn output accounted for about one-eighth of global production. Despite the revival of the US industry, by the end of the twentieth century, cotton textiles, a manufacturing activity that had loomed so large in the economic development of major industrialized countries, had largely abandoned its nineteenth- and twentieth-century homes. A new spatial distribution of this productive activity had emerged, in which new exporting centres took advantage of 26 Lobato, Chapter 2, this volume, ‘Argentina’. 27 Fowler, Chapter 9, this volume, ‘Great Britain’. 28 Hunter and Macnaughtan, ‘Japan’. 573

The Ashgate Companion to the History of Textile Workers cheap labour and low transport costs to become the textile workshops of the world. By 1998, China and the Indian subcontinent (India, Pakistan and Bangladesh) each accounted for about a third of the world’s production of cotton yarn, a picture that is reminiscent of that in the mid-eighteenth century.29 Exports from these new centres affected not only the United States and Japan, but also countries of the south. From the late 1970s, the Argentine textile industry, for example, was undercut by cheap imports of cloth from East Asia and by the 1990s, and the introduction of neo-liberal trade policies, textile activity in the country had virtually disappeared. This has reversed somewhat after the economic crisis of December 2001, which led to massive devaluation of the peso and steep drops in wages, making home production competitive once again.30 Another major shift in the global textile trade in the post-war period was the invention and rapid diffusion of synthetic fibres. These synthetics first began to be manufactured in the interwar period, but grew rapidly from the late 1940s. In 1960, synthetic yarn production in the world was about 3.3 million tonnes. By 1998,this figure had more than quadrupled to 13.8 million tonnes. Synthetics accounted for nearly 40 per cent of world’s fibre consumption and king cotton had declined to less than 50 per cent. These man-made fibres were produced from cellulose (rayon and acetate) or from petroleum (nylon, acrylic and so on) and the early centres of manufacture were in industrialized countries that possessed sophisticated chemical industries. In 1960 the United States, Japan and Europe accounted for 90 per cent of the total world output of these materials. By the late 1990s, sizable synthetic fibre industries were found in South Korea, China and India and these three countries accounted for a third of global production. Another 10 per cent of world production came out of Indonesia, Thailand and Turkey. The development of synthetics in the postwar period mirrors that of cotton in the nineteenth and twentieth centuries with the spread of the industry from a few centres to many areas of the world and the beginnings of a trade from the new centres to the old. Competitive advantage that derived from technological superiority gave way to the combined forces of technological diffusion and cheap labour in the global market for synthetic fibres. A final major shift in the global textile trade is the explosion in the international trade in garments from the 1950s, which has led to garments displacing textiles in international trade. From the 1950s, according to one source, ‘exports of clothing began to expand more rapidly than exports of textiles proper and increased nearly seven times as fast (1954-84) as exports of textiles.’31 In 1987 world trade in garments overtook that in textiles. By 1997, the major apparel exporters were in East Asia, with China, Hong Kong, South Korea and Taiwan accounting for about one-third 29 Calculated from B.R. Mitchell, International Historical Statistics: Africa, Asia & Oceania, 1750–2000, 4th edn (Basingstoke, 2003); B.R. Mitchell, International Historical Statistics: Europe, 1750–2000, 5th edn (Basingstoke, 2003); B.R. Mitchell, International Historical Statistics: the Americas, 1750–2000, 5th edn (Basingstoke, 2003). 30 Lobato, ‘Argentina’. 31 Richard M. Jones, The Apparel Industry (Oxford, 2002), pp. 81–82. 574

Global trade and textile workers of global exports. In these four countries garments were largely manufactured from locally produced cloth. Therefore, they took advantage of advantageous labour costs twice: first in the manufacture of the cloth itself and then in the cutting and sewing of that material into an item of clothing. Therefore, since the Second World War textiles has been vertically integrated with apparel manufacturing in order to take advantage of vast global differences in wages and standards of living.

Conclusions Regional and even intercontinental trade had long been important in textile manufacturing. From the mid-seventeenth century, however, a world market developed for the first time in textiles, spearheaded by the cotton cloths of the Indian subcontinent. This was a world trade because the buyers for this cloth were truly distributed throughout the globe, from East Asia to the Americas with many points in between in Europe, West and East Africa and West Asia. From the seventeenth century the global trade in textiles, most crucially those made from cotton, remade textile manufacturing throughout the world, transformed the lives of textile workers and changed the buying habits of consumers around the planet. From the late seventeenth century, Europeans protected themselves from global competition by enacting restrictions on imports of Indian cottons. European manufacturers then began to imitate Indian goods. These efforts were successful from the 1770s after the invention of machines for spinning yarn and the development of the factory system. On the basis of these devices and institutions, the centre of the world’s textile industry moved from the Indian subcontinent to the British Isles. A long-standing Indian competitive advantage that had rested on agriculture was replaced by one reliant upon technology and intensified work. In the nineteenth century, British textile manufacturing relied upon demand from global markets to maintain high levels of output. In the early days of the industry, these markets were in Europe, where British goods were shut out from the early years of the nineteenth century, then in America, where the same exclusion took place in several countries, including the United States, Mexico and Brazil, and finally in Asia. In this final place the captive colonial market in India would prove to be the last great stronghold of British cloth. Over the course of the nineteenth century, however, the British advantages of superior technology and work organization were eroded. Both with and without government assistance, textile technology and work systems were diffused throughout the globe. Once machines were put to work with far cheaper labour in India and Japan, British advantages were eventually erased. By the early twentieth century the heart of the world’s textile industry was winding its way back to Asia, where it had been in the eighteenth century. From the 1970s, the Asian shift was to accelerate as China and the Indian subcontinent developed enormous textile manufacturing capacity.

575

The Ashgate Companion to the History of Textile Workers Since the late nineteenth century, the global economy has had excess capacity in textile manufacturing, which has placed enormous pressure on the wages and work lives of textile workers. From the late nineteenth century, British workers began to be out-competed by the cheaper workers of India and then Japan, initially in the yarn markets of China, but increasingly elsewhere as well. In the United States, after the Civil War, low-wage labourers in the South undercut high-cost workers in New England. A similar pattern has unfolded since the 1970s when textile production in Japan and the United States declined in the face of low-wage production in China, Mexico and elsewhere. Yet nothing in textile manufacturing is fixed for long, as may be seen by the revival of US production in the 1990s with the new opportunities presented by NAFTA and in Argentina in the wake of the economic collapse of late 2001, which slashed wages and the value of the peso. By the start of the twenty-first century, United States manufacturing, however, was under threat from the Chinese industry as textile import quotas by nation were to be eliminated on 1 January 2005. Chinese textile mills combined sophisticated technology with low wages to ‘overwhelm competition’, in the words of one observer.32 Out of necessity, a global history of textile workers must, therefore, include the global economy and the global trade in textile goods.

32 ‘Textile Quotas to End, Punishing Carolina Towns’, New York Times, 2 November 2004, p. A1. 576

23 Proto-industrialization and industrialization and ‘modernity’ in a global perspective Donald Quataert

In discussing the themes indicated in the above title, this chapter begins with comments at the most general level of analysis and ends with specific factual comparisons. The primary intent is to offer questions that have arisen from reading the national overviews. In the final section, however, I summarize some of the findings offered by specific national overviews.

Why discuss textile industrialization? It seems legitimate to inquire, in the first place, into the reasons for this volume’s investigation into textile production and textile-makers. After all, beyond the intrinsic interest in the subject of textile production and producers, why should or do we explore this subject? At least hypothetically, the nature of the overall project implies that success or failure in creating a mechanized industrial infrastructure is linked to broader questions of power. Thus, for the earlier part of our period, until approximately 1900 or so, we need to consider the links between industrial power and political and military power and whether the presence of the former is necessary for the latter. Were there politically and militarily powerful states in the earlier part of our era that did not become major textile producers? And, as a corollary, what is the relationship between textile production and political/military power towards the end of our era, c.1900–2000? After all, the division of the world between the economies that industrialized and those that did not industrialize is hardly an irrelevant one. If proto-industry – the definition of which is discussed below – does/did lead to industrialization, we do need to try to understand why and how this happened. That is, protoindustrialization is a form of production that in its own right might well deserve inquiry. These patterns of proto-industrialization are important because they

The Ashgate Companion to the History of Textile Workers seemingly have value for a basic problem of the global age – namely inequality, the huge disparities of wealth existing in the world. Why are there rich nations and poor nations? If the key to these fates rested in the manufacturing sector, then we need to understand the patterns of industrial production, even as disparities of wealth are being altered by new forms of wealth today – for example information technology. How to deal with the issue of development must be at top of the global agenda. Thus, the debate over proto-industrialization may have some relevance for people in the World Bank, the UN and other agencies dispensing funds. If past patterns of proto-industrialization can point the way to likely manufacturing successes in the future, then we need to understand these patterns of success and of failure. Hence, our identification of small-scale workshops staffed with women’s labour might provide guidance to development agencies as they seek likely candidates for wealth-inducing patterns of production. Thus, at the global level, a discussion of proto-industry might be about jobs, gender tensions, quality of life and levels of living.

Proto-industry and industry Also, we need to explore the challenges offered by the editors of this volume and its preceding conference when they charged the writers of the national overviews and the comparative chapters to look at ‘proto-industry’ and ‘industry’ together? Certainly both are intrinsically interesting subjects in their own right and later I will try to summarize some of the rich narratives that appeared in the national overviews. But, at this juncture, I wish to discuss some of the issues raised when the editors paired up the rubrics of proto-industrialization and industrialization. Would the conference and the volume have been different if their charge had been to comparatively investigate the processes of manufacturing, without distinguishing between proto-industry and industry? Part of the difficulty lies in the shifting meanings undergone by the term ‘proto-industry’ since its introduction by Franklin Mendels in 1972. As seems evident in several of the national overviews – notably those on the Netherlands and Brazil – the editors’ pairing of the terms proto-industrialization and industrialization was understood to mean a connection or relationship between the terms, or processes. If I have read them correctly, these seem to be arguing for proto-industrialization as a stage in the road towards industrialization. On the other hand, some other national overviews, for example that concerning Argentina, specifically argued against such a necessary linkage.

 Franklin Mendels, ‘Proto-industrialization: The First Phase of the Industrialization Process’, Journal of Economic History (March 1972): 241–261.  Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’; Delson, Chapter 4, this volume, ‘Brazil’.  Lobato, Chapter 2, this volume, ‘Argentina’. 578

Proto-industrialization and industrialization Several points seem relevant here. Firstly, the precise meaning that the authors of the several national overviews have attached to the term proto-industry is not entirely clear to me. The rich and honourable history of the phrase dates back to Mendels, who seems to have changed his mind about the significance of proto-industrialization and its meaning as debate raged over the concept during the 1970s and 1980s. In his original formulation, I believe that Mendels at least implied a necessary link between proto-industry and the subsequent appearance of mechanized, factory-based manufacturing. By the early 1980s, however, he seems to have stepped away from a stage theory approach, likely because of the scholarship that his own work had generated. For example, scholars during the intervening years had shown that it was not merely the presence of proto-industry (however defined), but also of the necessary resources, notably coal, that often explained the emergence of industrialization. They also had demonstrated that, sometimes, proto-industry did not evolve into mechanized factory production. In his later formulation, Mendels cautiously said ‘historians find very few regions of Europe that had an industrial revolution without first going through a phase of proto-industrialization’. Despite all the nuances that have crept in over the past decades, the phrase itself – proto-industrialization – still seems to suggest or imply acceptance of a relationship between industrialization – meaning manufacturing by mechanical means – and certain kinds of manufacturing that went before. It continues, to my ears at least, to imply that proto-industrialization is a form of organization that necessarily precedes, paves the way for, makes possible, causes, gives way to industrialization. As a corollary, it seems to imply a lack of concern with the pre-mechanical forms of production that did not lead to mechanized factory production. And, it does not appear to accommodate itself very well to the non-mechanized production that is so prevalent in our own age and such an inherent part of textile manufacturing in age of mature capitalism. It seems to me that embedded in the pairing of the terms proto-industrialization and industrialization remains a sense of success or failure, approval for economies that successfully industrialized and disappointment in those that did not. This procedure seems rather judgmental to me and thus I am not entirely comfortable with the term. Indeed, there is abundant research showing the persistence of household as well as small workshop manufacturing well into the age of factory-based machines, indeed to the present day. There are very many examples, some of them detailed below, of substantial household-based, hand-weaving, manufacturing activities deep into the twentieth century. Here, I also would point to the example of the Saxon Oberlausitz that, according to imperial and Saxon state records, was among the most important textile manufacturing states of Germany at the turn of the twentieth century. For a long  For a summary and advancement of the research at the time, see Peter Kriedte, Hans Medick and Jürgen Schlumbohm (trans. Beatte Schempp), Industrialization before Industrialization: Rural Industry in the Genesis of Capitalism (New York, 1981).  See Mendels’ response, in The New York Review of Books 31 (16) (1984), to Lawrence Stone’s review. 579

The Ashgate Companion to the History of Textile Workers time, virtually all of this manufacturing was located in households, where both men and women wove, took care of the family garden and did household chores. The commitment of these households to weaving perhaps is made clearer by the following illustration. In this region, large looms were built into homes encased in their own frameworks (rooms within rooms in a sense), so that the vibration from the work of the loom did not pass into the living area of the house. The substantial investment such a structure requires indicates the involvement of these households in commercial production. Moreover, when mechanized factories opened in the region, the result was not urbanization and the collapse of these rural households but instead their retention. The households remained intact and from these both men and women commuted, on the newly built railroad system of the region, to the factories. In this example, industrialization, in the sense of factory-based and machine-driven manufacturing, did not result in the emptying out of the countryside but rather its continued occupation by household communities intent on self-preservation. If we abandon proto-industry as a useful term, should we instead turn to ‘puttingout’ as a more useful phrase? As I understand it, putting-out means the small-scale commercial manufacture of goods located in homes and workshops of varying sizes and organized to a greater or lesser degree by merchants. Originally, almost all of the inputs in the putting-out process were made by animate sources of power, that is by hand, sometimes supplemented by water power. Later on, machine-made inputs became an integral part of putting-out – for example machine-spun yarns or factory-produced dyestuffs. And sometimes machinery itself entered the homes as well as the workshops, for example in the form of powered looms or foot-powered sewing machines. In this definition, putting-out manufacturing is not seen as a stage of production leading to the factory but rather as a permanent feature of the manufacturing process. It is not necessarily a phase or stage although, I suppose, sometimes it historically has been. As the national overview on China makes clear, the contemporary China of our own age certainly is a successful textile producer and yet, clearly, its origins do not rest in its manufacturing past. Indeed, the author at one point flatly asserts: ‘The “putting-out” system that developed in North China in the early twentieth century increased merchant control over production without creating a wage-based factory system.’ So, is the term putting-out, by not implying a stage theory leading to mechanization, preferable to proto-industry? The likely answer is ‘No’, since putting-out always implies the presence of merchants in the production process.  See the following three articles by Jean H. Quataert: ‘Combining Agrarian and Industrial Livelihood: Rural Households in the Saxon Oberlausitz in the Nineteenth Century’, Journal of Family History (Summer, 1985): 145–161; ‘The Shaping of Women’s Work in Manufacturing: Guilds, Households and the State in Central Europe, 1648–1870’, The American Historical Review (December 1985): 1122–1148; and ‘The Politics of Rural Industrialization: Class, Gender and Collective Protest in the Saxon Oberlausitz of the Late Nineteenth Century’, Central European History (June 1987): 91–124.  Cliver, Chapter 5, this volume, ‘China’. 580

Proto-industrialization and industrialization Thus, it neglects other agencies, including the producers themselves, through which the manufactured good reached the market. And so, in the end, neither ‘proto-industry’ nor ‘putting-out’ seem terribly useful as terms delineating the non-mechanized manufacturing experiences before (and after) the rise of machine-based textile production. In the end, we may need to be content with rather vapid phrases, such as ‘hand-manufacturing’, to encompass these pre-machine experiences. Similarly ‘commercial hand-manufacturing’ may best describe the production processes involved in making a commodity for sale, whether directly by the producer or through a merchant or others. Whatever term we use, it remains important that we discuss manufacturing that is not factorybased or proletarian in style in a manner that is non-judgmental, that does not praise or condemn for the production process not reached. The above comments, moreover, underscore the importance of being very precise about chronology. When discussing textile manufacturing and the role of proto-industry, putting-out or whatever we call manufacturing outside of mechanized factories, we need to distinguish between hand-manufacturing in the era before the appearance of economies dominated by big mechanized factories and the hand-manufacturing networks that existed afterwards. More concretely, does or does not hand-manufacturing in the age after the emergence of industrial economies play a different role than it did in a world where there were only one or two mechanized industrial economies – mainly Britain? Are the putting-out networks that amassed capital in pre-industrial economies functionally the same as putting-out networks in contemporary places like Istanbul, Turkey?

A critique of de-industrialization A number of the national overviews offer hefty challenges to the notion that industrial production structures outside of Europe (and the United States) collapsed in the face of mechanized competition from the developed world. Persuasive evidence against de-industrialization emerges from quite disparate economies and political orders including India, the Ottoman Empire, Egypt, Uruguay and, perhaps, Mexico.10 In the example of Uruguay, during the 1970s textile producers moved out of the factories back into the homes in response to unfavourable conditions.  I am grateful to the editors of this volume for their comments on my use of the terms ‘proto-industry’ and ‘putting-out’.  For an early exploration of this issue in Turkey, see Çağlar Keyder, ‘Afterword: The Current Conditions of the Popular Classes’, in Donald Quataert and Erik J. Zürcher (eds), Workers and the Working Class in the Ottoman Empire and the Turkish Republic, 1839–1950 (London, 1995), pp. 147–158. 10 Roy, Chapter 19, this volume, ‘India’; Quataert, Chapter 18, this volume, ‘Ottoman Empire’; Beinin, Chapter 7, this volume, ‘Egypt’; Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’; Bortz, Chapter 13, this volume, ‘Mexico’. 581

The Ashgate Companion to the History of Textile Workers But such a decline in factory production did not mean an overall reversal in textile output. Instead, as the author notes, the industry became characterized mainly by ‘networks of informal homeworkers who did piecework for companies that sold their production in the domestic as well as foreign markets’.11 In this regard, the Uruguayan example remarkably parallels that of modern Turkey, where hidden home and small workshop production provides a vital segment of production.12 In Egypt, a European director-general of the Department of Agricultural, Industrial, and Commercial Education surveyed the textile industry in 1910 and made the following observation. ‘One can affirm that the textile industry, far from declining as is often claimed, is in a state of growth and, to the contrary, is first among the local industries.’ There, in Egypt, textile production was maintained through ‘expanded use of the putting-out system and feminization and ruralization of production’.13 In the cases of both the Ottoman Empire and India (and certainly others), historians’ understanding of the industrial past had been distorted by using mainly Western sources to describe the Ottoman and Indian economies. Such sources, not unnaturally, primarily discussed Ottoman and Indian trade in finished textile products with Europe and detailed a dismal picture of their collapsing exports to Europe. Indeed, these are accurate descriptions, as far as they go. But, because of their sources, they over-emphasized the importance of this export trade in textiles for the Ottoman and Indian economies and thus distorted the overall picture of their respective textile sectors.14 Indeed, by ignoring the two respective domestic markets, historians had left out most of the textile industry in the two regions. In the examples of the Ottoman Empire and India, the domestic market played the key role and international exports had been only a small part of the textile economy. Indeed, the cases of India, the Ottoman Empire, Egypt and Austria all indicate that stress on exports likely distorts the picture of many local textile industries. As Roy puts it: At the end of the eighteenth century, cloth exports from India amounted to about 50 million yards, whereas total production within India could not have been smaller than 1,800–1,000 million yards.15 And so, even though the export markets for both regions fell away at different times, the domestic markets remained largely intact and so did the overall textile sector. In early twentieth-century China also, domestic weavers provided most of the cotton cloth being purchased, even with the ready presence of imports.16

11 Camou and Maubrigades, ‘Uruguay’. 12 Jenny White, Money Makes Us Relatives (Austin, 1994); Keyder, ‘Afterword’. 13 Beinin, ‘Egypt’. 14 Quataert, ‘Ottoman Empire’; Roy, ‘India’. 15 Roy, ‘India’. See also Quataert, ‘Ottoman Empire’; Beinin, ‘Egypt’; Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’. 16 Cliver, ‘China’. 582

Proto-industrialization and industrialization A common key survival strategy in both the Indian and Ottoman cases was the adoption of imported yarns. As several authors note, spinning was an extremely labour-intensive activity and the abandonment of cotton yarn hand-spinning is a net gain, not loss, for the economy. (The China example of job losses and spinners’ riots, however, shows the human costs.) As Roy puts it: ‘The ruin of hand-spinning did see massive loss of jobs, but rather little loss of income.’ 17 Thus, the powerful emphasis that Western authors and those under the influence of development theory often give to the loss of cotton hand-spinning jobs cannot be used to argue for de-industrialization. The lost jobs in spinning were more than made up for by productivity increases elsewhere in the textile sector. The paths that Ottoman and Indian textile producers respectively took to survive seem somewhat different, likely because of differing comparative advantages. In the Ottoman Empire, mechanized spinning factories were built but, c.1914, supplied only about onequarter of all cotton yarn consumed. The major Ottoman successes lay elsewhere: textile manufacturers won new export markets in knotting carpets, reeling raw silk and knitting lace. They also retained most of their domestic consumers using low technology imports such as cotton yarn and factory-made dyestuffs. Moreover, producers paid close attention to local taste. Increasingly large numbers of women and girls wove, usually in rural homes and workshops linked by various puttingout networks.18 But in India, a rather vibrant mechanized sector emerged on the world market during the nineteenth and early twentieth centuries while, at the same time, handloom weaving remained strong. As Roy indicates, the volume of handloom cloth production more than doubled between 1900 and 1940.19 Overall, the various national overviews have offered considerable evidence in favour of abandoning de-industrialization as a construct. Thus, I concur with Roy’s conclusion that de-industrialization has not proven to be a very useful concept. Like proto-industry, de-industrialization as a term seems judgmental as well as tautological. De-industrialization also suggests a stage theory of industrial development where there are absolute winners and losers in a race to acquire the attributes of modernity that include mechanized factories full of proletarianized workers. And yet, global experiences with textile production reveal quite different trajectories involving large numbers of putting-out workers in the home and workshop. Indeed, de-industrialization often is an ideological construct that a priori presupposed the decline of hand-manufacturing because it was perceived to be a form of industrial organization inferior to factories. More concretely, since most mechanized factories until the twentieth century were located in Europe, the binary was not merely mechanized vs non-mechanized but Western vs non-Western. And so, de-industrialization was part of the ideological arsenal used to assert and thus prove the superiority of Christian Europe over the rest of the globe, thus justifying its political domination.20 17 Roy, ‘India’. 18 Quataert, ‘Ottoman Empire’. 19 Ibid. 20 See Donald Quataert, Ottoman Manufacturing in the Age of the Industrial Revolution 583

The Ashgate Companion to the History of Textile Workers Obviously not all allegations of de-industrialization are false or ideological. Some economies did de-industrialize – as one historian told me, one need only recall whole regions of Austria that were once covered with putting-out networks but are now agrarian countryside.

Examples from the national overviews In this level of analysis, that of factual comparisons, I take empirical data from the various national overviews to examine a number of different subjects.

Guilds The core issues surrounding guilds, that is, their prevalence and importance across the globe, deserve a closer look. Many of the national overviews point to the presence of guilds in cities as an important ‘push’ factor in the rise in textile production in rural areas or, sometimes, in other cities without guilds. Too often, I must add, they assume guilds to be exclusively male organizations and that they always had been so. Quite the contrary, women had long been an integral, if often minor, part of guilds in many European countries. Here, a discussion of the actions that drove women out of guilds on the European continent could have provided a useful point of departure for the nature of guilds around the world.21 The decline of the guilds offers rich ground for comparative analysis. I was struck by the discussion of guilds in Denmark and their fate, and the strong similarity to developments within the Ottoman Empire, two otherwise quite dissimilar places! In Denmark during the late eighteenth century, the emergence of a capitalist economy was beginning to undermine the guilds. As production became concentrated in larger units, the number of journeymen in relation to masters rose. Thus, a journeyman no longer could expect to eventually become a master himself. Guilds proved unable to handle the new types of disputes and mounting contradictions between journeymen as wage workers and masters as employers. ‘Furthermore, from the point of view of the proponents of liberalism, the guilds were harnessing free trade and constituted an obstacle to economic development.’22 Consequently, the liberal Danish government removed the rights of the guilds to control trade and made membership voluntary in 1862. These events are remarkably similar in both their chronology and detail to events surrounding textile guilds in the Ottoman Empire, and to the fate of those guilds in Damascus. In the 1860s, the central Ottoman (Cambridge, 1993), esp. the Introduction and Conclusion. 21 Sheilagh Ogilvie, ‘How does Social Capital affect Women? Guilds and Communities in Early Modern Germany’, American Historical Association Review (April 2004): 334–343; see also Quataert, ‘Politics of Rural Industrialization’ and the sources therein. 22 Christensen, Chapter 6, this volume, ‘Denmark’. 584

Proto-industrialization and industrialization state began formally dismantling the guilds by consolidating previously separate guild organizations into manufacturers’ associations. At that moment, journeymen weavers in Damascus guilds were encountering the same downward mobility as their Danish colleagues who found themselves without masterships into which they could move. And so they found themselves employed as wage workers. Over the remainder of the century, journeymen increasingly struck against the masters and the Damascus guild structure fractured.23

Technology On the issue of technology, it seems worthwhile here to briefly make several different points that will be taken up more fully below when discussing the organization of production. It is my impression that England, the United States and Germany were the major technological innovators in the textile industry during the era up until about the First World War. Elsewhere, a great deal of technology transfer occurred and only rarely were local innovations present. Nineteenth-century Japan seems to have been the great exception, as Hunter and Macnaughtan indicate in Chapter 12. The growth of production was sustained by a number of indigenous technological advances, which helped to increase both output and productivity. Some of these advances came in the sericulture process, but probably more important were advances in weaving technology, in particular the spread of the so-called ‘tall loom’ across the country by the mid-nineteenth century, offering greater efficiency than the traditional looms.24 Less successful innovative efforts occurred in the Vienna-based Habsburg Empire. Komlosy notes attempts to develop locally made competitive spinning machines. These innovative initiatives multiplied after the 1780s, thanks in part to government support and the granting of privileges for the use of certain devices. These efforts notwithstanding, there was not a decisive Austrian breakthrough in inventing an effective spinning machine.25 Also, during the first wave of mechanization that I define as lasting until the First World War there seems to have been a global chronological pattern regarding the respective adoption of mechanized spinning and weaving. The precise dates at which mechanized spinning and then later weaving occurred varied considerably, depending on the case in question. But in all the instances that I noted, the adoption of mechanized spinning preceded that of weaving by several or more decades. The national overview on China nicely summarizes the reasons for the relative lag.

23 Quataert, Ottoman Manufacturing; Sherry Vatter, ‘Militant Textile Weavers in Damascus: Waged Artisans and the Ottoman Labour Movement, 1850–1914’, in Quataert and Zürcher, Workers and the Working Class, pp. 35–57. 24 Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’. 25 Komlosy, ‘Habsburg Monarchy’. 585

The Ashgate Companion to the History of Textile Workers Whereas mechanized spinning quickly replaced hand-spinning due to a gain in efficiency of some 32 times, power-loom weaving was only four times more efficient than handloom weaving, and required much greater investment in machinery, training, maintenance and industrial plant.26 Thus, in Catalonia, the mechanization of spinning proceeded apace dating from the 1810s. This process was accelerated by the introduction of steam power, especially from the 1840s; and in Barcelona and other textile towns this also promoted the use of power-looms and the vertical integration of spinning and weaving within larger plants. By 1860, spinning had been totally mechanized while power-looms accounted for half the industry’s total output.27 The introduction of Western cotton spinning technology in Japan brought with it potential economies of scale, and from the turn of the century many cotton spinning firms employed in excess of a thousand workers. The scale of operation grew over time, particularly as many cotton spinning firms extended their operations into weaving, resulting in integrated mills. By the interwar years many of these mills were of very considerable size, and even some specialist weaving companies in the local production areas (sanchi) were employing hundreds of workers in a single mill.28 Similar patterns are present in Austria where between 1815 and 1873 mechanized spinning and printing expanded gradually. Furthermore, it spread from cotton to other fibres involving more and more people and locations in the process of industrialization, which increased the number of proletarian workers. Although the mechanization of weaving started in the 1840s, until the crisis in 1873 weaving largely remained a manual activity which entailed the simultaneous expansion of semi-proletarian hand-weaving.29 In England, there was a delay of as much as half a century between the widespread mechanization of spinning and of weaving. There, weavers only began to be replaced by power-loom weavers in mills as the pace of technological development accelerated in the 1820s and 1830s and was not complete until the 1840s. The cases of China and India, however, seem rather different from the trajectories of other areas. As we learned, mechanized spinning arrived rather later in China, at the end of the nineteenth century. In my view, this was partly due to the unfavourable international treaties that permitted the influx of cheap foreign yarns. It appears remarkable that hand-spun still accounted for a full one-quarter of Chinese cotton yarn needs as late as the 1930s. In China, mechanized weaving accounted for only about one-half of output in 1949. ‘Before the Second World War, household handicraft weavers continued to produce most of China’s cotton cloth.’30 These facts suggest that, if we had held our conference back in 1950, likely we all would have agreed that China’s textile manufacturing future was bleak indeed. 26 Cliver, ‘China’. 27 Smith et al., Chapter 17, this volume, ‘Spain’. 28 Hunter and Macnaughtan, ‘Japan’. 29 Komlosy, ‘Habsburg Monarchy’. 30 Cliver, ‘China’. 586

Proto-industrialization and industrialization Similarly, as seen above, Indian hand-weaving remained important and was expanding in the post-1900 era.

Organization of production I will begin with some brief comments on the use of slavery and on the gender composition of the workforce and then will turn to the subject of the organization of labour.

Slavery

The various national overviews reveal a host of different forms of labour that produced textiles during the period 1650–2000. In Argentina, the Spanish colonial authorities used the forced labour systems of the encomienda. These workers, clearly, were legally unfree whereas the wage workers in the spinning mills of eighteenth-century England, nineteenth-century India or the early twentiethcentury Ottoman Empire certainly were free workers in the legal sense.31 Whether their levels of living and conditions of existence were appreciably different from these encomienda workers is quite a different matter, one that was not pursued, if I recall correctly, in the various national overviews. To me, it is an open question as to whether these workers were more or less freer than the piecework females and males working in British factories during the 1840s or the sweatshop immigrant workers in contemporary New York City. Likely, the hourly wage workers in the spinning and silk reeling mills in China, the Ottoman Empire, England and a host of other countries enjoyed more personal freedom than the encomienda workers or the pieceworkers just mentioned. Except for Delson’s chapter on Brazil, the presence or absence of slave labour is not explored in any of the national overviews. It is possible that slaves were not part of the textile production processes in any of the regions explored by the several cases. In the New World, a centre for capitalist slavery, slave labour mainly worked in agriculture, for example producing the sugar that generated such vast profits. As is well known, the textile industry of early modern Europe vastly expanded output to clothe these sugar plantation workers. Slavery was commonplace in most other global areas. In the Ottoman Empire, there were many slaves but these were used almost always for domestic purposes. In the sixteenth century, however, some slaves wove expensive and elaborate silks for the palace and elite consumption. But, thereafter, slaves did not play any role in Ottoman textile production. Whether they did in the Indian, Chinese or South Asian textile sectors seems worthy of further exploration.

31 Lobato, ‘Argentina’; Roy, ‘India’; Quataert, ‘Ottoman Empire’. 587

The Ashgate Companion to the History of Textile Workers Gender

Regarding the gender composition of the textile workforce, there is a comparative chapter on the subject. Therefore, my remarks will be very brief. I begin by quoting from the Danish national overview. Christensen offers the following statement: There is no single cause for the distribution of job functions in relation to sex within the textile industry. As women carried out some of the most mechanized jobs (such as cotton spinning) as well as some of the least mechanized (such as burling), no simple relationship between mechanization and ‘feminization’ existed.32 On the one hand, his analysis seems undercut by some of the other national findings, notably those on Britain and Spain. In Spain, the use of spinning jennies had led to the employ of women in the factories. But then, import of the new mule-jennies after 1814 signified a profound revolution in this respect. They were large semi-automatic machines, unlike anything that had gone before, which required considerable physical effort and the spinner to work alongside a team of youngsters, known as piecers, who changed the spools and mended the broken thread. As in other European countries this encouraged the replacement of women by men – ‘heads of families’ in the spinners’ own words – within the trade. The process has not been studied in any detail. Crucial, no doubt, was the physical strength required to operate the new machines. In addition they made it possible for the spinners to play a supervisory role and train their piecers. Hence their position could be assimilated to that of an artisan … This situation was to change in the 1850s with the introduction of the self-acting mules. These were fully automatic and required less physically effort. Hence, the temptation among employers to cut costs by replacing male spinners with female spinners. By 1856 in Barcelona, while men operated 66.7 per cent of the mule-jennies, they were employed on only 21.2 per cent of the self-acting mules.33 This evidence seems quite solid – that the various mechanized spinning technologies led to different mixes in the gender composition of the workforce in Spain (and similarly in England and at Fall River in the United States). Namely, spinning jennies led to more female labour, while mule-jennies used mainly male labour which then decreased in importance as self-acting mules favoured female labour. Whether or not these patterns also appear during the technology changes of regions such as India, China, the Ottoman Empire and modern Turkey seems to be a subject well worth pursuing. 32 Christensen, ‘Denmark’. 33 Smith et al., ‘Spain’. 588

Proto-industrialization and industrialization On the other hand, I have a certain measure of discomfort over the assumptions, present in both the Spain and the India cases, regarding the gender division of labour in the household.34 The authors of both seem to assume a fairly rigid gender division of labour in which only women spun while men wove. My own research on the Ottoman carpet-making industries works against the assumption in the Spain and India national overviews and seems to support our Danish colleague’s assertion that technological change, by itself, is not a sufficient explanation for the gender composition of the textile production workforce in an area.35 My research traced the explosive growth of Ottoman carpet-making during the nineteenth century, thanks to European and American fascination with what they mistakenly thought were handmade works of art. At the turn of the twentieth century in some towns of western Anatolia only young girls and women knotted the rugs. But in other towns quite nearby, both men and women knotted the rugs. Thus, while in the Ottoman world generally rug knotting was seen as a female task, there were important exceptions. The situation in contemporary Iran, also an important centre of the rug industry, could not have been more different. There usually only men knotted the rugs.36 These remarks about the gender switchability of carpet-making tasks seem relevant to the comments in the national overview on the Netherlands about the significant presence of male spinners in Tilburg in 1810. The authors are surprised to find so many male spinners since ‘it is usually supposed that only women spun because they represent much cheaper labour’.37 Moreover, in very many of the national overviews presented, as just seen, spinning is stated to be women’s work, because of the ‘self-exploitation’ factor. And yet there is considerable additional evidence from elsewhere of gender switching of tasks in the textile industry. Research on the Saxon Oberlausitz decisively demonstrated that both men and women in their household cleaned house, tended the garden and wove for commercial sale. They routinely interchanged tasks, depending on the needs of the moment. This is very well established. But then, the state arrived at the household doorstep, for purposes of registering the members for its old age and worker insurance programme. And so, to avoid paying these insurance costs, the families reported the women’s work in a way that made it appear they were not working at all. Thus, they successfully conspired to not pay the insurance fees, which they could ill afford.38 My point here is: doesn’t this example suggest that women’s textile work has been under-reported or not reported all in many other locations of the globe? And so, because of these Oberlausitz findings, the studies on Iranian carpet workers and my research on the various Ottoman carpet-makers,

34 Ibid.; Roy, ‘India’. 35 Christensen, ‘Denmark’. 36 Donald Quataert, Ottoman Manufacturing, pp. 134–160; Leonard M. Helfgott, Ties that Bind: A Social History of the Iranian Carpet (Washington, 1994). 37 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 38 Jean Quataert’s three articles cited in n. 6 above. 589

The Ashgate Companion to the History of Textile Workers I am delighted with the findings from Tilburg regarding male spinners but I do not find them surprising. Thus, the various national overviews present two quite distinctive arguments pointing in very different directions. The one indicates a strong correlation between gender and technology while the other denies such connections. Allow me one final remark on the relationship between technology and gender, derived from the chapter on Mexico. In the case of Mexican textile factories before the contemporary age, we find mainly males working. But, nowadays, in the maquiladoras textile factories of the global age that sprawl along the US–Mexican border, a full 60 per cent of the workforce is female. And yet, the technologies in the earlier factories and those in the maquiladoras do not seem very different. Here the presence or absence of labour unions rather than technology seems important in explaining the presence or absence of men or women workers. Clearly technology and gender remains an important subject of additional study.39

Home, workshop and factory

Whenever the authors of the national overviews offer sufficient specificity on the pre-1900 period, they illustrate, for their respective national studies, the simultaneous presence of textile production in homes, workshops and factories, of both the un-mechanized and mechanized variety. In the chronologically later parts of their presentations, many point to the crucial persistence of home and workshop manufacturing, even in the age of massive factories with computerized spinning, weaving and knitting. In this regard, A.V. Chayanov’s notion of self-exploitation remains useful for explaining household production of textiles below subsistence cost. I also would point to the global use of real or fictive family ties to recruit and discipline labour. For example, the earliest filatures in rural South China hired young women from the surrounding villages, who were encouraged to go out to work by their lineage organizations.40 Other hierarchical authority figures also played key roles in the recruitment of women and girls for factory labour, if the mid-nineteenth-century examples of Ottoman Bursa in west Anatolia and of Ottoman Lebanon have any universal applicability. In Bursa, the Roman Catholic Pope issued a special decree affirming the morality of silk mill labour for the local Armenian women while, in the Lebanon, ranking ulema and local archbishops similarly approved the employ, respectively, of Muslim and Christian women and girls in the reeling factories.41 In Brazil, by the late eighteenth century textile production showed two distinctive, coexisting configurations: ‘home/fazenda production for self-contained use, which continued on much as it had in the earlier centuries, and proto-industrial factory production’.42 Production in homes and on estates remained dependent on 39 Bortz, ‘Mexico’. 40 Cliver, ‘China’. 41 Donald Quataert, Ottoman Manufacturing, pp. 128–129. 42 Delson, ‘Brazil’. 590

Proto-industrialization and industrialization spinning distaffs and looms but the state-run factories employed more complex, likely imported, machinery. Wage-earners in factories were both male and female but largely female in home production.43 Not dissimilarly, Mexico in the colonial period contained three forms of production: ‘household production, which the Spaniards called telares sueltos; small non-household artisan production, trapiches or obradores; and larger workshops, obrajes, with both forced and free labour’.44 During the period 1821–1910, industry became more concentrated, so that the scattered workshops yielded to a concentration of factories in a few textile zones. Therefore, the Mexican case seems to argue for the proto-industrialization development. In Spain, masters and artisans who worked in what remained of the urban workshops were men who usually had undertaken apprenticeships and worked in guilds. In rural areas, on the other hand, textile production was interspersed with agricultural labour, household chores and child-rearing. And while men normally were handloom weavers, women were said to have undertaken spinning and carding. As ‘a modern factory-based industry’ emerged, the structure and fate of the textile industry in the several Spanish peninsular regions varied greatly. In Galicia, linen-making derived from atomized production by small, family units, with scarcely any intervention by merchant manufacturers. In other areas, merchant manufacturers played a more important role in organizing production. These actions, Smith et al. note, only sometimes led to factory-based, mechanized production.45 At the turn of the twentieth century there were a smattering of large integrated weaving-spinning factories which employed between 500 and 2,000 workers, but the average-sized Catalan cotton factory had only about 87 workers (130 workers in spinning factories and 73 in weaving factories).46 In the early Qing period, imperial silk weaving factories in China increasingly contracted out work to private weavers. This trend was accompanied by a flourishing of private commercial silk weaving in the Jiangnan region. However, despite a similar commercialization and flourishing of cotton production and trade in the eighteenth century, we do not see the emergence of anything resembling a factory system in cotton textiles. This is surprising in light of the appearance of handicraft workshops and wage labour engaged in calendering and dyeing. Workshops processing cotton cloth increased steadily during the eighteenth and nineteenth centuries, and their workers proved sufficiently organized and conscious of their status to engage in strikes, as did silk weavers. However, there is no evidence of similar developments in cotton spinning or weaving.47 Cliver expresses surprise over the absence of workshop production of cotton cloth, since eighteenth-century

43 Ibid. 44 Bortz, ‘Mexico’. 45 Smith et al., ‘Spain’. 46 Ibid. 47 Cliver, ‘China’. 591

The Ashgate Companion to the History of Textile Workers cotton merchants were clearly able to amass large amounts of capital. And, he adds, technology does not appear to have been a significant obstacle. In eighteenth-century China, there emerged a kind of putting-out system but the integration of cultivation, spinning and weaving in the domestic production system usually made such putting-out arrangements unnecessary. In cottonproducing regions such as the Yangtze Delta, rural weavers sold their products to cloth merchants who then distributed them in distant markets. Sometimes weavers traded cloth for warp. But in general they grew the raw material, spun and wove their own cloth, marketing whatever surplus they produced through the cloth merchants. By contrast, in ‘regions that engaged in cotton textile production but did not cultivate cotton, the “putting-out” system was a natural development. This was, however, a commercial relationship rather than a wage contract.’48 In Qing China, household-based production reportedly remained the norm in cotton weaving but silk weaving occurred not only in rural households but also in urban workshops with wage workers. In explaining why household production of cotton textiles proved so resilient and long-lived, Cliver says that the most important reason for the durability of household cotton weaving is that farming families preferred to weave cloth to meet their own needs if this was at all possible. Even when markets declined in the nineteenth century, and many merchants left the cotton trade, farming families continued to spin and weave for their own consumption … In effect, households treated labour as a fixed factor of production rather than as a variable input. When combined with the fact that cotton production was necessary to meet families’ own needs, and the labour devoted to textile production enjoyed relatively low opportunity costs, the domestic production system proved highly resistant to efforts to develop factory-based cotton weaving well into the twentieth century.49 The China national overview, in common with several other national studies, points to distinctly different patterns of organization and technology among and between the various textiles. I should also note that, unlike Latin America and Mexico, but in common with the Ottoman Empire, local entrepreneurs played an important role in the Chinese textile industry.50 The putting-out system that developed in North China in the early twentieth century increased merchant control over production without creating a wage-based factory system. The minimal development of handicraft cotton weaving workshops during the 1870s–1930s serves to demonstrate the advantages of household-based textiles production in China. Based on wage labour and the supply of cheap 48 Ibid. 49 Ibid. 50 Cf. Cliver, ‘China’, Camou and Maubrigades, ‘Uruguay’, Bortz, ‘Mexico’ and Quataert, ‘Ottoman Empire’. 592

Proto-industrialization and industrialization machine-spun thread, these workshops struggled on the outskirts of the major cities, in the margins between rural handicrafts and urban factories. After 1949, the proportion of hand-woven cotton steadily declined and the domestic production system ‘that had heretofore exhibited such resilience’ all but disappeared within a decade. As earnings from independent handicrafts began to decline in the twentieth century, a great many weavers went to the cities to find work in urban silk factories. Thus, the paths to the Chinese industrialization of cotton and silk were very different, as were the histories of thread and cloth.51 In Japan during the seventeenth and eighteenth centuries, much textile processing was urban-based but shifted to rural areas thanks to urban guild restrictions that hindered the development of new technologies. In the nineteenth century, textile production expanded into areas with populations that were no longer very dependent on income from agriculture. Also, many farming families included one or more members who were engaged in textile production (including raw material production) on a full- or part-time basis. Generally, during the nineteenth century, systems of organization were very diverse and included handmanufactories as well as household production. Over time, however, the balance shifted towards larger scale capitalist enterprises. The growth of factory production came first in silk reeling but most silk reeling factories remained relatively small scale throughout the period.52 In India, weavers working within a household unit dominated the weaving industry during the pre-1850 period. In addition, in some towns (located in the northern Ganges plain) there were large workshops owned by rich nobles. In these worked one or more all-male master–apprentice teams. These two general types – households and male master–apprentice teams – persisted for a long time but overall tended to dissolve into various types of wage-labouring workshops in the late twentieth century. Before then, the usual form of industrial organization in handloom weaving was the household. Inside a weaving family, adult men reportedly worked usually as weavers, adult women on winding and sizing operations, and children as assistants in both weaving and winding. By the 1920s, however, a significant number of handloom factories had appeared in towns in western India. These factories, founded by rich weavers and merchants who had made money in the relatively new textile trades, employed migrant labour who used improved tools. In the late nineteenth-century Indian cotton mills women worked on the shop floor in tasks that earlier had engaged women in the handloom weaver household, namely winding, sizing and sometimes warping. These women almost always were the wives of workers in the same factory, working in teams with a female supervisor who mediated between the managers and the workers. In the case of India, the ‘transformation of the labour market has not necessarily involved a drastic move from the home to the mill shop floor. More often, it has involved a rather graduated shift from the home to the small handloom/power-loom workshop.’53 51 Cliver, ‘China’. 52 Hunter and Macnaughtan, ‘Japan’. 53 Roy, ‘India’. 593

The Ashgate Companion to the History of Textile Workers In Austria during the period 1650–1720, there were two organizational forms of textile production – household and guild. Komlosy argues that developments exactly follow the classic proto-industrial model. It was dominated by the puttingout system that mainly spread in rural areas, both relatively independent and subcontracting types. Proto-industrial domestic labour relied on a system of manufactories that was subsidized by the state. The textile sector combined centralized manufactories, concentrating on administration and marketing, preparation of raw materials and finishing, with decentralized activities, such as spinning, winding and weaving, carried out in the homesteads of mostly rural producers. Sometimes the centralized manufacture also included hand-weaving and, to a lesser extent, spinning. Eventually single factory villages or towns became so dense that contemporary observers were struck by the ‘continuous series of factories, just like in certain English regions’. But when mechanization set in, they had to increase their output in order to compensate for the lower wages that were being depressed by mechanization, first of spinning and then of weaving. Thus, many forms of rural manufacture vanished but home production moved into new sectors and expanded along with the factory system. The foundation of weaving mills in former proto-industrial regions effected a spread of the factory system and a gradual disappearance of traditional crafts, which fell victim to the overwhelming competition. Still, machine weaving did not put an end to homework.54 Lace-making became a notable home industry in which machines as well as finishing works were put out into the cottages of the surrounding village. In the case of the Netherlands, in Tilburg, a woollen weavers’ guild had been founded as late as 1767. Previously, cloth production was organized in various ways. There were independent drapers who bought their own materials and sold their cloth themselves and also merchants from outside Tilburg, who brought raw materials to the village for spinning and weaving on a commission basis. During the second half of the seventeenth century, many of the independent weavers could no longer compete with the production of wealthier entrepreneurs. Increasingly they wove for these ‘capitalist’ drapers and merchants in return for wages. The putting-out system thus came into being, making most Tilburg weavers dependent on wage labour.55

A few concluding remarks In the above pages, a major specific objective has been to call for the abandonment of terminology that is both misleading and judgmental and, more generally, to appeal for a rethinking of the trajectory and nature of textile production during the last several centuries. The term ‘proto-industry’ suffers from a double defect. In its 54 Komlosy, ‘Habsburg Monarchy’. 55 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 594

Proto-industrialization and industrialization original incarnation during the 1970s, it suggested stage theories of developments and, as scholars debated, the term became increasingly vague and general, while still retaining tinges of its earlier meaning. Alternative terms such as ‘puttingout’ are not terribly helpful and so we are left with ‘hand-manufacturing’ as a non-normative indicator of textile production without machines. Furthermore, based on the findings of a number of the national overviews and my own research, I have underscored the presence of non-factory production in the industrial and post-industrial era. Small-scale sweatshops and, sometimes, household production, I have stressed, are an essential aspect of twenty-first-century textile production, as they were in centuries past. This finding necessarily brings with it a sharp criticism of de-industrialization, a concept that has been blind to the significance of nonfactory production in the textile industry (among others). Collectively, the national overviews offer considerable evidence for the continued prevalence of both handand small-scale manufacturing in various areas of the globe during the era of mature capitalism and into the present day. And a number of them make clear that vital and vibrant textile sectors managed to survive. Certainly the notion of textile production in the non-European world inevitably collapsing during the era of the Industrial Revolution needs to be abandoned. The experiences recorded by the various national overviews, particularly that of China, also suggest the difficulty of predicting future production trends based on past performance in the textile sector. The particular variable mix leading to success in textile production is not readily apparent from the studies presented in this volume. In addition, this chapter has noted particular concentrations of female and male labour that various textile technologies have created in the factory workforce and calls for further research on this exciting and valuable subject. And, contrariwise, it cautions against too glibly assigning gender divisions of labour in the textile workforce at large. It notes that the various tasks historically have not been particularly male or female. Rather, the past shows us, gender switching of roles is commonplace.

595

This page has been left blank intentionally

24 The textile firm and the management of labour Arthur McIvor

Recently, the scope of labour history has widened to incorporate the study of employers, firms and management and there has been a sharper recognition of the need to critically analyse the nature of capitalism in order to understand social and labour relations. Research on the behaviour of firms has demonstrated a far more complex vision of the importance of different management policies and practices over time, divisions between different factions of capital, the often contradictory nature of employer strategy and the close interaction between capital and labour, resulting in a more nuanced and diverse history of workplace relations across time and space. However, while the historical studies of employers within particular country settings have increased, international comparative historical analysis of employer practice has been rare. Such approaches are important in developing theoretical conceptions of employer practice beyond the contexts of specific national economic settings. Two studies that stand out in this respect are Littler’s The Development of the Labour Process in Capitalist Societies and Tolliday and Zeitlin’s edited volume The Power to Manage? Both provide historical and comparative analysis of employer policy and practice in a range of different country settings. A key finding from these studies is that rather than a single means of capitalist control, employers have diverged significantly between countries in their labour 

I am particularly grateful to Patricia Barton, David Moon and Mary Blewett for helpful comments, information and references, and to all those at the conference in November 2004 who raised questions and made comments on the first draft of this chapter.  R. Chandavarkar, The Origins of Industrial Capitalism in India: Business Strategies and the Working Classes in Bombay, 1900–1940 (Cambridge, 1994); D. Nelson, Managers and Workers: Origins of the New Factory System in the US, 1880–1920 (Madison Wis., 1975); S. Jacoby, Employing Bureaucracy: Managers, Unions and the Transformation of Work in American Industry, 1900–1945 (New York, 1985); S. Jacoby (ed.), Masters to Managers: Historical and Comparative Perspectives on American Employers (New York, 1991); H. Gospel, Markets, Firms and the Management of Labour in Modern Britain (Cambridge, 1992); C. Wright, The Management of Labour: A History of Australian Employers (Melbourne, 1995); A. McIvor, Organised Capital (Cambridge, 1996).

The Ashgate Companion to the History of Textile Workers management policies as well as in the timing of the evolution of control from indirect, simple and more coercive means to more sophisticated, bureaucratic and consensual practices. Moreover, what these comparative studies reveal is a more complex interpretation of factors which have shaped such variation, with both economic contexts and differing institutional environments playing important roles in shaping employer strategies and behaviour. This chapter focuses on the textile firm, providing a comparative examination of the changing patterns of organization and behaviour among employers, managers and others who have controlled production in global textile manufacture since about 1700. Attention is concentrated on five issues. The first section explores labour control mechanisms within pre-industrial and proto-industrial modes of production, examining homework, artisan and guild textile manufacture, together with the role of merchants in the ‘putting-out’ system. The second section investigates the textile firm in the era of the modern, mechanized factory system and the evolution of more direct and frequently authoritarian work regimes, tempered in some cases by traditions of company paternalism. The third section explores the responses of textile firms to the challenge of trade unionism and organized workers’ protest movements, including the role played by employers’ organizations in industrial relations, supported, in some cases, by the state. The fourth section evaluates the key changes in managerial practice in the textile firm associated with scientific management and the bureaucratization of work in the twentieth century and the extent to which the textiles sector shared in this ‘managerial revolution’. The final section makes some brief comments about international and multinational textile firms.

The guild and merchant era: from home-working to ‘puttingout’ Throughout much of the world through most of the period under review the spinning and weaving of cloth took place within the home, as a primary activity or as a supplementary craft undertaken by family members whose main livelihood 

C. Littler, The Development of the Labour Process in Capitalist Societies: A Comparative Analysis of Work Organisation in Britain, the USA, and Japan (London, 1982); S. Tolliday and J. Zeitlin, (eds), The Power to Manage? Employers and Industrial Relations in ComparativeHistorical Perspective (London, 1991). For the best example in textiles, see M.B. Rose, Firms, Networks and Business Values: The British and American Cotton Industries since 1750 (Cambridge, 2000). Other examples include M.-G. Dezes, K. Lunn, A. McIvor and K. Tenfelde, ‘Employers and Trade Unions in the late Nineteenth Century in Britain, France and Germany’, in J.-L. Robert, A. Prost and C. Wrigley (eds), The Emergence of European Trade Unionism (Aldershot, 2004), pp. 204–214; A. McIvor and C. Wright, ‘Managing Labour: UK and Australian Employers in Comparative Perspective, 1900–50’, Labour History (Australia) 88 (2005): 45–62. 598

The textile firm and the management of labour came from working the land. Within this system, the obtaining of raw materials, the supervision and management of the labour process, the maintenance of the basic technology (such as the spinning jenny or the loom) and the distribution of the final product (where there was a surplus beyond the needs of the family) frequently lay in the hands of the workers themselves. A high level of workers’ control and autonomy thus existed over the pace and rhythm of work, epitomized, perhaps in the tradition within much of Western Europe of taking a rest day on ‘St Monday’. Within this mode of production we know very little about the exercise of authority and control, though it appears that social relations varied within the family unit, ranging from matriarchal dominance in some countries through to the more dominant West European and Islamic country norm of patriarchal control. In the latter, the sexual division of labour normally saw men responsible for strategic decision-making and undertaking the weaving processes and women and children in subordinate positions, doing the carding and spinning. In urban centres in the seventeenth and eighteenth centuries, textile production was frequently regulated by associations of artisans who, together with the merchants, exercised varying degrees of power within town councils. The guilds of wool, cotton and silk spinners and weavers typically operated to protect the interests of the textile artisans by restricting entry to the craft and fixing product prices. In some places guilds imposed strict apprenticeship regulations laying down training and probationary periods of up to 10 years and used their political influence to introduce licensing laws to forbid ‘interlopers’ from the trade, with fines and imprisonment acting as deterrents. In the urban centres, textile production thus increasingly took place in small workshops owned by a master craftsman or merchant who employed a handful of journeymen or ‘probationers’ and the requisite number of apprentices. Often the latter would have been kin. Again, we know little of the internal dynamics of the early-modern textile workshops. Apprentices though appear to have been subject to a very strict regulatory regime as they learnt the trade, but when fully-trained artisans they exercised a great deal of autonomy and independence over their work hours and pace of production within what appears to have been very loose managerial control structures. The extent of ‘responsible autonomy’ reflected medieval values such as the exaltation of skill, ‘community’ and distrust of competition. Vertical integration also characterized this mode of production in that the guilds were usually associations of both the craft masters and the fully trained artisans. While predominant in Europe, the guild system was widely exported to European colonies. Guild control began to erode in the eighteenth century in Britain, though in other areas of the world – such as the Middle East – the system survived intact well into the twentieth century. In Egypt, for example, urban textile production was dominated by artisans working in small workshops (typically employing 3–4 workers, often kin), controlled by the urban guilds (with Cairo dominating production) well into the twentieth century.

 For example see Bortz, Chapter 13, this volume, ‘Mexico’.  Beinin, Chapter 7, this volume, ‘Egypt’. 599

The Ashgate Companion to the History of Textile Workers Commonly merchants were also involved in textile production in the seventeenth and eighteenth centuries, developing what was known as the ‘puttingout’ (or contracting-out) system. There were various forms of such decentralized manufacture worldwide, though typically it involved a merchant purchasing raw material in bulk and arranging to distribute it to homeworkers and to workshops to be spun and woven, with the final product then being uplifted by the merchant for selling on. In Spain, such a system of atomized production was referred to as ‘industrialists without factories’. In the Netherlands, this involved merchants ‘putting-out’ raw material for manufacture to the family farms largely in the Twente district. Such merchants appear to have rarely interfered directly in the labour process itself and the workers typically retained a great deal of discretion and control over the rhythm of work within a system constrained mainly by the imperatives of completing the task. However, in some cases, merchants extended control through the provision of machinery (sometimes rented by the workers) and through the provision of credit which could lead to spiralling indebtedness, as in Russia and Mexico. In Turkey, the merchants came to directly control textile workshops in places like Damascus, Nablus and Mosul, imposing direct forms of supervision over the work. In areas of South and Central America and in Tsarist Russia such contracting out involved more coercive regimes, where forced or ‘bonded’ labour was employed.10 In Brazil, Portuguese colonists controlled textile production using black slaves in supplementary production on large sugar and tea plantations prior to the abolition of slavery there in 1888.11 Home-working, the guild system and merchant-controlled contracting systems prevailed in many countries well into the twentieth century. As Beinin has commented, where labour and food were cheap there remained little incentive to mechanize production or move to more sophisticated managerial techniques.12 In India in the 1930s, for example, the textile factories employed around half a million while there were several million handloom weavers.13 Similarly, in China household-based production and ‘putting-out’ predominated well into the twentieth century – indeed effectively until the communist era, when rapid, state-initiated industrialization over two decades replaced handicraft methods with modern, factory-based production of textiles.14 For the merchants, decentralised contracting systems had distinct advantages, involving relatively little capital investment in fixed plant and machinery, maximum flexibility, low levels of risk and the option to allow workers to manage themselves.  Smith et al., Chapter 17, this volume, ‘Spain’.  Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’.  Pretty, Chapter 16, this volume, ‘Russia’; Bortz, ‘Mexico’.  Quataert, Chapter 18, this volume, ‘Ottoman Empire’. 10 Pretty, ‘Russia’; Bortz, ‘Mexico’. 11 Delson, Chapter 4, this volume, ‘Brazil’. 12 Beinin, ‘Egypt’. 13 Roy, Chapter 10, this volume, ‘India’. 14 Cliver, Chapter 5, this volume, ‘China’. 600

The textile firm and the management of labour

Mechanized textile firms and the real subordination of labour? Textile manufacture was the first sector to make the transition to mechanized factory production, initially in Great Britain, and this mode of production involved quite different social relations and structures of authority and control. Undoubtedly, the development of the capitalist labour process, as Marx hypothesized, facilitated greater degrees of subordination over labour as machines de-skilled work, factory owners and overseers imposed draconian discipline over their workers (what Marx termed ‘despotic’ control) and, to varying degrees, governments supported or colluded in such intensified exploitation. However, in reality this was a complex and uneven process and there is now much debate over the validity of Marxist labour process theory. Disagreements exist over the timing and extent of transformations in authority and control at work at a number of levels, not least among those scholars influenced by Marxist ideas.15 Recent work suggests that it would be wrong to assume that the factory system necessarily involved a change from indirect to direct forms of supervision and control.16 Moreover, the process of industrialization is now recognized to be much more uneven than was once imagined – hence ‘traditional’ modes of work organization and control persisted and frequently proliferated in tandem with an emerging modernized sector. It follows that employers were not as omnipotent, united or as empowered as the orthodox Marxist model suggests, nor workers as powerless. Comparative analyses across countries (such as this project) throw up evidence of common patterns of employer strategies and managerial practices but also highlight a rich mosaic of experience at any given point in time across global textile production within different cultural and institutional contexts. Nonetheless, with higher levels of capital investment in machinery and in the provision of work space (workshop or factory), textile firms increasingly came to expect greater degrees of control over labour. Sharpened notions of employers’ right (or ‘prerogative’) to manage emerged with the factory system and this was encapsulated in more coercive modes of supervision through overseers, foremen and plant managers. The rather loose and informal monitoring of the labour process common in homework and the workshop gave way to tighter discipline and control – as E.P. Thompson famously hypothesized from the English case, irregular ‘task’-based work gave way to regularized time discipline, with the factory clock (or bell) determining the work day.17 Moreover, what was new was that the speed of the power-driven machines determined to a large extent (though not fully) the pace of work. While technical control was taken to a more sophisticated level by Henry Ford, it was pioneered in cotton textile factories. The new textile factory 15 Most famously, perhaps, Braverman’s disagreement with Marx over the periodization of ‘real subordination’ of labour. See Harry Braverman, Labor and Monopoly Capital (New York, 1974). 16 Littler, Development of the Labour Process, pp. 64–69. 17 E.P. Thompson, The Making of the English Working Class (London, 1963). For a comment on these developments, see Fowler, Chapter 9, this volume, ‘Great Britain’. 601

The Ashgate Companion to the History of Textile Workers regimes were thus blatantly more exploitative (for example of child labour), especially in the early unregulated phase of textile factory development (in Great Britain through to the 1830s, and elsewhere frequently until late into the nineteenth century and even beyond). Thus, traumatic injury rates were extremely high in the early phases of mechanized factory production, as were long-term chronic industrial diseases associated with textile production, such as respiratory disability (especially in the dustiest preparatory processes in the card and blowing rooms). The maximization of profits within such factory regimes invariably damaged workers’ bodies to a lesser or greater degree.18 Coercive modes of control dominated this early phase of factory-based textile production. However, the authority of the firm could be and frequently was challenged, rather than passively accepted by the workers. The USA provides a good example of the shifting frontier of control in factory textile production in the nineteenth century, the contested nature of authority in the mills and, ultimately, the capacity of owners and managers to impose their will upon mill workers. Blewett emphasizes the way in which US mill owners successfully exploited divisions within the labour force based on ethnicity, religion, race and gender in both the traditional sector in New England and the emerging sector in the southern states.19 In response to trade union organization, strikes and protest, mill owners countered with draconian tactics, including widespread use of the lockout, blacklisting, victimization and the use of anti-trade union contracts. US mill owners in the main refused to bargain collectively with unions and insisted on their unilateral right to manage their mills without ‘interference’, as they thought fit. Consequently, they rejected any open scrutiny of the order books as a breach of their sacrosanct prerogative to manage their businesses. Yarn and cloth were also stockpiled as a safeguard against strikes, while the growth of state intervention (for example in Massachusetts with the 1874 Factory Code) in the late nineteenth century was widely ignored. In nineteenth-century US textile mills, an authoritarian employer regime of fear and intimidation appears to have prevailed across most of the sector, though there did exist a significant paternalist strand of family-owned companies in Philadelphia engaged in higher quality, specialized batch production where more consensual social relations largely prevailed.20 The latter perhaps illustrates the point pressed strongly by Gospel (after Chandler) that product markets, labour markets and company structures significantly influenced employers’ labour relations policies.21 18 For a discussion of the impact of cotton textile firms’ policy on occupational health standards in British mills, see A. McIvor, ‘State Intervention and Work Intensification: The Politics of Occupational Health and Safety in the British Cotton Industry, c.1880– 1914’, in Ad Knotter, Bert Altena and Dirk Damsca (eds), Labour, Social Policy and the Welfare State (Amsterdam, 1997), pp. 125–142; A. Fowler, Lancashire Cotton Operatives and Work, 1900–1950 (Aldershot, 2003). 19 Blewett, Chapter 21, this volume, ‘USA’. 20 Ibid. 21 Gospel, Markets, Firms and the Management of Labour. 602

The textile firm and the management of labour In Great Britain, and other Western European countries, this struggle for control at work was less one-sided, the dominance of textile firms was never ensured and was ultimately subject to mediation due to sustained worker organization and protest. Here social control mechanisms developed by the firm were invariably more flexible and more diverse. In some regions, the discipline of the firm remained relatively loose. In Barcelona, artisanal autonomy prevailed in the late nineteenth century, with male spinners controlling the technology, curtailing the employment of female labour and winning concessions including the right to smoke and drink on the job and to start work after five minutes ‘leeway’.22 Paternalist modes of control also persisted in many textile firms across Europe where employers strove to dominate social relations within the wider community as well as within the mill, attempting to cement workers’ sense of loyalty and attachment to the firm through the provision of a wide range of benefits beyond the wage. The company welfare ‘package’ could include the provision of accommodation, schools, shops, community halls, parks and playing fields, as well as company pensions and other forms of social insurance, such as sick pay and ad hoc accident compensation. Joyce has shown such factory regimes to be prevalent in cotton and wool manufacture in northern England in the Victorian period, and such company paternalism persisted longer in the Coats mills in Paisley, Scotland.23 Again, the influence of product markets, labour markets and company structures are evident in the determination of firms’ choices. Textile company paternalism appears to have developed deepest roots in more isolated rural settings (where a labour force had to be attracted and retained), where family-ownership prevailed (as in Bolton, England and Paisley, Scotland) and where the firm produced finer quality or specialized products, monopolizing markets (such as thread manufacture by Coats) or at least where competition was less fierce. At one extreme, this could produce a much more humanitarian factory regime, as with the New Lanark mills in Scotland during the managerial era of the communitarian Robert Owen and at the Amoskeag mills in New Hampshire, USA in the nineteenth century.24 At the other extreme, paternalist control could be overbearingly despotic, as with the textile mill dynasties in the Twente region of the Netherlands, epitomized perhaps by Jacob Spanjaard in the early twentieth century, dubbed ‘the god of Borne’,25 and in early twentieth century rural mills in Catalonia, where such control could extend even to the banning of left-wing literature.26 Nor was such company paternalism solely a Western phenomenon. There was a strong tradition of welfarism in Japan where the provision of housing

22 Smith et al., ‘Spain’. 23 Patrick Joyce, Work, Society and Politics (London, 1980); W. Knox, Hanging by a Thread: The Scottish Cotton Industry, c1850–1914 (Preston, 1995). 24 T. Hareven, and R Langenbach, Amoskeag: Life and Work in an American Factory-City (New York, 1978). 25 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 26 Smith et al., ‘Spain’. 603

The Ashgate Companion to the History of Textile Workers and education were well-established strategies used by late nineteenth-century textile firms for managing a largely female labour force.27 In Asia and South America, in different cultural settings and especially where workers rights were limited and traditions of democratic government weak or non-existent, textile factory regimes could be notably more draconian, fitting more closely, perhaps, the Marxian model of ‘despotic’ control. In some countries, state ownership of textile firms in the nineteenth century facilitated autocratic modes of control. Egypt provides an example where textile workers organized in guilds retained a considerable degree of control over the labour process in the state-owned factories in the early/mid-nineteenth century, but the armed forces were used to ensure that forced labour remained in the mills.28 Textile firms in colonial regimes could use forced and slave labour, as in Brazil where the labour force consisted of a mix of slaves and free wage labour toiling for up to 17 hours a day in the 1880s, frequently under the control of English managers.29 In Mexico before independence (1811) Indian and ‘unfree’ labour worked in urban obrajes (larger workshops) under the repressive authority of Spanish businessmen and supervisors (one such firm was described as ‘a dark prison’).30 Autocratic control and domination continued to characterize the mills in Mexico City through the nineteenth century, with long 15-hour work days imposed upon ‘Mexican and dark-skinned’ workers by company owners who were mostly ‘foreign and white’.31 Elsewhere, as in Shanghai before the Second World War, textile firms practically enslaved a largely young, female workforce recruited from the countryside, housing them in squalid dormitories and imposing military-style discipline upon them. Bribery and corruption permeated China’s main textile centre in Shanghai, with textile firms acquiescing in the control that organized crime (the Green Gang) exercised over labour recruitment (‘buying’ single women on contracts of 1–3 years).32 Late nineteenth-century Russia provides, perhaps, an archetypal example of coercive control within textile production. Mechanized cotton manufacture developed rapidly from the early 1840s, in part through the sponsorship of the Tsarist state, and became very concentrated geographically with a relatively small number of very large plants employing over 1,000 workers. As Dave Pretty has shown, draconian factory regimes characterized by over-bearing supervision, complex fining systems for indiscipline and the widespread employment of child labour prevailed.33 The autocratic authority and control structures of the firm in Russia were also bolstered by a repressive state, where trade union organization and strikes remained illegal until 1905. Transgressors could find themselves exiled to Siberia. Textile firms maximized profits in this period by increasing the monitoring 27 Hunter and Macnaughtan, Chapter12, this volume, ‘Japan’. 28 Beinin, ‘Egypt’. 29 Delson, ‘Brazil’. 30 Bortz, ‘Mexico’. 31 Ibid. 32 Cliver, ‘China’. 33 Pretty, ‘Russia’. 604

The textile firm and the management of labour of workers, ignoring the weak and ineffectual factory legislation that was introduced (for example in Factory Acts in 1885 and 1897), extending payments by results wage payment systems, rate-cutting and increasingly replacing more expensive male textile workers with women. Tight control over labour costs was deemed imperative by the autocratic family dynasties in Russia who ran these giant textile corporations, necessitated by their dependency upon less efficient technology and low profit margins. The extent to which textile workers were central to the strike waves and revolutionary ferment in Russia from 1905 to 1917 owes much, as Pretty has demonstrated, to such a volatile, provocative and alienating combination of workplace and civil repression.34 All this is not to suggest, however, that factory regimes were necessarily worse in the largest firms. In several countries evidence indicates that it was the smaller textile firms, trying to compete on the margin, where conditions were worse, local and state labour codes were subverted and the work impacted most critically upon the bodies of employees.35 Some of the larger firms were also notably paternalistic, even in countries where repressive modes of control dominated, such as Brazil.36 However, relatively speaking there only appears to have been a very thin strand of genuinely welfarist textile firms scattered across the industry globally in the eighteenth and nineteenth centuries. Where corporate paternalism was most developed, perhaps, was in Japan where textile firms pioneered a distinctively welfarist labour management strategy – with state support – as an alternative to class confrontation in Western capitalism. The example set by textile firms congealed into a more widespread paternalist managerial philosophy in Japan in the twentieth century.37

Rising to the challenge of the trade unions: organized capital in textiles Textile firms across the globe showed varying tendencies to organize together, often playing important roles within national employers’ movements in the nineteenth and twentieth centuries. The cementing pressures that drew mill owners away from their traditional ‘rugged individualism’ towards a sharpened sense of class consciousness exhibited in collective organization and coordinated action also differed markedly in different contexts. However, there appear to be three primary interlocutors: the rise of trade union organization and worker militancy, 34 Ibid. 35 Lobato, Chapter 2, this volume, ‘Argentina’; for the USA, Nelson, Managers and Workers; for Britain, Joyce, Work, Society and Politics. 36 Delson, ‘Brazil’. 37 Littler, Development of the Labour Process, p. 155; A. Gordon, The Evolution of Labor Relations in Japan Heavy Industry, 1853–1955 (Cambridge, Mass., 1985), pp. 414, 421–425. 605

The Ashgate Companion to the History of Textile Workers government intervention in matters which impinged upon profitability in textile manufacture and, finally, changes in market conditions – usually associated with perceived or actual loss of market share. Where much debate continues to rage, however, relates to just how powerful and united employers were. One view, based largely on Marx, defines employers as omnipotent and united in ruthless pursuit of their class interests. An alternative ‘revisionist’ interpretation stresses the relative weakness of employers in the face of organized labour and their inability to combine together to impose their will. Here centripetal as well as centrifugal tendencies are recognized. Like other employers, textile firms were divided by their competitive relationship with one another, by differences in company structures and styles of management, and by diverging product markets in different fibres, yarn and cloth qualities.38 Nonetheless, what appears evident from this comparative survey of textile firm behaviour across a number of countries is just how prevalent and effective collective activity by firms could be in neutralizing the challenge from organized labour. Some of the earliest examples of collective organization among textile firms can be found in Great Britain. In the 1740s and 1750s cotton textile merchants and manufacturers in the Manchester region combined together, firstly to force a reduction in wages and subsequently to deal with a strike of check weavers.39 Over the following century, as industrialization developed, a tapestry of local town-centred employers’ associations emerged in northern England and central Scotland. The Glasgow Master Spinners’ Association and the Oldham Cotton Masters’ Association – in permanent existence from the 1830s – provide archetypal examples. Prior to the middle of the nineteenth century the labour relations strategies of such nascent textile firm associations were blatantly coercive. Firms came together to provide a united front to preserve managerial prerogatives where companies felt threatened by the ‘encroachments’ of worker militancy and trade union formation. In essence this was a struggle over the distribution of power and in this early period the textile employers’ organizations mustered an impressive array of weapons against the unions. These included use of the lockout to widen the area of struggle, neutralize the unions’ use of the selective strike tactic and literally starve workers into submission. This would typically be supplemented with attempts to replace recalcitrant workforces with non-unionist, ‘blackleg’ labour and various methods of victimizing strikers and labour activists, ranging from enquiry and discharge notes to character referencing systems and formal blacklisting to prevent troublemakers getting work elsewhere. Moreover, such strike-breaking methods were usually sanctioned by the state and the law courts, creating an environment 38 For a survey of interpretations and analysis of the emergence and development of employers’ organizations in cotton textiles in north-west England, see A. McIvor, Organised Capital, pp. 11–56. For India, see also R. Chandavarkar, Origins of Industrial Capitalism in India, pp. 333, 393–396. For comparative studies of employers’ organizations, see Tolliday and Zeitlin, The Power to Manage? and J.P. Windmuller and A. Gladstone, Employers Associations and Industrial Relations: A Comparative Study (Oxford, 1984). 39 McIvor, Organised Capital, pp. 38–39. 606

The textile firm and the management of labour of fear and intimidation which could castrate early attempts to unionize textile workforces. By the end of the nineteenth century cotton textile firms in Great Britain were very well organized with both the associations of master spinners and master weavers recruiting over half of the machine capacity in the industry. While divisions between mill owners remained significant, collective organization empowered textile firms, with the early phases of the employers’ movement often characterized by inflexible and confrontational labour relations strategies. In Mexico, for example, when workers went on strike to reform draconian work regimes in 1906, textile firms organized together to form an association and to lock-out the entire industry in a particularly bloody confrontation. The companies had the support of the Diaz government which imposed a settlement favourable to the bosses and ordered troops to disperse strikers in Orizaba, leading to the killing of ‘scores’ of strikers. This succeeded in maintaining unilateral managerial authority in the Mexican textile mills, at least until Diaz was opposed in the Revolution of 1911.40 In the Netherlands, the late 1880s saw the emergence of particularly aggressive textile manufacturers’ associations in the Twente region which used the lockout weapon effectively to undermine the textile trade unions in the two decades or so before the First World War.41 In the early years of the interwar depression the Netherlands Textile Employers’ Federation went on to spearhead a costs’ reduction drive which included a 10 per cent wage cut and a rise in work hours from a 48- to a 53-hour working week. In Spain, Catalan industrialists combined in the last quarter of the nineteenth century to counter a strong and militant tradition of craft organization in textiles. Among their tactics were heavy use of the lockout weapon and exploitation of their political muscle in a period of Conservative government in Catalonia. The outcome was a phase of work intensification, deteriorating wages and work conditions for textile workers in Barcelona from the 1870s and a resurgence of strict labour discipline in the mills.42 Similarly in Japan the collective organization of the largest textile companies (in the Japan Spinners’ Association) significantly bolstered the capability of textile firms to keep unions out of the workplace and retain company-level bargaining and largely authoritarian paternalistic work regimes up to the 1940s.43 In Egypt, the repression of trade unions lasted even longer with employers organizing together to impose crippling lockouts to destroy any attempt by workers to organize. As Beinin has shown, the 1938 textile lockout in Egypt led to the arrest of around 100 strike leaders and the imprisonment of 50, and was followed by the reorganization of the industry into smaller workshops to neuter trade union power.44 The persistence of unilateral managerial control in textile firms in Egypt is indicated in contemporary labour relations; where company unions prevail, some workers are working on only three-month recruitment contracts and others are required to sign undated letters 40 Bortz, ‘Mexico’. 41 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 42 Smith et al., ‘Spain’. 43 Hunter and Macnaughtan, ‘Japan’. 44 Beinin, ‘Egypt’. 607

The Ashgate Companion to the History of Textile Workers of resignation when they are hired to facilitate instant dismissal.45 As Beinin notes, these authoritarian responses typify a country with an abundant labour supply and a textile industry attempting to compete using older technologies and work methods against more developed and technically superior competitors. To an extent the same was true of late Tsarist Russia, though here the responses of textile firms to the growth of trade unions and strikes from 1870 to the 1910s was particularly atomized, with little evidence of employer collusion or formal association.46 Such coercive modes of control by textile firms and their collective organizations gave way over the course of time in many countries – and especially in the developed economies of Western Europe – to increasing levels of collaboration with the trade unions. Experimentation with ‘corporatist’ or procedural forms of control superseded coercive methods in the West. The textile employers’ associations frequently played a key role in this institutionalization of industrial relations. In part this was a rational response to the growing capacity of textile workers to get organized, as well as their increasing propensity to successfully exploit the strike weapon. In some countries, such as Mexico, this transition was precipitated by political change and characterized by much violence. The Mexican Revolution of 1911 marked a watershed in textile firms’ autocratic factory regimes with the textile unions exploiting the opportunity to challenge managerial authority and win reformed labour codes, including industry-wide collective bargaining, enshrined in the Labour Contract of 1927.47 Symbolic of this erosion in the authority of Mexican textile firms was their failure to reverse this extension of workers’ control in this period in the law courts. In countries such as Great Britain a symbiotic relationship between the unions and the employers’ associations developed from the late nineteenth century and the transmogrification of labour relations towards ‘corporatist’ accommodation occurred more peacefully. Each side fed off the other with the trade unions and employers’ movements growing largely in tandem. Recent research has shown the extent to which workers were an agency here (rather than powerless victims) and how flexible and adaptive firms were in reacting to changed circumstances (such as times of full employment and periods of war such as 1914–1918), developing more sophisticated control mechanisms or simply making the best out of prevailing power relations. For some, this involved attempting to incorporate unions into bureaucratic systems and use unions to police their own members. This was most evident, perhaps, in the emergence of national wage bargaining agreements and stage-by-stage disputes procedures whereby no strikes or lockouts were permitted until formal discussions had been exhausted. However, the pace of change differed across and even within countries. In Great Britain, for example, Fowler has shown how the generally larger firms in the wool and worsted sector in Yorkshire widely employed female labour and succeeded in largely keeping trade unions out of the workplace before the First World War, in marked contrast to the experience 45 Ibid. 46 Pretty, ‘Russia’. 47 Bortz, ‘Mexico’. 608

The textile firm and the management of labour in cotton textile firms in Lancashire.48 Scotland followed a similar pattern so that in both Yorkshire and Scotland levels of collective organization among firms were weak and collective bargaining poorly developed in contrast to Lancashire.49 For textile firms, these ‘corporatist’ methods acted as an alternative labour control strategy and could be extremely effective in the twentieth century in containing the challenge of labour. The Netherlands, Denmark and Britain provide examples.50 Sometimes, as in Britain, Denmark and Japan, these institutionalized arrangements for dealing with industrial relations followed major confrontations between capital and labour in textiles. In Denmark, formal industry-wide collective bargaining came in 1898 after a nationwide industry lockout for three months in which the Copenhagen textile association amalgamated with provincial manufacturers. The so-called ‘September Compromise’ which emerged saw the employers recognizing the unions’ right to bargain collectively, while the unions accepted the employers’ sacrosanct prerogative to manage their own workplaces without interference.51 In Japan, annual collective bargaining in textiles spread following an infamous, bitter strike in 1954 at the Omi Kenshi works – a notoriously autocratic employer.52 As workers became better organized across the global textile industry in the twentieth century, textile firms and their associations were invariably forced to give ground and concede degrees of worker participation in the determination of wages and conditions of employment. At the extreme, this could significantly curtail textile firms’ ability to introduce new technology or changes in work organization, such as increasing the ratio of looms operated per worker.53 However, outcomes of such struggles over the organization of work differed markedly across different countries and even in different sections of textile manufacture within countries.

‘Scientific’ management and the bureaucratization of work in textile firms The extent of delegation of labour control to employers’ associations differed significantly across the global textile industry. In many countries firms remained 48 Fowler, ‘Great Britain’. 49 See T. Jowitt, ‘The Retardation of Trade Unionism in the Yorkshire Worsted Textile Industry’, and I. Magrath, ‘Protecting the Interests of the Trade: Wool Textile Employers’ Organisations in the 1920s’, in J. A. Jowitt and A.J. McIvor (eds), Employers and Labour in the English Textile Industries, 1850–1939 (London, 1988), pp. 44–63 and 84–106, respectively. 50 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’; Christensen, Chapter 6, this volume, ‘Denmark’; McIvor, Organised Capital, pp. 118– 145. 51 Christensen, ‘Denmark’. 52 Hunter and Macnaughtan, ‘Japan’. 53 Christensen, ‘Denmark’. 609

The Ashgate Companion to the History of Textile Workers the primary locus of employer power and traditions of collective activity were transitory and at best apparent only in periods of crisis. Russia would be an example. Elsewhere, while employers’ associations dealt with labour contract issues, such as wages, work hours and conditions, firms retained much autonomy over the organization of production, including the introduction of new technology and management systems, such as Taylorism (for example the USA and Japan). In some cases in the twentieth century the initiative in labour management and personnel policy was shifted back from the collective organizations to the individual textile firms. This was evident in the UK and other Western European countries, with the declining membership and eroding importance of the textile employers’ associations from the 1930s on.54 These developments accelerated in the second half of the twentieth century and were also associated with changes in company structures in an era of textile multinationals, globalization and in the pervading bureaucratization of work in textiles worldwide. In short, textile firms were increasingly moving from indirect forms of labour management, including puttingout, internal subcontracting and external delegation to employers’ organizations, towards more direct forms of recruiting, controlling and managing their labour forces within the firm. These tendencies in the behaviour of firms were evident earliest, perhaps, in the USA, the birthplace of the architect of ‘modern’ scientific management, Frederick Winslow Taylor. Taylorism essentially challenged employers to directly manage their labour, to discover how the work was done through time study (using the stop watch), then to reorganize production to benefit from maximum division of labour. In the process, craft trade unionism was undermined as work was progressively de-skilled and downgraded. In the USA and in Western Europe one of the main mechanisms used by textile firms to implement Taylorism after the First World War was the Bedaux system. In his study of the bureaucratization of work Littler has shown how a significant number of textile firms brought in the Bedaux management consultants as a precursor to radical reorganization of the labour process.55 It is likely, however, that many more firms experimented with elements of scientific management while eschewing the whole package and sometimes the expense of bringing in outside agencies such as Bedaux. The extent, timing and pace of this internal management revolution differed significantly, however, across the textile industry globally. Textile firms operating in overstocked labour markets (such as in Egypt and India) or where forced labour persisted (such as in China) had little incentive to introduce new technologies or such costly (and sometimes provocative) reorganizations of work. Moreover, as Chandarvarkar has argued, divisions within the employers’ ranks in Bombay (the premier cotton textile city in India) prevented a common front and undermined the rationalization of work drive in the interwar years.56 Such divisions within textile capitalism undoubtedly constrained managerial work reorganization 54 McIvor, Organised Capital, pp. 208–210. 55 Littler, Development of the Labour Process, pp. 114–115. 56 Chandavarkar, Origins of Industrial Capitalism, pp. 394–395. 610

The textile firm and the management of labour elsewhere. Conversely, scientific management spread rapidly in the USA and areas of US economic influence after the Second World War (including Germany and Japan). Elsewhere, managerial complacency and poor systems of management education retarded the spread of new ideas on personnel management, including in the UK textile manufacturing sector. It was not until well after the Second World War that work study was taken up in earnest in British textile firms, a factor which contributed in no small measure to the sharp contraction of the industry from the 1920s. Moreover, the managerial revolution associated with Taylorism did not occur within textile firms in a vacuum in the twentieth century. The capacity of textile firms to introduce and sustain such sophisticated systems of labour control depended upon a range of variables, including prevailing enterprise culture, the structure of the firm, the nature of product and labour markets, the attitudes of labour and unions, and other contingent circumstances, such as the attitude of state agencies and war. Importantly, some firms chose not to initiate such changes, preferring systems of responsible worker autonomy or negotiated and agreed evolutions in labour processes. In the Netherlands it was the Second World War and post-war labour shortages that provided the key stimulus to the bureaucratization of work in textile firms, but this was initiated within a context of worker participation and joint consultation, with the pill sweetened by a significant extension in company welfarism.57 The process also occurred more rapidly in the Netherlands in the cotton sector (based in Twente) than in wool manufacturing (centred around Tilburg). In Japan, scientific management was grafted on to a pre-existing and deep-rooted commitment to corporate paternalism, with the latter largely surviving intact in Japan in contrast to the USA, where welfare capitalism atrophied from the 1920s and Taylorism largely triumphed. This success of American capitalism was the product not just of distinctive proprietorial and anti-union cultural values, but also of the economic and political resources that US firms could muster to neutralize trade unionism and facilitate a thorough-going transformation of work organization and control.58 The sheer size of US textile companies represented an enormous conglomeration of power in itself: the average unit size of US textile firms by the 1970s was almost 10 times the West European average.59 By 1980, only around 7 per cent of US textile workers were unionized.60 Where trade unions were powerful and militant, however, the imposition of Taylorism, through Bedaux or other guises, could be fiercely contested, either officially, or through militant rankand-file protest. The outcomes of such skirmishes across the fluid and ever-changing managerial frontier of control differed widely. In South America, the Bedaux system appears to have spread into textile firms most extensively. In Argentina, the Textile 57 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 58 Jacoby, Masters to Managers, p. 199. 59 F. Clairmonte and J. Cavanagh, The World in their Web: Dynamics of Textile Multinationals (London, 1981), p. 168. 60 Ibid., p. 175. 611

The Ashgate Companion to the History of Textile Workers Employers’ Confederation spearheaded the drive to rationalize work. This led to a massive confrontation with the unions in 1959–1960 over the introduction of time and motion studies, culminating in the wholesale sacking of strike committees, worker takeovers of plants and the drafting in of the police to break up strikes. The resulting 1961 Productivity Agreement sealed the employers’ victory and opened the way to unrestricted exploitation of Taylorite work systems in Argentina. These developments were endorsed after the right-wing military coup in 1976 which saw militant trade unions outlawed and employers’ power consolidated.61 A similar if somewhat less bloody process unfolded in Uruguay where the major textile companies in wool and cotton introduced Taylorism after 1918, resulting in fundamental changes in work organization. Among the changes were a marked rise in the degree of monitoring, with the ratio of supervisors and planners to production line workers increasing rapidly. The collective agreement struck in 1988 registered the power of textile employers in Uruguay to manage their firms as they thought fit, without any interference from the unions, especially in the sphere of work organization and production.62 However, in a number of cases employer innovations and ‘scientific’ managerial systems were moderated by organized labour, resulting in a more humanized production regime. This happened, for example, in countries such as Denmark and the UK in relation to the ‘more looms’ issue in the period 1920–1950. However, Denmark also provides a good example of how collective bargaining could undermine labour resistance to the rationalization of work. In a renowned ‘unofficial’ strike at the Silkeborg mill in 1934 the Textile Union disowned strikers, refused benefits and ultimately expelled all strikers from membership. While negotiations ensued, the breakdown in discussions left management open to introduce the Bedaux system on their terms, with the union counselling workers to obediently follow the orders of management.63 Therefore, it would be wrong to interpret the spread of scientific management and ‘rationalization’ as entirely unproblematic, inevitable or, indeed, as the most advanced weapon in the armoury of modern-day competitive capitalism. Moreover, it was not only the developed capitalist economies of the West that utilized such methods, but also some of the underdeveloped capitalist economies and centrally planned economies. Of the latter, the USSR stands out as a major protagonist for Taylorism, counting among its strongest supporters none other than Lenin in the early 1920s. Furthermore, in the West, including in Britain and the USA, textile firms operating in declining markets were as likely to pursue traditional modes of intensifying work as to embark fully on a systematic programme of work study and reorganization. In other words, continuities in labour management strategies are evident as well as mutations over time. The Boott Cotton Mills in Lowell, Massachusetts provides an example of such a company which commissioned but never implemented a Taylorite efficiency survey in the 1900s, continuing thereafter 61 Lobato, ‘Argentina’. 62 Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’. 63 Christensen, ‘Denmark’. 612

The textile firm and the management of labour to squeeze labour costs using traditional methods. The outcome was a deterioration in work conditions and wages which may have exceeded anything that scientific management would have achieved.64

The textile multinationals The multinational firm represents, perhaps, both the most recent transformation in ownership structure in global textiles and the most powerful expression of capitalist power over labour to date. In the second half of the twentieth century the trend in textile firms in the developed capitalist economies was towards horizontal and vertical integration, with mergers, takeovers and acquisitions occurring at an astonishing and unprecedented rate. Moreover, the biggest textile conglomerates opened up subsidiary factories throughout the world, and especially in the underdeveloped capitalist economies. Initially, this expansion beyond national boundaries occurred in regional spheres of influence – for the USA into South America, for Japan into the Far East underdeveloped economies (such as Indonesia) and for West European countries into their colonies and ex-colonies. Clairmonte and Cavanagh estimated that in 1980 ‘control of textile processing is in the hands of a loose oligopoly, with approximately 35 to 40 large textile corporations exerting a paramount force on world markets’.65 In Europe, Courtaulds and Coats became the two largest textile manufacturers by 1980 and both invested heavily outside of their base in the UK. Coats employed 66,000 with interests in 30 countries (including India, Pakistan, Hong Kong, the Philippines and Turkey), while Courtaulds employed 153,000, with factories in China, Thailand, Sri Lanka, Turkey, Tunisia, Morocco and the Philippines. By the end of the century, about two-thirds of the turnover of these companies was generated outside of the UK.66 Similar developments occurred in the USA, across Western Europe and in Japan, representing a massive, seismic decline in investment and in employment in textiles in the developed capitalist economies. The calculated economic rationality of such giant firms is expressed by this statement by the Netherlands textile multinational Gamma Holding NV in 2004: ‘Activities that show a structural lack of profitability or no longer fit in with the strategy are divested. Risk is limited by spreading the activities over various regions and market segments.’67 Concurrent with these developments was a major shift by textile capitalists from natural fibres into synthetic fibres, starting with nylon and rayon, and this sector became the most concentrated and transnational. In 1950, man-made fibres 64 See L. Gross, The Course of Industrial Decline: The Boott Cotton Mills of Lowell, Massachusetts, 1835–1955 (Baltimore, 1993). I am grateful to Mary Blewett for pointing out this reference. 65 Clairmonte and Cavanagh, World in their Web, p. 165. 66 www.knittingtogether.org.uk 67 www.eumanufactur.com/textile_multinationals 613

The Ashgate Companion to the History of Textile Workers constituted just 20 per cent of total world textile output; by the 1980s, they had surpassed 50 per cent.68 Table 24.1 World’s leading producers of synthetic fibres, 1979 Percentage share of world fibre capacity

Percentage share produced overseas

Firm

Country of ownership

DuPont

USA

14

21

Akzo

Netherlands

6

85

Celanese

USA

6

27

Monsanto

USA

6

17

Toray

Japan

4

30

Taijin

Japan

4

43

Hoeschst

Germany

3

43

Rhone-Poulenc

France

3

53

Courtaulds

UK

2

24

ICI

UK

2

76

Source: V. Cable and B. Baker, World Textile Trade and Production Trends (London 1983), Table 35, cited in P. Dricken, Global Shift, Industrial Change in a Turbulent World (London, 1986), p. 243.

The textile multinationals reaped massive benefits from the shifting of manufacturing to underdeveloped countries. In many cases they enjoyed political privileges, subsidies and tax breaks in adopted countries and invariably fewer ‘social’ overheads, such as not having to contribute to company pension schemes or insure against accident risks. Double standards were common in home-based transnational plants compared to subsidiaries, not least on occupational health and environmental controls. Among the attractions were cheaper transport costs and, crucially, the potential to exploit abundant, cheap and poorly organized labour, including child labour, in textile factories, but also through subcontracting and ‘out-sourcing’. The multinationals could bypass the more extensive labour codes of the developed capitalist economies and the well-established and powerful trade unions. In turn, these companies forced alien management regimes, including Taylorism, upon labour forces (often predominantly female) lacking 68 P. Dicken, Global Shift: Industrial Change in a Turbulent World (London, 1986), p. 223. 614

The textile firm and the management of labour traditions of collective organization and the capacity to protest as a consequence of over-stocked labour markets and, invariably, state repression of unions.69 The expansion of Japanese multinational conglomerates into neighbouring South Asian states provides a case in point. The ‘high tech’ factories established with Japanese capital in Indonesia were estimated to have made around 300,000 Indonesians working in the handicraft sector unemployed.70 In Bangladesh, workers’ rights were widely ignored by the multinationals, as Nazma Akter, the General Secretary of the Bangladesh Independent Garment Workers’ Union Federation, noted: Freedom of association is a fundamental right that almost all employers in the textile industry deny their employees. The big multinationals … are concerned about the working environment though not about workers’ basic rights … Long working hours and employers’ refusal to grant women maternity leave are some of the other recurrent problems … Some employers use the very nasty device of forcing workers to sign a letter of resignation at the same time as their recruitment contract.71 In the developed market economies, a combination of technological change (including computerization and water jet looms), import penetration and corporate strategy to invest in cheaper labour markets (largely in underdeveloped economies) led to massive job losses in textile manufacturing. The EEC countries alone lost over one million jobs in textiles over the period 1960–1985 (and a further half million in clothing).72 As textile manufacturing capacity and employment levels declined sharply in the developed capitalist economies, the textile conglomerates reaped the additional benefits of enhanced work discipline and control in their surviving home-based plants. The threat of closure and transfer of investment abroad facilitated wage-cutting, labour discipline and work reorganization and ‘speed-up’. These deleterious effects were more apparent because such corporate strategies impacted severely upon specific communities in the developed market economies because of the regional nature of textile employment. Hence areas such as Lancashire in England, Paisley in Scotland, Quebec in Canada, Massachusetts and the Carolinas in the USA, Lorraine in France and the Wallonian region of Belgium have been hit particularly hard.73 In Paisley, Scotland, the Coats plant was run down from employing 14,000 in 1960 to around 1,000 in 1981, while thread production was concentrated in the Madura Coats plant in India – employing 22,000 on wages at 13 per cent of those in Scotland.74 One activist in the Gap textile plant in Belfast 69 P. Stearns, The Industrial Revolution in World History (Boulder, 1993), p. 222. 70 B. Bolton, ‘Standing Up to the Multinationals’, New Internationalist 117 (Nov. 1982), consulted at www.newint.org 71 www.bangladesh-web.com 72 Dicken, Global Shift, p. 251. 73 Ibid., p. 252. The lack of transferable skills of textile workers and limited alternative employment opportunities in such localities only made matters worse. 74 See review of A. Sinclair, Sewing it Up: A Scottish Company in India (Edinburgh, 1982), 615

The Ashgate Companion to the History of Textile Workers commented on the futility of trying to compete with Russian Gap workers paid only 11 cents an hour, Indonesian Gap workers sacked when they tried to form a union and Cambodian Gap employees who were shot at during a protest meeting. This had additional advantages for textile management in Western Europe and the USA because these developments severely undermined a well-unionized, solidaristic work culture, resulting in a collapse in union membership, the neutering of strike activity and severe loss of worker bargaining power in what remained of textile manufacturing in the developed capitalist economies.75 We do need to keep this in perspective, however. Looked at from the point of view of the under-nourished, poverty-stricken citizen in the underdeveloped world, textile factory employment in a transnational company frequently offered more regular work, at higher wages, with more job security and greater individual freedom.76 One study comparing the wages of Vietnamese factory workers employed by textile multinationals found annual wages to be roughly double that of workers employed in Vietnamese firms.77 There were also some ‘ethical’ multinationals, such as the clothing group Benetton, though like the paternalist mill owners in the nineteenth century, these were few and far between in the second half of the twentieth century. There have also been some attempts recently to address child labour and improve work conditions and workers’ rights in the multinationals, with pressure being exerted by the ILO and the International Textile, Garment and Leather Workers’ Federation (ITGLWF) to establish ‘Codes of Conduct’. However, while more work remains to be done on such questions, it appears that these countervailing forces have had little significant effect on the degenerative impact of the textile multinationals upon workers’ rights and work conditions in both the underdeveloped and the developed capitalist economies. Significantly, in his 2005 study of international textile trade union efforts to improve conditions for textile workers, Miller argued that such initiatives had proved difficult, not least because of entrenched anti-unionism among textile employers’ associations and some of the largest multinational textile companies.78 In 2000, Neil Kearney, the General Secretary of the ITGLWF, dismissed the ‘Codes of Conduct’ initiated by some textile multinationals as ‘public relations exercises’ designed more to maintain the brand image than to genuinely improve workers’ rights in developing market economies.79 New Internationalist 112 (June 1982), consulted at http://www.newint.org/issue112/ update.htm (30 September 2009). 75 J. Greer and K. Singh, ‘A Brief History of Transnational Corporations’, consulted at www.globalpolicy.org 76 Dicken, Global Shift, p. 253. 77 J. Bhagwati, ‘Do Multinational Corporations Hurt Poor Countries?’, The American Enterprise Online, June 2005, consulted at www.taemag.com 78 See D. Miller, ‘Negotiating International Framework Agreements in the Global Textile, Garment and Footwear Sector’, Global Social Policy 4 (2) (2004): 215–239. See also www. citinv.it/associazioni/CNMS/archivio/lavoro/childlabourview 79 ITGLWF, Press Releases, ‘Can Corporate Codes of Conduct Tame the Multinationals?’, consulted at www.itglwf.org 616

The textile firm and the management of labour

Conclusions This chapter has focused on the textile firm and examined the changing patterns of employer behaviour and management policies over time, both within the firm and collectively. Common features emerge in the narrative as textile firms evolved historically from small-scale enterprises and proto-industrial forms servicing local markets to large-scale transnational corporations competing in world markets using sophisticated power-driven and latterly automated and computerized technology. Over time, indirect and loose forms of labour control which allowed for considerable degrees of ‘responsible autonomy’ for workers (and especially the skilled artisans) gave way to more direct forms of control, tighter discipline and managerial authority. Thus coercive or ‘despotic’ forms of labour control clearly dominated in textile firms in the eighteenth and nineteenth centuries and the national overviews in this volume indicate a dogged persistence of such methods in many countries deep into the twentieth century. Over time, draconian factory regimes gave way to other more consensual methods whereby textile firms sought to legitimize their authority, including through the recognition of trade unions and the extension of collective bargaining. In this process of protecting and advancing the interests of capital, employers’ organizations of textile firms came to play a key role, though latterly across countries such as Britain, the USA, Sweden and France there has been a degree of disintegration of collective organization and atomization of industrial relations.80 With such transitions, labour management methods were transformed, with a major shift within the firm towards more bureaucracy, higher levels of supervision and monitoring and incremental de-skilling and fragmentation of job tasks. These broad trends in the behaviour of textile firms and their associations, theorized as a transition from coercive to more consensual modes of labour management, embrace a wide range of variation in experience and heterogeneity in behaviour, both between and within national boundaries. In reality, the prevailing picture at any moment in time appears to have been much more uneven, and differing, even sometimes contradictory, policies coexisted (such as scientific management and company paternalism). Indeed, this chapter has emphasized the wide variety of company-level and organized responses in textiles to the rising challenge of worker organization and strikes, ranging from company paternalism, institutionalized welfarism, mass production and automation, more flexible combinations of out-working and factory production, subcontracting and delegation, to outright coercion, collusion with the state, collective bargaining and forms of ‘procedural control’. Moreover, and importantly, the evidence suggests that the control that textile firms sought over labour was rarely complete, was frequently challenged and almost everywhere subject to mediation as a consequence of workers’ organization, resistance and protest. Textile employers appear to have been rarely omnipotent, frequently divided and sometimes completely powerless to withstand fundamental change. 80 Tolliday and Zeitlin, The Power to Manage?, pp. 322–323. 617

The Ashgate Companion to the History of Textile Workers Another major difference was between state-owned enterprises and private, competitive textile capitalism. In China, Russia and elsewhere through much of the twentieth century textile production came under the aegis of the state, ostensibly under more worker-oriented management regimes where production for needs was prioritized over the maximization of profit. While traditional managerial hierarchies continued to exist in textile plants in the communist bloc (with enterprise managers in Soviet Russia responsible directly to the relevant Minister for Textile Production), the degenerative pressures of competing in the world marketplace were nullified by a policy of producing for a ‘captive’ internal market itself closed off to external competition. Hence, wage standardization, improved welfare (in a sector dominated by female labour) and more job security characterized such work regimes. On the other hand, labour control and discipline was maintained by the outlawing of free trade unions and repression of dissenters, and the inefficiencies associated with securing inputs (of raw materials and labour especially) often meant a staccato, ‘stop-go’ cycle of production, with episodes of short-time working followed by periods of intense activity and overtime in order to make up the planned production targets. Indeed, the imperatives of reaching (and indeed expectations of exceeding) ‘the plan’ may well have placed similar pressures upon textile managers and workers to those experienced by workers and managers in the West. Moreover, in both communist China and Soviet Russia textiles played an important role in the industrialization process with the transition to the factory system concentrated in a relatively short time frame – literally over a couple of decades in both cases. While there were differences between the centrally planned economies, factory-based textile production developed in a distinctively labourintensive fashion, lagging somewhat behind the technological innovations and economies of scale evident in the West, with labour discipline sustained through a judicious mix of repression, concession and much social ‘conditioning’.81 In a sense the last-mentioned substituted as a managerial strategy for the company paternalism of Japanese mills (and elsewhere) which worked to cement attachment to the work regime. Such systems merit much more extensive analysis than has been possible here. What is apparent too is that the organization of textile production and the management of labour in the centrally planned economies have been changing rapidly since the 1980s as these countries opened up to the market.82 China is a case in point. Over the last quarter of the twentieth century textile production in China underwent a rapid transformation as the regime reorganized the industry along modern lines, re-aligning itself to export markets and bringing in much foreign expertise and investment to manage this transition, especially, though not exclusively, recruiting textile capitalists from Hong Kong. Shanghai was the fulcrum of this transformation and the main centre of the new export-oriented textile 81 In the late 1970s it was estimated that labour productivity in textiles in China was half that of Hong Kong and almost a fifth that of the USA. See Clairmonte and Cavanagh, World in their Web, p. 191. 82 Stearns, Industrial Revolution, p. 220. 618

The textile firm and the management of labour factories.83 This has attracted much interest from the dominant West European and US textile multinationals, including DuPont, who have invested heavily in textile manufacture in China, especially in artificial fibre production.84 What is evident, then, is that textile firms across the globe adapted and developed their organizational forms and labour relations policies in an organic, experimental and incremental fashion in response to product and labour market pressures, technological change, the growth of trade unions and other prevailing circumstances. Organizational forms and strategies were also influenced by national culture, values and national institutional factors, including labour law and government policy. This produced prevailing common tendencies within capitalist textile production globally, including the key transitions towards the bureaucratization of authority and work within the firm, the delegation of aspects of labour management externally, to employers’ organizations, and the shift towards collective bargaining with labour as a key mode of legitimizing the authority of the firm. However, it must be emphasized again that within these broad and converging patterns what is also evident is the diversity of textile firms’ labour relations policies. Evidently textile firms exercised strategic choice, and much divergence, division and lack of consensus existed within the ranks of textile capitalists globally, and frequently within national boundaries. More research is needed to disentangle this and to understand both the patterns and the determinants of textile firms’ behaviour in labour relations. However, what is apparent is that textile firms’ labour management policies varied widely across time and space, that workers and the state were powerful agencies influencing the strategies of textile firms, and that technology and markets were important determinants of the policies of textile firms and textile employers’ organizations. While recognizing the plurality of organizational forms and strategies, widely differing trajectories and the heterogeneity of textile capitalism worldwide, it is important in the final analysis to appreciate the central importance of the profit motive and the market in determining the key historical mutations in the textile firm globally, from merchant-capitalist, through the familyowned enterprise of the early factory system, to the joint-stock corporations and the vertically and horizontally concentrated multinational conglomerates that now dominate world textile production.

83 Clairmonte and Cavanagh, World in their Web, pp. 189–190. 84 Xinhua News Agency, 14 March 2002, cited at www.china.org.cn 619

This page has been left blank intentionally

25 Spatial division of labour, global interrelations and imbalances in regional development Andrea Komlosy

Explaining the significance of space in textile production requires a multi-level approach, taking into account local, regional, national and global aspects as well as their interrelationship. A main question is how textile production was geographically distributed. Textile production for direct subsistence or local markets occurs almost everywhere people settle and local materials are available for spinning and weaving, in order to meet the basic human need for clothing. It represents a general phenomenon, which develops under specific local and regional circumstances. However, our focus is not on basic textile production, but on specialized production for markets beyond local and regional boundaries. As production for local demand and export markets often interrelate, we have to keep in mind all forms of production. While basic production shows a pattern of equal spatial distribution, specialized export production is usually concentrated in specific locations. Generally speaking, production centres for export are the result of a coincidence of productive competence, demand and accessibility. Export centres are embedded in a regional environment, which provides them with raw materials and semi-processed products (such as yarn, dyeing stuff and so on) as well as sales markets. Spatial division of labour occurs at the regional level, and it equally occurs at the transregional level, including international relations on a global level. Regional divisions of labour can function independently from the global interactions, or the small divisions of labour can be part of a larger, further reaching network. Interaction on both regional and transregional levels can be based on trade or on production chains. Trade acts as a mediator allowing for the exchange of different varieties of textiles, each typical for a certain production centre. In this case the connection is between different centres, rather urban than rural, each offering a complete finished product. We may speak of a horizontal interrelation. Conversely, in the case of production chains, several localities of production are combined in an interlocal or interregional division of labour, each contributing only a single

The Ashgate Companion to the History of Textile Workers step to the process of production. Here interaction takes place on a vertical level. A general agent, usually a merchant or a merchant-capitalist, coordinates the process and skims off the profits. In this case the spatial division of labour contains urban and rural locations, and the different places fit into a centre-periphery structure. Trade exchange, as well as the combination of production sites, leads to regional specialization and interregional division of labour. Splitting up production into chains may result in a spatial separation of production and consumption, some regions being responsible for certain low-paid steps of production only (for example plantation of raw materials, cheap production sites for outward processing and so on) and others for finishing, management and marketing, offering higher wages to the workers and allowing them to become consumers of industrial products.

Chronological sequence versus synchronicity? From 1650 until today, the reach and character of transregional interaction has constantly changed. We can observe a gradual transition from trade exchange towards a transregional organization of production, which is characterized by interregional divisions of labour (production chains). Also, the distance and frequency of transregional interactions has increased immensely and a general transition from small-scale towards more global interactions has occurred. The periodization generally applied to economic development, in the literature as well as in the national overviews of this project, distinguishes between a ‘pre-industrial phase’ (until 1700/1750), a ‘proto-industrial phase’ (1700/1750–1800/1820), an ‘industrial phase’ (1800–1900 mechanization, 1900–1970 automation, 1970 and later digitalization) and a ‘post-industrial phase’ (from the 1970s onwards) of textile production. Although there is some evidence for this periodization scheme, it has to be rejected for being Euro-centric. It draws on a notion of progress derived from developments in those regions of Western Europe that started to hegemonize the world economy in the eighteenth century. Enlargement, intensification and acceleration of global exchange are only registered when they are realized, promoted and imposed on other parts of the world by European actors. But textile history should also be viewed from the perspective of textile centres outside of Europe, namely in Asia, which set up long-distance trade in textiles long before European merchants first participated in and later monopolized global trade. Asian regions had their first industrial climax in pre-mechanical time (before 1820), then suffered enormous pressure from the introduction of the factory system in Western Europe, which became a synonym for ‘industrialization’. However, from 

We prefer the term ‘production chains’ over ‘commodity chains’, because it refers to the combination of several localities of production within an interregional division of labour. For the concept of ‘commodity chains’, see Immanuel Wallerstein et al., ‘Commodity Chains in the World-Economy, 1590–1790’, Review: Journal of the Fernand Braudel Center 23 (1) (2000): 1–199. 622

Spatial division of labour the point of view of the Asian workshops, a definition of industrialization does not necessarily imply the centralization of industrial production in power-driven factories. While Western Europe went through the factory period (1820–1970), Asian regions were ousted from the position of leading world suppliers of textiles, which in some cases resulted in de-industrialization (Bengal and other Indian, Anatolian and Arab regions). In other cases the local textile producers adopted new strategies to meet global demands (Ottoman raw-cotton, Chinese raw-silk), mainly falling back on local and regional markets. Mechanical production in factories in most non-western countries was introduced only in the twentieth century, preparing the grounds for ‘catch-up’ industrialization. The first phase (1914–1970), the building up of a national industry, aimed to substitute home-produced goods for imports. From the 1970s onwards, national programmes were replaced by new strategies of multinational textile and garment producers to dislocate labour-intensive steps of production to cheap locations all around the globe. The above-mentioned periodization is also problematic from a European perspective. To conceive progress in a chronological dimension of production stages conflicts with many European regions’ experience of a simultaneous existence and interrelationship of different technological means and modes of production at each moment in time. Viewed this way, the spatial division of labour appears rather as the ‘synchronicity of the non-synchronous’ (Gleichzeitigkeit des Ungleichzeitigen) than as a chronological conception of production stages. If we overcome the European narrative, global textile history represents a variety of textile competences, relying on different modes of production, which mutually exchanged their specialized products. Until the eighteenth century the highest demand was for Chinese silk, Indian calicoes, English and Dutch woollen cloth, Lake Constance region linen, Persian and Anatolian carpets, and Ottoman damasks and embroideries. Only from 1800 onwards did Western European textiles, based on factory-manufacturing, take the world market lead, conquering export as well as internal markets of Asian producers, until in the twentieth century textile locations moved from the Western European (considering themselves as ‘old industrialized’ countries’) to so-called newly industrializing countries (NICs) in Asia, Latin America and Eastern Europe. This process contains a chronological dimension. Production relying on the coexistence and the equal competence and legitimation of each single location was superimposed and finally replaced by an unequal division of labour. The whole production process became reorganized so that different wages and prices and different ways of organizing and regulating production were combined in order to raise profits for those merchants or producers who were in control of the entire process. The regions where their headquarters were located hence became ‘centres’ of capital accumulation, characterized by relatively high wages and standards of living for workers. Conversely, in the peripheral parts of the production chain, where the more labour-intensive steps of production were carried out, wages were low and people survived because they relied equally on subsistence and  Ernst Bloch introduced this term in his book Erbschaft dieser Zeit (Zürich, 1935). 623

Figure 25.1 The synchronicity of the non-synchronous

Spatial division of labour agrarian activities. So the peripheral households helped to save costs for the central enterprise. It is therefore necessary to look at the unequal division of labour from the perspective of each location or region that was part of the ‘chain’ and contributed to the process, regardless of its function or position within the hierarchy. This general pattern of capital accumulation can be observed in each period of history, since industrial production was organized in ‘chains’, each time combining locations, characterized by specific differences, which change and adapt to the exigencies of the time. Unequal division of labour – in other words the synchronicity of the non-synchronous – can be observed on a regional as well as a global level, the smaller imbalances fitting perfectly in with the broader ones. The dates and extents marking distinctive periods in the history of non-western textile regions to a certain extent coincide with the Western European ones, because from the moment that mechanization of textile production established new standards for competitiveness, all other textile producers had to adapt to the new challenge, albeit in different regional and national forms. However, we can use similar but differently defined periods that are constructed with a completely different aim. These periods do not serve to show historical progress, modelled by the developments in one leading centre, but serve to analyse the combination of inequalities, resulting from the unequal division of labour between centres and peripheries of the world economy. Each period is characterized by a specific spatial arrangement of interregional cooperation. The linear sequence model of the Euro-centric conception of stages must therefore be transformed into a more complex picture, considering the regional diversity of worldwide textile production and its changing combinations of location, techniques, organization of labour and production. The linear model is restricted to a specific world region, which is considered to be standard and against which developments in the rest of the world are rated (from progressive to backward). A multi-focal model better allows for consideration of the diversity of the different regions involved. At the same time it allows the identification of the interactions between different places in each period, and the extent to which these interactions are characterized by cooperation or by dependencies, by equality and mutual respect or by imbalances and the transfer of values from peripheral locations to the centres of accumulation. The following sections present the coexistence of different modes of textile production, combined by trade exchange and/or an unequal division of labour, in five distinct periods. The focus is on the synchronicity of very different levels of organization of both textile production and labour in the various regions participating in a transregional chain of production.

 Immanuel Wallerstein, The Capitalist World Economy (Cambridge, 1979). 625

The Ashgate Companion to the History of Textile Workers

Before 1650/1700 Before 1650/1700 the overwhelming majority of textile production took place at the local level, whether in rural or in urban households, often combined with agriculture. In some strategic branches there were initiatives to concentrate production in (state) manufactories, in other branches putting-out relations were set up in order to increase output; the latter had an impact on space, as rural areas turned into suppliers for urban production centres. Exchange of local specialities was exercised by merchants, who built up trade networks between different textileproducing areas. Specialized textile competence was concentrated in South Asia from the Bosporus to the Chinese Sea for a long period of time, before Europe – inspired by the contacts with the Arab World across the Mediterranean – first joined and then penetrated the Asian centres of competence. European textile expertise was restricted to the south, the centre and the west; Eastern European textile producers did not specialize in exports, neither did North Asian, African or American producers. Many of those regions nevertheless had a high level of textile production for domestic markets. They also took part in short-distance trade. This occurred not only in Europe, but also in Latin America, and in pre-colonial as well as colonial Africa. In Asia the most prominent items for export were Chinese silk, Indian calicoes, Persian and Anatolian carpets, and fine woollens and silks from various parts of the Arab (including North Africa), Persian and Byzantine (later Ottoman) world. Europe was receiving incentives from Asia via the Mediterranean South (Italian city-states, Arab Spain until the end of the caliphate), which were transferred northwards along the following – competing – trade routes: the first line of transfer was across the Champagne fairs to the French northwest and to Flanders (thirteenth–fifteenth centuries); the second was via Upper Italy (Venice, Lombardy) to Upper Germany, where St Gallen and Augsburg became leading textile towns, building a strong axis with Antwerp in the sixteenth century. When the European centres shifted northwestwards because of the dynamics of the overseas expansion, the Netherlands (seventeenth century) and England (eighteenth century) took over as the leading centres of trade. Within each state specialized textile production was limited to specific regions, usually with an urban character. One can distinguish between long-distance trade within single world regions and some trade connections interlinking those regions across the continents. Regional long-distance trade started from all centres of competence; it was most developed within and between Asian regions, in which Europeans successfully interfered after their arrival in the Indian Ocean. The ‘Oriental Trade’ between Asian regions and the Mediterranean Sea originally concentrated on spices and luxury goods, but after the sixteenth century textiles played a growing role. As there was no Asian interest in European textiles, European merchants exported silver conquered in the  Fernand Braudel, Civilization and Capitalism, 15th–18th Century (3 vols, Berkeley and Los Angeles, 1992); Immanuel Wallerstein, The Capitalist World System (3 vols, New York, 1974, 1980, 1989). 626

Spatial division of labour Americas, thus linking the ‘Atlantic Triangle’ with the ‘Oriental Trade’. A part of the goods imported by English, Dutch and French merchants was re-exported to other destinations, for instance other European countries or – in exchange for African slaves – to the Americas. Asian textiles represented a major European input into the ‘Atlantic Triangle’, based on the export of consumer goods to Africa, African slaves to America and Caribbean cash crops back to Western Europe. In the sixteenth and seventeenth centuries cotton textiles were rarely produced in Europe. Until the fifteenth century Venice, Lombardy and the Augsburg region were centres of cotton or fustian production, which declined when the supplying routes for raw cotton were interrupted by the wars with the Ottoman Empire in the sixteenth century. In the seventeenth century the growing European demand for cotton cloth was met by imports from Asia. The multi-centric character of the world economy is reflected by the existence of various centres of textile competence. There was a clear Asian dominance in the world economy, with Europe receiving skills and imports from Asia.

1700–1820: ascent of the manufactory and putting-out system In many publications on textile history as well as in many of the national overviews in this volume, including my own on the Habsburg Monarchy, this period is labelled the ‘period of proto-industrialization’. Here, however, this term is avoided because it represents the Euro-centric view, suggesting that the process of industrialization by definition ends with the introduction of the factory system, as the culmination of all efforts to industrialize. We plead for a broader definition of industrialization, which includes all types of industrial production, regardless of the use of mechanical technology. Therefore this period shall be characterized by the rise of centralized manufactories and extended putting-out systems, which allowed for the increase of market production and a larger and deeper division of labour. As a consequence the organization of textile production included several localities with an increasing distance between them. Manufactories and putting-out activities already existed in former times, but they did not develop into a dominant mode of production until the eighteenth century. By manufactory, we understand a centralized unit of production, which is characterized by division of labour, mostly based on manual work. With regard to space, manufactories centralize production, which was previously scattered in several workshops, in one single location. Nevertheless,  For the concept of proto-industrialization, see Markus Cerman and Sheilagh C. Ogilvie (eds), Protoindustrialisierung in Europa. Industrielle Produktion vor dem Fabrikszeitalter (Wien, 1994).  Andrea Komlosy, ‘Chinesische Seide, indische Kalikos, Maschinengarn aus Manchester. ‘Industrielle Revolution’ aus globalhistorischer Perspektive’, in Margarete Grandner and Komlosy Andrea (eds), Vom Weltgeist beseelt. Globalgeschichte 1700 – 1815 (Wien, 2003), pp. 103–134, at p. 104. 627

The Ashgate Companion to the History of Textile Workers this centralizing tendency of manufacture went hand in hand with decentralized forms of productions. The geographical extension of the division of labour could be organized in different forms: independent artisans selling their products to a merchant (seller’s market), merchants putting out certain processes to be realized in the workshop of an artisan, who worked under his own responsibility (buyer’s market) or who became a more or less decentralized location for a merchant-capitalist, who was responsible for the whole production process (putting-out system). If the putting-out system was entirely dominated by centralized workshops, one can speak of decentralized manufacture. Whether centralized, put out or decentralized, spinning, winding and weaving were carried out by hand-operated devices; mechanical power was only used for fulling, calendering and so on. These mechanized workshops represented small factories, which formed part of a larger manufactory or were operated as independent firms. Centralized manufactories relied on wage labour, although often in combination with forms of dependent or coerced labour. Labour in decentralized units showed an even broader range of organizational varieties. To understand the putting-out system, space is a decisive category. There were urban forms of putting-out, which integrated several locations within a single town into the organization of production. These locations differed by centrality, by land rent, by access to water, water power and waterways, representing a small but distinct universe of an unequal division of labour. Usually putting-out was linked with the transfer of certain processes to decentralized rural producers, who were coordinated by a central organizational unit: a merchant, a merchant-capitalist or a manufactory. Work was put out because of urban shortages of labour supply, especially as labour was more expensive in towns and guild regulations restricted entrepreneurial freedom. Putting-out resulted in far-reaching, sophisticated chains or networks, which were organized around local centres with various types of regions supplying labour, showing various forms of labour relations. These relations ranged from – more or less – paid to unpaid forms of labour, from independent to wage labour, from full-time to part-time labour with an unequal division according to qualification, gender and age. Management and specialist artisans, mainly for complicated weaving, dyeing and finishing, operated the organizational centre. Spinning, winding and ordinary weaving was put out to spinning and weaving regions. Spinning and weaving regions sometimes overlapped and sometimes represented different types of regions, with textile work playing a different role in the working year and the household income. Spinning usually was an additional occupation for rural women and children carried out in periods when agriculture required less work. Wages were low and often below the costs, because families’ living was based on subsistence agriculture. Weaving could also be a rural industry complementing agricultural work, but was more often done by rural full-time weavers, who were assisted by family members. As spinning required more hands than weaving and had lower opportunity costs, the putting-out of spinning encompassed a much greater radius than the putting-out of weaving, reaching into rural regions at a distance of several hundred kilometres. The system changed when 628

Spatial division of labour spinning became mechanized or homemade hand-spun was replaced by imported factory-made yarn and putting-out shifted to the weaving sector. Prominent examples for the far-reaching character of the putting-out system in Europe are Holland, eastern Switzerland with Vorarlberg and southern Germany, and Lower Austria. With the colonial settlement the European system of manufactory also spread to the Americas: we meet colonial textile processing in the Spanish provinces of South America (Argentina, Mexico)10 and in Brazil,11 but we do not find it in regions that only served as plantations, such as the Caribbean region. In North America local textile production only took off after independence. At the beginning of colonization the settlers imported European textiles, but soon the growing demand for everyday cloth was met by a mixture of household and craft production and – in the eighteenth century – centralized manufactory. Only in the case of Brazil did colonial textile production exceed colonial markets and was exported to Portugal. When Portuguese mercantilists, in order to overcome the peripheral role imposed by the Treaty of Methuen (1703) by England, supported the building up of a national textile industry, Brazil was not only seen as a supplier of raw materials, but as a participant in increasing industrial output. Only in 1785 did Portuguese producers put pressure on the government to push back Brazilian exports to Portugal. One can observe similar patterns of the elaboration of the division of labour in eighteenth-century Asian textile regions. The ‘Thread & Money’ system was set up by the British East India Company to meet the growing European demand for cotton cloth.12 Chinese silk production in the eighteenth century took the form of the so-called zangfang system. Merchant-capitalists controlled both sericulture, performed by rural agricultural producers, and silk weaving, which was an urban activity, carried out in centralized workshops and by decentralized individual producers, both hiring wage labourers and being assisted by a range of helpers and specialists who were coordinated by the zangfang agent.13 Unlike the silk industry, the Chinese cotton industry is often denied industrial character, because it has been a rural industry carried out within an agricultural household for many centuries. With rising demand, cotton spinning and weaving in the eighteenth century were increasingly done for the market, allowing for many households to rely on textile incomes rather than on agriculture. In times of  Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’.  Albert Tanner, Das Schiffchen fliegt, die Maschine rauscht. Weber, Sticker und Fabrikanten in der Ostschweiz (Zürich, 1985); Ebeling et al., Chapter 8, this volume, ‘Germany’.  Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’. 10 Bortz, Chapter 13, this volume, ‘Mexico’; Lobato, Chapter 2, this volume, ‘Argentina’. 11 Delson, Chapter 4, this volume, ‘Brazil’. 12 Dietmar Rothermund, Europa und Asien im Zeitalter des Merkantilismus (Darmstadt, 1978), p. 100; Roy, Chapter 10, this volume, ‘India’. 13 Kriti N. Chaudhuri, Asia before Europe: Economy and Civilisation of the Indian Ocean from the Rise of Islam to 1750 (Cambridge, 1990), p. 320; Xu Dixin and Wu Chengming (eds), Chinese Capitalism, 1522–1840 (Basingstoke and London, 2000), pp. 207–212. 629

The Ashgate Companion to the History of Textile Workers crisis, however, they were able to switch back to agriculture. Compared to other examples of mass production of textiles, this system must be considered a huge sellers’ market with independent producers coordinated by merchants, thus the Chinese form of large-scale cotton manufacture.14 In Japan, which was almost entirely isolated from the world market during the Tokugawa Shogunate, different systems to organize the work process also coexisted, including individual handicraft production, manufactories and various forms of putting-out industries. While specialized forms of textile production – apart from sericulture – were urban-based in the seventeenth and eighteenth centuries, a process of ruralization took place at the turn of the nineteenth century, to avoid guild restrictions and higher urban wages.15 Except for Chinese cotton, in all these forms of manufactory and putting-out production the intermediary link between different locations was not realized by trade exchange, but by a division of labour attributing certain production stages to single sites and areas, which all formed part of a multi-local production chain. Transregional demand promoted the putting-out organization vis-à-vis local artisans, eventually controlled by guilds. Generally, putting-out merchants introduced new notions of profitability, ideas commonly attributed to industrial capitalism. Hence they conflicted with independent handicraft producers, who were less able to expand and to accumulate. Both were competing for the support of the state, which – for fiscal reasons – over the long run everywhere shifted from artisan to more commercialized forms of production. Traditional handicrafts could only survive in local niches. Putting-out also favoured a new type of entrepreneur, who ruled out urban (guild) artisans, but integrated rural producers for supplying semi-finished products. Earlier, many of the rural producers had exercised textile activities for subsistence, for local markets or local artisans. They now performed the same activities, but within a different context: they became a chain in a line of locations, which cooperated in order to compete in an export market. This line, or chain, was controlled by a merchant-capitalist, whose profit was based on the low costs of the local producers. These households were able to survive because they relied on various activities, paid and unpaid, each of which supported the others. Consequently, the merchant-capitalist not only profited from the suppliers’ low wages, but also from their families’ unpaid activities. Wherever it came up, the putting-out system and decentralized manufacturing promoted geographical expansion and multiplication as well as a more sophisticated division of labour with stronger hierarchies of command and control in order to meet rising demand, standardization and in-time production. It did not necessarily promote mechanization, as all Asian and Arab examples show. Only in Western Europe did the putting-out system reach the limits of labour supply and was gradually overcome by the introduction of the factory system based on powerdriven machines. The explanation for this labour bottleneck is less a demographic 14 Cliver, Chapter 5, this volume, ‘China’; Kenneth Pomeranz, The Great Divergence: China, Europe and the Making of the Modern World Economy (Princeton and Oxford, 2000). 15 Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’. 630

Spatial division of labour issue than a question of relative market dominance. Encouraged by the good performance of Asian textiles on European markets, industrial entrepreneurs strove to replace Asian imports by domestic production. They put pressure on governments to subsidize their industries and to protect them against imports, carried out by merchant-entrepreneurs who were active in overseas trade. This process is a perfect example of import substitution. It began in England, France and some other Western European states during the eighteenth century without leading to a factory system however; on the contrary, the strive to increase domestic production led to an increase and a geographical expansion of the putting-out system itself. The labour shortage did not result from the substitution of imports for the internal market. It was caused by the ambition not only to conquer the internal market, but also to oust Asian producers from their export markets and to replace them in their domestic markets as well. Such an ambitious expansion of production was only possible if labour-saving techniques were introduced to raise productivity and output. Power-driven machinery represented such a means. After all, it did not replace workers, but it directed them into other activities, which promised higher output. The introduction of the factory system is commonly referred to as the ‘Industrial Revolution’; it arose from the British intent to achieve global industrial hegemony.16 As soon as industrial capitalists had prevailed over merchant-capitalists – a struggle provoking heavy conflicts – the mechanization and centralization of production were declared a national strategy, the state providing all means of market integration, market protection and market conquest. For the workforce the factory system brought along a new type of wage labour, which was characterized by the loss of workers’ control, a new discipline and subordination to the exigencies of the machine. By 1800, the factory system had been introduced into the British spinning industry (starting with cotton and then including wool and linen); weaving followed some decades later.17 It was not until the end of the Napoleonic Wars that British factory-produced yarn started to conquer the world markets, with various consequences. Some European nations successfully introduced the factory system to compete with England, other European textile producers with elaborated systems of manufactory and putting-out had to adjust to the pressure of machine-spun yarn and cloth. Asian exporters, who had until then dominated the world market, were now pushed out of their previous markets and trade connections. Equally, colonial textile producers in Latin America and in New England were confronted with British exports challenging their capacity to supply local markets.18 So the spatial arrangements of textile production were challenged both on the local/regional and on the global level. However, there were vast regions and some entire states, such as Japan, which were not integrated into the world market of 16 Andre Gunder Frank, Re-Orient: Global Economy in the Asian Age (Berkeley, Los Angeles and London, 1998), pp. 276–296. 17 S.D. Chapman, The Cotton Industry in the Industrial Revolution (London, 1972). 18 Blewett, Chapter 21, this volume, ‘USA’; Bortz, ‘Mexico’; Delson, ‘Brazil’; Lobato, ‘Argentina’. 631

The Ashgate Companion to the History of Textile Workers textiles at that time; therefore they were not (yet) affected by the new standards of productivity established by Western mechanization.

1820/1830–1914/1918: regional textiles in view of industrial Westernization As soon as factory-spun yarn and factory-woven cloth appeared on the market, prices fell and put pressure on all those who still span and wove by hand. The labour theory of value, which is a centrepiece of classical political economy from its liberal to its Marxist wing, consequently predicts a victory of new production methods, which will inevitably out-compete more traditional ones. One can indeed observe a process of concentration, in which outdated modes of production were replaced by new ones, leading to a modernization of the economy. In some former industrial regions this pressure provoked de-industrialization. Modernization was never complete, however, and it was ruled out by other means by which local producers could manage to survive and to maintain a competitive position, that is, resist de-industrialization. The more we leave the British perspective behind, the more these other paths of development – which do not correspond with the British model, but which were released by it – gain ground. Two main differentiations have to be taken into account. The first concerns the different levels of textile competence, which was high in some European and most Asian states, but which was less elaborated in many other European states, in Africa and in the Americas. The second concerns the time lag in the introduction of mechanization as well as the different impacts that were provoked by the advent of factory-spun yarn and, later, of factory-woven cloth. With regard to these two differentiations, the worldwide geographical pattern of textile production from 1820 onwards experienced fundamental restructuring. Western and Central European regions, where textile manufacture enjoyed a high standard, were pushed to react on the advance of British machine-spun yarn, which appeared in the 1780s. If they did not want to leave spinning to British mills, they had to enforce tariff protection and to promote the development of domestic mechanical spinning capacities.19 Almost simultaneously spinning mills were opened from Catalonia to Lower Austria, and from the German states to Lombardy, replacing hand-spun first in the cotton sector, followed by wool and linen. The same happened in New England. Hand-spinning only survived in very remote and backward areas and for yarn spun for subsistence purposes. Asian regions did not react in the same way. Hand-spinning was maintained in many regions as the prevailing method, and it was complemented by imported machine-yarn, which put pressure on the wages paid for hand-spinning.20 19 Komlosy, ‘Chinesische Seide’, pp. 116–119. 20 Quataert, Chapter 18, this volume, ‘Ottoman Empire’; Roy, ‘India’; Cliver, ‘China’. 632

Spatial division of labour The late mechanization of weaving resulted in a period of 20 to 40 years when yarn was produced in mills, while cloth was still woven on the handloom.21 As a consequence European textile production changed its geographical pattern. Let us take the example of Austria and Bohemia.22 Some regions became ‘industrialized’: they housed the new mechanical spinning mills, which were overwhelmingly driven by water power. Others specialized in printing, and some in hand-weaving. In the first half of the nineteenth century in many former putting-out regions home-spinning was replaced by home-weaving. Although the weaving techniques remained more or less the same, the organization of the work process often changed significantly. Home-weaving households primarily relied on their textile incomes, and exercised agriculture, if at all, only to feed themselves. They worked for textile entrepreneurs, who often set up their headquarters in the big towns, while puttingout weaving to peripheral regions specializing in home industry. While in the era of manufactory and putting-out spinning and weaving mostly was an additional activity to raise income, rural weavers now completely relied on the entrepreneurs’ orders. When demand dropped and prices fell, the weavers had to increase their working hours, including all members of the family. So they showed a specific vulnerability, while at the same time resisting becoming proletarians. Mechanization of weaving in continental Europe only started in the 1850s, again changing the spatial arrangement of the interregional division of labour. Mechanization went hand in hand with the centralization of the production in factories, leading to the development of factory towns and factory villages in many of the former homeweaving regions. Usually putting-out went on, although the fields in which it took place changed. Home-weaving was restricted to complicated patterns and unusual materials, while special applications and finishing processes, which required manual work, took most of the homeworkers’ time. The second half of the nineteenth century was characterized by fully mechanized spinning and printing, concentrated in the more industrialized regions, while weaving still relied on a combination of centralized factory and home work, represented by a sort of putting-out factory, which reproduced the unequal division of labour on a local scale. Throughout Western Europe, the economic crisis of 1873 accelerated the search for new industrial locations in rural areas in order to mechanize and save both labour and mortgage costs.23 The rural weaving mills often were mere branch establishments, depending on a leading enterprise where management and marketing was concentrated. These company headquarters were usually located in large urban or industrial agglomerations.24 In the United States the modernization of the textile industry was stimulated by the Civil War. Cotton shortage and war damages accelerated the build-up of new industrial patterns: capital moved from New England to rural regions in the South, namely Georgia and the Carolinas, 21 Almut Bohnsack, Spinnen und Weben. Entwicklung von Technik und Arbeit im Textilgewerbe (Reinbek bei Hamburg, 1981). 22 Komlosy, ‘Habsburg Monarchy’. 23 Joseph Schumpeter, Business Cycles (2 vols, London, 1939). 24 See, for instance, Komlosy, ‘Habsburg Monarchy’. 633

The Ashgate Companion to the History of Textile Workers where labour was cheaper and not experienced in industrial action like the New England textile workers.25 To sum up, the textile factory system of the last quarter of the nineteenth century in Europe and the USA showed two faces: a factory type characteristic of core regions, comprising headquarters, research and development, production as well as sales, and a factory type characteristic of peripheral industrial regions, which depended on extra-regional management and specialized in labour-intensive activities requiring a combination of factory and home work. In the old centres of Asian textile excellence the mechanization of spinning and weaving provoked different reactions. In the national overviews in this volume two lines of interpretation coexist with regard to former industrial regions oppressed by British exports. The first one emphasizes the devastating effects of British industrial exports on the domestic industrial landscape, in its worst case provoking de-industrialization.26 The second one rejects the measurement of regional developments by British standards and emphasizes the local producers’ ability to maintain industrial traditions in spite of the challenge of the factory system.27 They follow De Vries and Sugihara, who distinguish between two routes of global industrialization: an industrial path in the West, based on laboursaving technologies, and an industrious path, aiming at mobilizing human rather than non-human resources and at developing technologies for labour-intensive production.28 Spinning must again be distinguished from weaving. As already mentioned, the mechanization of spinning introduced the factory system in some, mostly Western European, regions which turned into industrial landscapes, proletarians becoming the dominant type of workers. Other textile regions partly maintained handspinning, but more and more replaced homespun by machine yarn imported from the spinning mills. With regard to spinning they underwent de-industrialization. If hand-spinning was replaced by hand-weaving, which at that time generally was still a manual task, the transition from spinning to weaving could open new market possibilities. As a cheap input into weaving, imported factory-spun yarn promoted Asian weaving at lower costs. The first half of the nineteenth century saw a boom in hand-weaving. For some of the regions, which previously had specialized in spinning, this replacement, however, meant de-industrialization. The bifurcation between industrializing and de-industrializing regions can be observed in old European as well as in Asian textile regions as long as the handloom still dominated the process of weaving. 25 Blewett, ‘USA’. 26 Parthasarathi, Chapter 22, this volume, ‘Global Trade’; Cliver, ‘China’; Lobato, ‘Argentina’. 27 Delson, ‘Brazil’; Roy, ‘India’; Quataert, ‘Ottoman Empire’. 28 Jan de Vries, ‘The Industrial Revolution and the Industrious Revolution’, Journal of Economic History 54 (1994): 249–270; Kaoru Sugihara, ‘Labour-intensive Industrialization in Global History’, paper presented at the 13th International Economic History Congress, Buenos Aires, 26 July 2002. 634

Spatial division of labour By the time the factory system encompassed both spinning and weaving, Asian and European textile regions had taken different paths. Since weaving became mechanized in the mid-nineteenth century, European mass textiles had been produced in factories that covered the whole process from the preparation of yarn to the finishing of the cloth. Some former European hand-weaving regions now experienced de-industrialization; others became integrated into the factory system, albeit in the dependent position of extended workbenches. The factory-made cloth produced under these conditions became a competitor on domestic and export markets, for which national governments fought heavy battles, which easily turned from economic into military ones. Regional producers in traditional Asian textile centres were in a different position. They faced competition first from English and then from factory-made cloth from other European producers, who had successfully introduced the factory system and strove to find markets for their products. Generally, this advance of mechanized mass production, which usually is referred to as ‘industrialization’, went along with the repression of those textile producers relying on hand – or ‘non-industrial’ – work. In most Asian textile regions neither spinning nor weaving became mechanized during the nineteenth century.29 As a consequence the former leading textile regions no longer were able to compete with factory-made cloth on the world market, and even faced heavy competition on the domestic markets. Direct interference by conquest (in the case of British, Dutch and French imperialism in India, Indonesia and Indochina) or unequal trade treaties enforced by military power (in the case of the Ottoman Empire, Egypt, Persia and China) exercised pressure on the textile producers to open their markets for industrial commodities produced in the West.30 The consequences differed according to the specific specialization of a textile region on certain products for export. Bengal, which overcame the British import restrictions on printed calicoes (1700) by specializing in white cotton cloth for the British printers’ demand, suffered the breakdown of its industry when British producers began to substitute cotton cloth imports by setting up domestic cotton spinning and weaving and no longer depended on Bengal cloth after 1780.31 Bengal often serves as an example for the destruction of regional textile competence by colonial interference. A closer investigation of other Asian textile regions shows that the Bengal case cannot be generalized, however.32 Other regions in southern and central India, which had not relied so much on exports, survived the arrival of factory cloth on the Indian markets and developed products with which they could compete. If we look at the Indian textile landscape, the nineteenth century showed not only withdrawal, but also the development of new forms of textile 29 With the exception of western India (Bombay) and Japan, where textile mills were introduced in the second half of the nineteenth century. See Roy, ‘India’; Hunter and Macnaughtan, ‘Japan’. 30 Wallerstein, The Modern World System, vol. III, ch. 3 passim. 31 Rothermund, Europa. 32 Roy, ‘India’. 635

The Ashgate Companion to the History of Textile Workers production. These processed (imported) machine-made yarn as the basis for a new type of manufactory combining centralized with putting-out activities, thus offering conditions to survive on the domestic market in spite of the presence of imported cloth. A similar adaptation occurred in the Ottoman Empire, where some textile regions gave up, but others developed products, techniques and organizational methods to defend certain internal markets against ‘industrial’ (factory-made) textiles.33 In the face of the manifold ways to maintain textile competence, the concept of ‘de-industrialization’ of Asian textile regions requires revision.34 Conversely, we cannot separate the Asian textile industries, whether successfully surviving or declining, from the impact of the European attempts to repel them from their former leading position in the world market. Even if the European advances were not always successful and stopped neither Asian nor American producers from producing and innovating, they set up a new world economic framework. Until the nineteenth century the textile trade was dominated by centres in Asia, from which European merchants and companies sought to import. Then the European states, headed by England and followed by others, shifted towards the domestic production of what they had previously imported from Asia, fuelled import substitution by mechanization, social and political reforms and put economic and military pressure on the Asian competitors to withdraw from the world markets and to open their domestic markets for British (and other) imports. Asian producers faced a new framework, which provoked a broad variety of reactions and adaptations. Neither for all producers nor for all regions did the crisis lead to a withdrawal. New products and methods of production were developed, and new regions became involved. However, the various regional strategies depended upon the changes in the global relationship. Therefore they must be analysed in relationship to the changing role of Western Europe on the world markets. I agree that the Euro-centric claim to establish the specific British process of import substitution, put forward and accompanied by certain social and technological reforms, as the model that serves to evaluate industrialization processes in other parts of the world should be rejected. However, the impact of Western Europe taking the world lead in industrial production and in declaring its regional path to be the universal way must not be denied, even when the whole process is seen through the perspective of non-European regions. The national overviews in this volume clearly show that Asian producers acted ‘in reaction’ to European competition, that they ‘adopted strategies’ in order to find ‘niches’ and to adjust to a situation of British – or Western – industrial hegemony. Successes were rather restricted to the domestic markets and the specific demands of local consumers, while they had almost no access to export markets. In particular, their successes relied on the ability to compensate the rationalizing effects of factory production by introducing new methods to lower costs by putting pressure on the wages of the workers. 33 Quataert, ‘Ottoman Empire’. 34 As suggested by Roy, ‘India’; also Sugihara, ‘Labour-intensive Industrialization’. 636

Spatial division of labour China and Japan, both leading Asian textile producers with huge internal markets and a protectionist trade regime until the mid-nineteenth century, deserve special attention. Both empires opened their markets to Western competition only in the second half of the nineteenth century – as a result of military and economic pressure from the Western powers. Both countries remained important textile producers. China’s industry continued to be rural-based; it became an importer of Indian cotton textiles, which were exchanged in a triangular trade for Chinese tea for British markets.35 Conversely, Japan was able to modernize its textile industry according to the Western model of mechanization in the second half of the nineteenth century. On the one hand, Japan became a main exporter of raw silk, replacing China in the Western markets; on the other hand, the Japanese also exported cotton yarn and cotton cloth to Asia, Africa and Latin America, where they successfully competed with local producers as well as with Western exports. At the same time Japan defended internal markets against Western imports.36 The discussion also relates to Latin America, which after its colonization became integrated into the world economy as a supplier of silver, gold and agricultural commodities. Because of the distance from Europe, colonial settlers developed domestic textile production for the home market, which copied European models. It was particularly developed in Mexico, Argentina and Brazil. Conversely, it did not exist in those regions that specialized in plantation products. Spain and Portugal did not play a major role in textiles; neither did their exports to the colonies. Textile exports to colonial Latin America, on the one hand, consisted of luxury cloth for the colonial elite; on the other hand, textiles formed part of the ‘Atlantic Triangle’. This was established in the mercantile period, when European merchants and trading companies set up a triangular trade, in which manufactured goods such as textiles were exchanged for slaves who were brought to American plantations to produce cash crops for European markets. Textiles were largely of Asian origin and were re-exported to Africa and America by European companies. Asian textiles were not replaced by European-made textile exports until the first half of the nineteenth century.37 In the early nineteenth century Asian textiles ceased to play a role in both European and American markets. Conversely, British and French textiles, especially after mechanization, were looking for new export markets. The Latin American colonies, which had become independent from Spain – or in the case of Brazil, became the centre of the Empire in 1822 – represented prospective markets. Local textile producers were challenged by European imports. In Argentina local textile production nearly disappeared,38 in Mexico and Brazil local producers were able to resist British imports, to maintain domestic production and modernize their

35 Rothermund, Europa. 36 Hunter and Macnaughtan, ‘Japan’; Kaoru Sugihara (ed.), Japan, China, and the Growth of the Asian International Economy, 1850–1914 (Oxford, 2005). 37 Roy, ‘India’; Quataert, ‘Ottoman Empire’. 38 Lobato, ‘Argentina’. 637

The Ashgate Companion to the History of Textile Workers textile industry according to the European model.39 They produced for the internal market, where domestic textiles and imports from Europe coexisted. In the cash crop and plantation economies like the Caribbean the supply of cloth was met only by European exports. Compared to the period of mercantile capitalism, industrial capitalism in the nineteenth century faced a regionalization of production chains. Textile manufactory and putting-out in the mercantile period seized the globe. In order to enlarge profits (at the expense of their Asian competitors), Western European producers in the eighteenth century sought to develop domestic production capacities in those branches previously relying on imports. They put pressure on governments to protect domestic industries and domestic markets against imports, thus contributing to the nationalization of the states and economies. In order to take over the domestic markets of the former leading textile regions, they promoted market globalization, which, however, was based on the political and military power of their respective nation states. In the Asian textile regions the advance of factory-produced commodities in most of the cases did not provoke a national protection of the market. National protectionism was not on the agenda and if it was, as in China (traditionally until 1840) or Egypt (1810–1840), it was fought by unequal treaties, which could be achieved because of the military supremacy of the European powers.40 Wherever Asian regional textile production resisted the pressures of European imports, it was restricted to domestic demand. Again Japan has to be mentioned as an exception. As in China, the opening of the country visà-vis Western commercial interests was due to Western military supremacy. Unlike in China, Japan’s ruling classes were able to introduce British technologies at the end of the nineteenth century and to develop textile export capacities, which were not restricted to raw materials, but also included manufactured goods.41 Brazil and Mexico set up tariff protection against European imports, while they imported European machines and specialists in order to develop a national textile industry.42 As for the supply of raw materials, overseas regions played a strategic role. In the first phase of industrialization it was not possible to substitute natural fibres. Cotton was not cultivated at all in Western and Central Europe, and there was a lack of wool because of the rising market production of woollen cloth. There were, however, successful efforts to increase the domestic cultivation of mulberry trees to raise silkworms in the southern regions of Europe. The significant role of cotton is rooted in the absence of European grown cotton and the rising demand for cotton textiles, one of the leading sectors of the Industrial Revolution. All European cotton producers had to organize the cotton supply, including cultivation and preparation as well as transport and security. Cotton supply became a national task, each state or region competing against the others and forming alliances with different regions of cultivation. Western Europe 39 Bortz, ‘Mexico’; Delson, ‘Brazil’. 40 Beinin, Chapter 7, this volume, ‘Egypt’. 41 Hunter and Macnaughtan, ‘Japan’. 42 Bortz, ‘Mexico’; Delson, ‘Brazil’. 638

Spatial division of labour imported cotton mainly from America, where the centres of cultivation had shifted from Brazil to the Caribbean Islands, before they arrived in the North American South.43 France and Spain also relied on American suppliers. When France was cut off from her Caribbean planters during the wars with Great Britain, it tightened contacts with the Ottoman Empire and improved land transport across the Balkan Peninsula. The Habsburg Empire, but also Switzerland and some German states, relied on Ottoman cotton, since the Peace Treaty of Passarowitz/Požarevac (1718) had opened commercial contacts with the Ottoman Empire.44 It can be considered a sign of economic decline and subordination under European hegemony that the Ottoman Empire, which had previously exported precious manufactured products, promoted its exports of cotton yarn and raw cotton in exchange for European manufactures. One could have expected that India would also turn from manufactured textiles to raw materials, when its cloth exports to Europe were substituted and its markets partly conquered by British production. India, however, did not become an exporter of raw cotton to Britain after its incorporation into the British Empire. This can be explained by the huge demand for cotton by weavers, who went on to produce for domestic demand. Another reason is the rise of a triangular trade between Britain, colonial India and China, which was set up by the East India Company (EIC) after 1780. China had cottoncultivating regions, but by the end of the eighteenth century they no longer fulfilled the growing domestic demand. So China was interested in importing Indian cotton yarn and raw cotton in exchange for Chinese tea, which was increasingly asked for in Britain. In this way, India became indirectly involved in a triangular trade, being a supplier of unprocessed cotton, albeit not for the British market. In the second half of the nineteenth century, when cotton-processing factories developed in Gujarat, China became a market for Indian cloth too. British and other Western European demand for raw cotton was met by the southern provinces of the United States of America.45 From 1780 onwards, when the introduction of the spinning mill led to a steady increase in British demand, the US South specialized in cotton plantations. They were continuously expanding, expelling the Native American population from their land and driving them westwards. The plantations were worked by black slaves. After the slave trade was halted, the slaves reproduced themselves, providing the plantations with local manpower. The cotton gin, invented by Eli Whitney in 1793, allowed productivity to increase in one of the most labour-intensive fields of cotton production: the cleaning of raw cotton. American cotton was of good quality and because of its long fibres was most suitable for mechanization. In Egypt a similar variety of long fibre cotton was developed. Although part of the Ottoman Empire, Egypt enjoyed political suzerainty. After the end of the French occupation, Governor Mehmed Ali, who had come to power by fighting 43 Lydia Potts, Weltmarkt für Arbeitskraft. Von der Kolonisation Amerikas bis zu den Migrationen der Gegenwart (Hamburg, 1988). 44 Komlosy, ‘Habsburg Monarchy’. 45 Potts, Weltmarkt; Eric R. Wolf, Europe and the People without History (Berkeley, 2000). 639

The Ashgate Companion to the History of Textile Workers the French, forced the cultivation of cotton and opened state manufactories for spinning and weaving. The modernization of the textile industry aimed at building up domestic industrial capacities on an Egyptian level. The process of regional emancipation was stopped by Britain, which intervened in favour of the Ottoman army when Mehmed Ali tried to conquer the Ottoman province of Syria in 1841. As a consequence of the defeat, the Anglo-Turkish trade agreement of 1838 was imposed on Egypt, too. Consequently, Egypt became a main supplier of the world market with raw cotton.46 The more the American cotton was used by domestic manufacturing entrepreneurs, the more Britain counted on West Asia and Egypt, which became a British protectorate in 1882. With the incorporation of Australia and New Zealand into its empire, Britain had also acquired colonies for wool. Despite the association of industrial production with the concept of national strength, the globalizing character of the idea of ‘national industrialization’ is evident: on the supply side it depended on cheaply accessible raw materials; on the sales side it depended on open markets for the manufactures, which – at the same time – were prevented from being industrial competitors by protectionist measures and military deterrence.

1918–1973/1980: the post-colonial period of import substitution If we include the dismantling of the Spanish colonial empire and the Ottoman domination of the Balkans, there are four periods of de-colonization. The first affected Latin America in the early nineteenth century, the second the Balkan Peninsula between 1830 and 1912. After the two world wars a large number of former colonies or internal peripheries in Africa, Asia, Latin America and Eastern Europe became politically independent. National sovereignty enabled a state to pursue a national policy of industrialization, aiming at overcoming its former position within the international division of labour. In many cases textiles were chosen to serve as a leading sector. The old international division of labour with its existing patterns of regional/ national specialization and division of labour was put into question. According to a region’s position within the old division of labour, the attempts to change appear in different perspectives. Although the focus here is on textiles, these strategies always have to be seen in relation to other economic branches. When national industrial strategies were based on textiles, while growth had shifted to other sectors, their success in promoting ‘catch-up’ was limited. Equally, regional hegemony that was based on textiles risked decline as soon as a new economic cycle promoted new branches to take the lead.

46 Beinin, ‘Egypt’. 640

Spatial division of labour Here we distinguish the perspective of the following three types of textile regions. The first were the old European textile regions that were affected by war damages, the Great Depression and the declining significance of the textile sector for economic growth. In the case of changing state borders, as in Austria or Czechoslovakia, the old interregional division of labour was disturbed and the industries had to adapt to the new market situation.47 The interwar period was a time of modernization: spinning faced the introduction of the ring spinning machine, weaving the introduction of automatic looms.48 However, there were many regions and companies that maintained outdated technologies. As a consequence different technological standards and types of enterprises coexisted and traditional ways of production, including homework, survived. Former leading textile regions missed modernization; they also faced the challenge from newly industrializing regions, which set up new textile capacities. This challenge started with a first wave of industrialization in the interwar period and it grew when new peripheral states and regions joined or intensified their attempts to achieve industrial modernization after the Second World War. The shift of textile production from central to peripheral regions took place on the international as well as the interregional level. Dislocation of textile production to newly industrializing regions in former colonies and internal peripheries went hand in hand with the move from central industrial areas to cheaper rural locations within single countries. The second type of textile region consists of former agrarian regions in the Third World or in Eastern and south-eastern Europe, some of which had only reached independence after 1918, or after 1945. They began to play an important role as textile producers in the twentieth century. In some cases, for instance the Caribbean, Africa and New Zealand, the new independent states continued to deliver raw materials, such as cotton or wool, for manufacturing regions. However, there was a striving towards setting up domestic industrial capacities in the new nation states. To start with, import-substituting industrialization was emphasized in the consumer industries, and last but not least in textiles. Among the countries with a textile take-off in the interwar period, Argentina and Uruguay deserve attention in the southern hemisphere,49 as well as Hungary, Romania and Yugoslavia in Eastern and south-eastern Europe.50 Import substitution very soon met the boundaries of available capital, machinery, know-how and purchasing power, impeding the ‘catch-up’ with the old industrial regions. However, the Great Depression, which led to a revival of protectionism, and the Second World War enlarged the possibilities for some of the peripheral competitors. For Latin America this favourable period had already started in the 1930s, and it concerned not only the NICs but also those nations – like Mexico or Brazil – that had already been established as suppliers of their own domestic textile markets for a long time. In Eastern and south-eastern

47 Komlosy, ‘Habsburg Monarchy’. 48 Bohnsack, Spinnen und Weben. 49 Lobato, ‘Argentina’; Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’. 50 Komlosy, ‘Habsburg Monarchy’. 641

The Ashgate Companion to the History of Textile Workers Europe, which were theatres of war, ‘catch-up’ was interrupted by the Second World War and only taken up under communist conditions.51 The third type of textile regions were non-European regions, in India, China, Japan and the Ottoman Empire, the leading textile regions of the world before industrial growth in European regions took off in the eighteenth and nineteenth centuries. Others, in Asia, but also in West Africa, in Mexico and in Brazil, produced for internal markets rather than for export, yet had a diversified market production. The competition of European factory-made textiles had different impacts, reaching from isolation and trade protection (China until the Opium wars; Japan until the Meiji restoration in 1867), to de-industrialization (Indian coastal regions such as Bengal, Ottoman regions, Argentina) or a reorientation on domestic markets, or at least on some sectors of the domestic market where the local products were able to resist British and other Western imports. When China, India, Turkey, Egypt, Mexico and Brazil became leading producers of cloth in the twentieth century, it would not be correct to call this expansion a process of ‘catch-up’ industrialization, comparable to the industrial development of an agrarian society, which until then had supplied the world economy with raw materials. They had been industrialized countries long ago; in the eighteenth and more so in the nineteenth century their textiles were ousted from the world market which had been conquered by Western manufactures. When in the Western centres of the world economy the dynamics of growth shifted to other sectors, however, peripheries had the chance to take over a leading role in textile production. The restructuring of the international division of labour opened the way for peripheral locations to become textile producers for the world, or for the domestic market. As a consequence of textile proliferation, in the interwar period and after the Second World War the production of textiles was widely spread over the globe. Enterprise and labour organization, as well as technological levels, showed a broad range, including handicraft, home, factory and putting-out production. A wave of automation, which spread to peripheral world regions, coexisted with traditional methods, based on the availability of industrious working hands. Despite manifold import and export activities and an ongoing international division of labour, the spatial reference of production was the nation-state: in the old industrial countries economic nationalism was due to the rise of new competitors in the South and East, and to the political tensions before and after the world wars; in the newly and renewed industrializing countries the national focus was an immediate consequence of the attempt to build up or to strengthen domestic economic forces. Import-substituting industrialization as well as the modernization of existing industries was conceived as a national project, making use of the following supportive and protective measurements: tariff protection,

51 Hannes Hofbauer and Andrea Komlosy, ‘Capital Accumulation and Catching-Up Development in Eastern Europe’, Review, Journal of the Fernand Braudel Center 23 (4) (2000): 459–502. 642

Spatial division of labour active industrial policy, state regulation of labour conditions and social conflicts, integration of trade unions in state politics and so on.52 To evaluate success and failure, one would have to distinguish between different regions and states. Among those newly and renewed industrializing countries, where textiles represented a leading sector, we can observe very different paths. Regardless of the difference, each state acted in a rather isolated way, trying to overcome or to avoid dependency, and therefore reduce the international division of labour to a minimum. The conditions, however, to start and to compete differed from the old industrial powers. As the new competitors – many of them former colonies or internal peripheries themselves – did not dispose of colonies, they had to rely on internal sources in order to raise capital, labour and methods to reduce costs. So they were trapped in a vicious circle: if lack of capital was compensated by credits or foreign direct investment, profits flew abroad. If capital was to be replaced by more labour-intensive forms of production, wages had to be kept low, with a negative impact on domestic purchasing power. Keeping wages low had a regional impact too. In order to lower costs, production was dislocated in peripheral regions, where people were able to compensate for low wages with various other incomes, among them those provided by migration into industrial centres. So the attempts to catch up or to compete internationally contributed to the rise of regional imbalances in single countries. The former leading Asian textile producers as well as some NICs had again become leading world textile and garment producers. But textiles no longer represented a driving force of industrial development. To rely primarily on textiles can hence be considered an indicator of a country’s very vulnerable position in the world economy. The more national industrialization programmes turned out to be successful, the more textiles lost their leading role and were replaced by other industrial branches to lead growth.

1973/1980 onwards: re-globalizing textile production During the period of reconstruction after the Second World War, in the industrialized West priority was given to national consolidation as well as to the re-introduction of multilateral economic relations among the centres of the world economy. Although being important providers of raw materials for the industrialized world, peripheral regions in the South and East were somehow neglected. Paradoxically, this neglect enlarged their possibilities to pursue ‘catch-up’ and industrial modernization.53

52 Henryk Szlajfer (ed.), Economic Nationalism in East-Central Europe and South America 1918–1938 (Geneva, 1990); Joseph L. Love, ‘Theorizing Underdevelopment: Latin America and Romania, 1860–1950’, Review, Journal of the Fernand Braudel Center 11 (4) (1988): 453–496. 53 Hofbauer and Komlosy, ‘Capital Accumulation’. 643

The Ashgate Companion to the History of Textile Workers In the 1960s extensive growth had reached its limits, leading to stronger competition among the old industrialized nations and new attempts to integrate peripheral locations all over the world, including internal peripheries in the developed countries themselves. Dislocation of industries, outsourcing, labour migration and credits were indicators of increasing interactions between the centres and peripheries of the world economy. The crisis of 1973 can be seen as a turning point. It was rooted in the decline of profits because of sharpened competition. The crisis led to new ways to renew profits, mainly the rationalization of production and the search to lower production costs. While rationalization introduced the shift from industrial to knowledge-based, post-industrial capitalism in the old industrial countries, the search for wage reduction opened the field for peripheral production sites to take over global manufacture. The term ‘new international division of labour’54 refers to the shift of industrial production to peripheries in the South and the (at that time Communist) East. Their contribution to the world economy was no longer reduced to raw materials, but to cheap industrial labour. The perspective may be a Euro-centric one, neglecting the fact that many of the NICs disposed of an industrial tradition that dated back to pre-colonial and early colonial times or – in the case of interwar and post-war import substitution – to the early periods of de-colonization. In other cases, mainly in the Caribbean and in parts of Latin America and South East Asia, industrial capacities were only set up now, while many cash crop and raw material producing countries – for instance in Africa – were not at all involved in the process of industrial transfer, which followed the search for cheap production sites. However, the international division of labour changed after 1973 and deserved a ‘new’ label. The changes became visible only in the 1980s, leading to a shift from a self-reliant to a neo-liberal political paradigm, replacing a – state-centred and state-regulated – national development policy by the idea of transnational integration, administered by international monetary institutions. Communist Eastern Europe had been affected by the new international division of labour since the 1970s; the change of regimes in 1989 only accelerated the transformation of many Eastern European regions into extended workbenches for Western companies.55 Hence the spatial order of manufacture changed. Textiles and garments had long ceased to be leading industrial sectors; nevertheless textiles and garments are paradigmatic examples for the shift from nation-based production, which had been typical for the ‘short’ twentieth century, to present-day global manufacturing, which combines different locations all over the world to one global production chain. The reason for the special involvement of textiles and garments in the global restructuring is rooted in the low profitability and labour intensity of cloth and clothes production. In order to compete, the big companies of this sector were among the first to look for cheaper production sites in peripheral world regions. 54 Volker Fröbel, Jürgen Heinrichs and Otto Kreye, Die neue internationale Arbeitsteilung (Reinbek bei Hamburg, 1977). 55 Bettina Musiolek (ed.), Ich bin chic, und Du mußt schuften. Frauenarbeit für den globalen Modemarkt (Frankfurt am Main, 1997). 644

Spatial division of labour Within a short time, production of textiles and garments sharply decreased in the old industrial regions, which underwent a process of ‘de-industrialization’, while peripheral regions in South and East, with or without a textile tradition, were chosen for new investment in textile and garment production. National governments and international development agencies often welcomed the investment of multinational firms, hoping that they would strengthen domestic industrial capacities and promote modernization. The transfer of labour-intensive manufacturing to old and new production sites in the ‘Third World’ (now also including parts of the former ‘Second World’) was completely opposed to national modernization and import substitution, which had guided the post-colonial restructuring after independence. Under the new regime of global outsourcing, research, development and company headquarters remain in the world economic centres in Western Europe, the United States and Japan, while the peripheral production sites are mainly activated because of their cheap labour. They produce for the global production chains. There is no continuity between the industrialization, which aimed at modernizing the national economy and developing the internal market, and the global enterprise. On the contrary, they are in direct opposition to each other. The multinational firms produce in special free production zones, often exempted from taxes and national labour regulations, outsourcing work to a myriad of suppliers and subcontractors, who guarantee still lower wages and total flexibility. The spatial and social picture completely changes: we observe new locations, and a new social and geographical composition of the workforce. Peripheral regions get trapped in mutual competition to offer the cheapest set of conditions. The general trend is towards a total withdrawal of the state from shaping the process of industrialization. However, this process would not be possible without the elites of the peripheral states, and it takes place with the approval of the World Trade Organization, which replaced state regulations with international market regulations. In China, the state still exercises control over mass production for export markets, but it equally assures private investors that labour and production costs are kept to a minimum. In the last 20 years in the old centres of the world economy, textiles and garments faced a sharp decline. These sectors now only play a marginal role, reduced to some highly rationalized processing of special fibres and cloth (in factories without workers). Global management, marketing and financing does not follow dislocation, but is concentrated in the global financial and commercial centres. Textile enterprises can only survive in rare niches that meet specialized demand. Hence we observe a re-globalization of the global production chains. A partial reversal of the globalization process can recently be observed in the garment sector, where some labels started to outsource sewing and finalizing activities to metropolitan sweatshops instead of peripheral regions. This process is limited to high-price fashion, however. As the workforce of the metropolitan sweatshops is composed of immigrants from peripheral regions of the world, it is precisely a spatial mirror of global outsourcing.

645

This page has been left blank intentionally

26 How will we get our workers? Ethnicity and migration of global textile workers Roberta Marx Delson

Textile workers are examined here in global perspective. Using the criteria of ethnicity and migration to measure the composition of textile labour forces, the discussion analyses the national overviews collected in this volume and focuses on the period from the sixteenth century to the present. The principal finding of this chapter is that a model of coerced-proto-industrialization of textiles (especially in the Americas) coexisted with, but is distinguishable from, contemporaneous European modalities. Even into the nineteenth and twentieth centuries, the ethnic and migratory profile of textile workers in the Americas continued to display far more diversity than its counterpart in Europe and elsewhere. The fact that textile industries are often cited as the historically largest employers in any given region, and frequently represent the greatest value with regard to manufacturing capacity, adds little to our understanding of what the composition of that labour force was or the processes via which those workers were recruited. Regardless of whether textiles were produced at home for self-consumption, in the putting-out systems and small workshops associated with proto-industry or at large mechanized mills, the labour component is fundamental; in modern  I thank the editors of this volume, Lex Heerma van Voss, Els Hiemstra-Kuperus and Elise van Nederveen Meerkerk for inviting me to participate, for suggesting this assignment which has proved so fruitful, and for helping me to come to Amsterdam to present the first version of this contribution. I am also grateful to the other members of the project for their comments in Amsterdam and especially to Dr Jan Lucassen for his review, which helped me to focus on specific modes of production. I also thank Ingrid Lennon of the American Museum of Natural History for her help with interlibrary materials, my colleague Dr Sharon Sundue at Drew University for her insights on slave production in the United States and Dr Eric Delson of the AMNH and Lehman College/ City University of New York for his editorial assistance.  Rondo Cameron, A Concise Economic History of the World: From Paleolithic Times to the Present (Oxford, 1989), p. 196.

The Ashgate Companion to the History of Textile Workers terminology it is the ‘bottom line’. Each of the national textile histories written for the Textile Project has in some manner raised the issue of securing labour. Using their findings and data, my assigned objective in this chapter is to advance the discussion of textile labour recruitment beyond what we already know about Europe by focusing on the history of migration and ethnicity of textile workers in the global context.

Defining the problem It is perhaps worthwhile observing, at this juncture, that the discourse of this project has addressed itself primarily to capitalist-oriented textile production, whether in proto-industrial or commercial settings. Choosing this direction, however, automatically discounts the vast amounts of fabric produced globally by skilled workers in non-commercial settings for ceremonial, religious and/or status reasons. Ethnographic studies confirm that certain fabrics were and continue to be produced only by specific ethnicities (for example in Okinawa); it follows that migration patterns may also be affected by this mode of production. For the purposes of this chapter, however, discussion of ethnicity and migration is limited to the context of commercial textile production. Towards that end I offer the following standard, anthropologically accepted definition of ethnicity: ‘ethnicity refers to a sense of group affiliation based on a distinct heritage or worldview as a “people”’. It goes without saying that such group affiliations may clearly represent racial distinctions as well. What the national textile overviews have revealed in toto is that even within a single nation’s textile labour pool there may be considerable ethnic variation; additionally, workers may comprise an ethnicity distinct from that of the management and/or the owners of textile enterprises. Further, the gulf between skilled and non-skilled workers may similarly represent different ethnic/ racial or religious divides. Complex as the concept of ethnicity may be, it pales in comparison to migration as an analytical category. Migration patterns (essentially ‘the movement of a person or people from one place to another’) can be further broken down into categories in which migrants seek employment on a seasonal basis, returning occasionally to their home locales, or those who make more permanent moves (for example textile

 See, for example, Justine M. Cordwell and Ronald A. Schwartz, The Fabrics of Culture: The Anthropology of Clothing and Adornment (The Hague, 1979).  Louise Allison Cort, ‘The Changing Fortunes of Three Archaic Japanese Textiles’, in Annette B. Weiner and Jane Schneider (eds), Cloth and Human Experience (Washington, 1989), pp. 377–415, at pp. 396–404. Banana fibre cloth produced by the Okinawans became a status symbol for Japanese monarchy. Today its production is perpetuated under the aegis of the Folk Craft Movement.  Barbara D. Miller, Cultural Anthropology, 3rd edn (Boston, 2005), p. 14. 648

How will we get our workers? workers who arrived in the Americas without plans to return to their homelands). Additionally, the labour composition of a specific textile industry may change over time as migrant labourers become naturalized and, thus, permanent residents. From an analytic point of view, patterns of migration exhibit so many dichotomies that they are best conceptualized as opposing heuristic pairs; these would include temporary vs permanent migration; long distance vs short distance migration; internal migration vs international migration; migration for political or religious reasons (that is refugees) as opposed to migration for purely economic motivation; migration of skilled vs non-skilled labourers; network migration vs paternalistic migration; and non-voluntary migration (coercion both of slaves and non-slaves) vs voluntary migration. Finally, none of the above categories is necessarily mutually exclusive; as the national textile overviews have clearly demonstrated, any given textile industry may have experienced multiple patterns of migration and ethnic composition of workers over time. A further impediment to fully understanding the global textile labour force is the presumed universality of the European experience. Without getting ahead of myself (as this is the crux of the present discussion), I will argue that the well-understood and well-studied division of European textile production into three modalities, that is home manufacture for self-use, home or workshop manufacture for the market by individuals who were paid for their labour (sometimes referred to as putting-out, piecework or proto-industry) and factory manufacture for widespread commercial distribution, may not be sufficiently encompassing to offer explanations for production configurations outside of the European context. Moreover, the recruitment patterns outside of Europe and the subsequent ethnic composition of the labour force may vary greatly from the model of predominantly ethnically homogeneous, wage-earning workers who travelled short distances to textile enterprises which is the hallmark of European industries. Even the push and pull factors influencing migration outside of the European context can be quite distinctive. In this chapter I have treated the issues of migration and ethnicity chronologically and comparatively. To obtain the necessary data I have synthesized both the ‘national’ textile overviews and the comparative, thematic treatments (to which this chapter belongs) that were compiled for this project; additional data from other sources have also been incorporated. While this discussion is not totally comprehensive, I believe that the data presented here do confirm my hypothesis

 Ibid., p. 346.  This point was apparently reached in Leiden in the late seventeenth century. See Leo Lucassen and Boudien de Vries, ‘The Rise and Fall of a West European TextileWorker Migration System: Leiden, 1586–1700’, in Gérard Gayot and Philippe Minard (eds), Les ouvriers qualifíes de l’industrie (XVIe–XXe siècle). Formation, emploi, migrations (Lille, 2001), pp. 23–42.  For example, see Maxine Berg, The Age of Manufactures: 1700–1820 (Oxford, 1985); Fowler, Chapter 9, this volume, ‘Great Britain’. 649

The Ashgate Companion to the History of Textile Workers that significant deviations from European models are evident when textile labourers are considered globally. For clarity’s sake, I have divided this discussion into two segments: what I have called ‘Stage I’ of this discussion outlines European textile worker modalities, and contrasts them with non-European examples from the sixteenth to the end of the eighteenth century. This section is organized by geographic region. Stage II, ‘The Modern Period’ is divided chronologically into two segments, that is the nineteenth century and the twentieth up to the present. Here the discussion privileges specific types of ethnic/migration patterns within those time frames.

Stage 1: The sixteenth to the eighteenth century The period from the sixteenth to the eighteenth century was characterized by the growth of capitalistic (and specifically textile-producing) networks within Europe and the Middle East. It also saw the emergence of specialized textile centres and, towards the end of the period, the beginnings of mechanized textile production. Those who produced textiles for the market were free labourers who were transformed by their work into wage-earners either in urban/semi-industrialized settings or in proto-industrialized, rural workshops. On the other hand, this time frame was also marked by the expansion of European colonial empires abroad, largely through the efforts of enslaved, unpaid labour. This phenomenon would have specific and lasting ramifications for the production of textiles as well as for the labourers who produced them.

Europe In the time frame under consideration, Europe presents a diversified picture in which the workforce recruited for textile manufacture either consisted of local labourers, both rural and urban, who were ethnically homogeneous with the mainstream population, or migrant labourers of differing ethnicity. Again there was no mutual exclusivity and mixes of both types of labour were present. For most of Europe, the involvement of labourers in textile production took three basic forms: people spun and wove in their homesteads for family use; or they worked at home, or in rural workshops, producing textiles (or spun yarn) which were sold in the larger market (this is the proto-industrial model); or they migrated to large textile centres (either seasonally or permanently) which had need of substantial numbers of workers to prosper, a phenomenon which steadily increased as mechanization began to influence production at the end of this period. For obvious reasons, where textiles were produced at home for self-consumption, migration was not an imperative although, as Moch correctly points out, this does

650

How will we get our workers? not imply that there was no movement in the rural areas. In the second instance, in which cloth or yarn was produced at home for commercial sale in the market, the incentive was obviously to remain at home, or in the village, in order to maintain the manpower necessary for spinning and/or weaving. Engaging in some aspect of textile production was sometimes done on a seasonal basis, presenting an opportunity to supplement income when agricultural work slackened; this had the effect of introducing wages (albeit sometimes for piecework) to rural industry.10 Sometimes, but not always, the switch to employment in textile production necessitated temporary relocation to an urban centre.11 Finally, there was substantial, permanent migration to large textile centres where factory-like conditions existed, an exceptional situation which will be discussed below. Regardless of which pattern is being examined, it was certainly the case that European textiles were produced mostly by ethnically homogeneous workforces. In this regard, Denmark presents a ‘typical’ case in which textile production was done by local populations and represented a mix of home production, weaver villages and some regions of specialized cloth production. Textiles were not only produced by peasants in their homes: ‘In the rural areas, textiles were also produced by artisans: weavers, employed by an estate, or working in their own shop in a village.’12 Interestingly, the overwhelming local nature of this textile industry did not preclude Danish artisans from travelling through central Europe for experience and in that way incorporating some technical German words regarding textile production into the Danish vocabulary.13 Towards the beginning of the nineteenth century, the pattern had shifted somewhat (although remaining ethnically homogenous) to a situation in which in some towns had so completely dedicated themselves to production that almost every household contained a loom.14 Whatever differences existed among the textile workers were likely more a reflection of degree of skill than ethnicity. This description also rings true for textile workers in Spain and Portugal, where ethnically undifferentiated populations provided the workforce for textile enterprise; these were largely non-migratory. Where the pattern begins to diverge is with the emergence of large textile centres, a phenomenon which predated the age of mechanization of the late eighteenth century. These centres, often historic cities which now experienced increased textile production, functioned as poles for increasingly long-distance, but still national, migration. Specific textile producing areas emerged, often associated with fabric specializations. In Spain, for example,  People moved to get married, harvest teams moved from farm to farm and so on. See Leslie Page Moch, Moving Europeans: Migration in Western Europe since 1650 (Bloomington, 2003), p. 2. 10 For example, seasonal migration had a role in the wool industries of Amiens. See Moch, Moving Europeans, p. 52. 11 Ibid. 12 Christensen, Chapter 6, this volume, ‘Denmark’. 13 Ibid. 14 In Dragor, for example. Christensen, ‘Denmark’. 651

The Ashgate Companion to the History of Textile Workers the linen industry was centred in Galicia, silk in Valencia, wool in Castile and cottons in Barcelona.15 Additionally, government intervention in the eighteenth century resulted in the creation of Royal Factories in urban zones; it is not clear if the labour which staffed these ventures was recruited from rural areas or obtained locally. Some regions experienced a division of labour between rural and urban areas; for example, the weaving of cotton fabrics took place in the more urbanized factories of Barcelona, while the actual spinning of the cotton yarn was carried out by peasants (women) in the more rural areas of Catalonia.16 In Portugal, textiles were similarly produced throughout the country (including the Atlantic Islands). As in Spain, government sponsorship in the eighteenth century resulted in the creation of centralized ‘spinning’ schools for women in various locales across the country, thus precluding long-distance migration.17 In the Iberian case it would seem that there was no need for external or international migration, as sufficient labourers could be recruited from within national borders. Thus, these textile industries would perforce remain ethnically homogeneous. (Of course, one must acknowledge that among themselves workers may have perceived differences in those who hailed from unfamiliar, but still national, regions.)18 For Iberia, then, as for most of Europe, where major movements of the textile labourers occurred it was in the migration of ethnically homogeneous workers to large textile centres, a pattern which intensified with the introduction of mechanization.19 There were, however, some significant exceptions to this general pattern of ethnic homogeneity and short-distance migration among European textile labour forces. The first exception occurred when politics and/or religious wars made it uncomfortable for workers to remain in certain areas. The second exception is associated with economic factors and, as Leo Lucassen and Boudien de Vries have pointed out, was equally or even more important than the first.20 The eclipse of local industries (either from competition or sometimes as a dislocation due to war) often resulted in the necessity of long-distance migration, sometimes sporadically undertaken as workers moved from town to town. On the other hand, newly emergent textile centres, such as Leiden, actively solicited skilled, long-distance migrant textile workers. This recruitment effort was so widespread that it resulted in what Lucassen and de Vries have termed a ‘Textile-Worker migration system’, which functioned in Leiden from the late sixteenth century until its demise at the end of the seventeenth century.21 15 Smith, et al., Chapter 17, this volume, ‘Spain’. 16 Ibid. 17 Nuno Luís Madureira, História do Trabalho e das Ocupações, vol. I: A Indústria Têxtil (Oeiras, 2001), pp. 47–48. 18 Regional suspicions were not unique to Spain. Lucassen and De Vries, ‘The Rise and Fall’, p. 27 write that even though Flanders and Holland were under the rule of Charles V until the Dutch revolt, the Flemish were considered outsiders. 19 On this point see Moch, Moving Europeans, pp. 100–160. 20 This is exactly the point of Lucassen and De Vries, ‘The Rise and Fall’. 21 Because of the success of the textile-worker migration system, Leiden eventually had 652

How will we get our workers? Lucassen and De Vries, as well as Elise van Nederveen Meerkerk, have examined the migration of textile workers to Leiden from areas as diverse as France, Flanders and Germany.22 Many of those who relocated to the city were already skilled textile workers whose expertise was actively solicited; Lucassen and De Vries point to competition between Gouda, Haarlem, Leiden and Amsterdam for such workers.23 Others arrived of their own accord because of recession in the textile industries of their native regions. Either way, the authors conclude that the fact that workers from specific Flemish areas chose to relocate to one particular city in the northern Netherlands over another may be attributed to a combination of chain migration (where one migrant already working in a textile centre encourages a friend or family member to relocate to that industry) and active recruitment.24 But there was another trajectory which resulted in long-distance migration. I am referring to that migration which was occasioned by religious persecution. Again, this type of migration was very much evident in the northern Netherlands, which attracted French Protestants fleeing from inhospitable conditions in Catholic France, as well as Dutch Protestants determined to escape the unstable religious climate of the southern Netherlands. Having migrated to the friendlier northern cities of the United Provinces, they found work in Haarlem and Leiden in the woollen and linen industries. The latter city apparently presented a particularly welcoming milieu since one did not have to be a citizen to work in the textile industry. Jews fleeing religious persecution in Spain and Portugal, as well as in Eastern Europe, also sought positions in the spinning as well as weaving sides of the industry. While such textile occupations were thought of as ‘immigrant’s positions’, workers were not necessarily paid lower wages than Dutch-born workers. However, such immigrants ‘were the first to suffer from the depression in the second half of the seventeenth century’.25 The Leiden experience was duplicated on a smaller scale in the textile centres of East Anglia, England. The city of Norwich, for instance, which ordinarily depended on local labour for cloth production, also solicited expert Protestant Huguenot workers from Flemish areas. These skilled Flemish workers came in two successive waves, representing two separate forms of migration; the first arrived in the late sixteenth century as a result of religious persecution, while the second wave (a chain migration) was initiated in the 1680s by the invitation of a descendent of one of the first-wave Flemish workers.26 enough workers to preclude recruiting outsiders. See Lucassen and De Vries, ‘The Rise and Fall’, pp. 38–39. 22 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’. 23 Lucassen and de Vries, ‘The Rise and Fall’, p. 28. 24 Ibid., p. 29. 25 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 26 Leslie Page Moch, Moving Europeans, p. 48. Lucassen and De Vries, ‘The Rise and Fall’, p. 30 suggest that descendants of the initial wave of Flemish immigrants to England actually ended up in Leiden in the seventeenth century. 653

The Ashgate Companion to the History of Textile Workers Huguenot migration continued even into the next century. In Catholic Ireland, skilled Protestant immigrants (who were largely Huguenots) were encouraged by Protestant landlords to settle in flax-weaving towns. Sir Richard Cox’s offer of premiums for such immigrants to work in his linen business was so successful that the town grew from 87 to 117 households between 1735 and 1749.27 Such premiums suggest that here the primary motivation for migration was economic rather than religious. In those areas of Europe where workers of multiple nationalities were subject to imperial authority migration was facilitated by official decree. Leaving aside the very real concerns raised by Donald Quataert in his comparative paper on the difficulties of comparing nation-states with empires,28 Andrea Komlosy makes it clear that the structure of the Habsburg Empire eased the process of migration in the eighteenth century, especially for skilled textile workers. This relocation involved workers travelling great distances from the more remote parts of the realm to centralized textile facilities in Austria itself. In this manner, linen and wool cloth weavers arrived from the Austrian Netherlands, tapestry-makers from Lorraine, barchent weavers from Swabia and silk weavers from Italy. As Komlosy points out: ‘A great number of foreign technicians, apprentices and other artisans followed their masters or entrepreneurs from the industrial provinces on the Habsburg fringes, when these were successfully established in the Austrian heartlands.’29 The author further observes this migration was essential to the development of the specifically Austrian textile industry: ‘Thus the Austrian Netherlands, the Vorlande and Lombardy, which did not take part in the unification of the Habsburg market, were of critical importance for the transfer of capital, technology and know-how, and gave important impetus to the industrial development of mainland Austria.’30 This was not the only pattern of textile production in the Habsburg Empire. While the long-distance migration described above was geared towards centralized manufacturing facilities, it was paralleled by a localized migration to artisan workshops. In these latter settings workshops maintained the guild system and recruited nearby workers on the basis of patrimony, marriage and professional career. ‘Unskilled labour for domestic industries and manufactories was recruited in the rural neighbourhoods of the proto-industrial districts.’31 The laws ending serfdom encouraged this pattern, as did the creation of new settlements in protoindustrial areas. Both factors resulted in a growing class of house and garden owners, who used their small plots of land to feed their families while still working in the textile industry. 27 Jane Schneider, ‘Rumpelstiltskin’s Bargain: Folklore and the Merchant Capitalist Intensification of Linen Manufacture in Early Modern Europe’, in Annette B. Weiner and Jane Schneider (eds) Cloth and Human Experience (Washington, 1989), pp. 177–214, at pp. 188–189. 28 Quataert, Chapter 23, this volume, ‘Proto-industrialization’. 29 Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’. 30 Ibid. 31 Ibid. 654

How will we get our workers? Textile facilities in fringe areas of the empire likewise attracted migration; Brno (Moravia) and Linz (Upper Austria) experienced a growth of the woollen cloth industry, while cotton-printing was expanded in Prague. But, the arrival of such proletarian textile workers in spinning or printing mills did not assure their immediate status. Indeed, the lingering imbalance associated with serfdom was quite evident in these facilities. Komlosy states that At the time industrial wage work in factories involved extreme exploitation. The new factory villages disregarded traditional social norms, and new conventions such as social bargaining, charity or workers’ organization had not yet developed. Mills were comparable to military barracks: they had a stranglehold over their employees from morning to evening.32 Serfdom was also a factor in the production of Russian textiles. In the eighteenth century, well before their emancipation, Russian serfs were either assigned work in textile factories or factory owners were allowed by a 1721 law to buy an entire village of serfs who would then be forced to migrate to the mill. Immanuel Wallerstein suggests that such serfs resented working in the factories, even though they had considerably more freedom than in agricultural settings.33 Such sentiments suggest deplorable working conditions, as well as coercion. For these reasons, I consider these serf/workers as falling within the same category as other non-voluntary migrants (for example African slaves to the Americas).

The Ottoman Empire Textile manufacture was extensive in the various culture areas which came under the jurisdiction of the Ottoman Empire during Stage I, and coarse and fine cloth was produced in settings which represented home production, peasant production for re-sale and more sophisticated, specialty textile centres. In this regard, the Ottoman pattern is similar to the European one, with many regions having extensive textile industries by the late eighteenth century (this would include the Balkan region of present-day Bulgaria, Romania and Greece).34 Such industries likely attracted localized migration of ethnically homogeneous workers but not necessarily skilled personnel. A different migration stream, in which skilled workers were enticed to relocate, was peculiar to certain regions; as early as the 1660s skilled weavers from neighbouring towns of Mardin and Diyarbakir were recruited to work in emerging textile centres like Aleppo. Aleppo alternated with Damascus for the top rank in textile production, and this migration to Aleppo provides a parallel with the coeval movement of skilled textile workers to Leiden and its competition for workers with 32 Ibid. 33 Immanuel Wallerstein, The Modern World-System, vol. III: The Second Era of Great Expansion of the Capitalist World-Economy, 1730–1840s (San Diego, 1989), p. 163. 34 Quataert, Chapter 18, this volume, ‘Ottoman Empire’. 655

The Ashgate Companion to the History of Textile Workers other northern cities in the Netherlands. By the late eighteenth century, however, guild restrictions on membership tended to halt this easy migration within the empire as a whole; it was not until the nineteenth century that the pace of migration was re-invigorated.35

Southern and eastern Asia The national overviews in this volume on Japan, China and India give clear indications that during Stage I in these countries as well, local textile industries conformed to the basic patterns noted above for Europe, relying on an ethnically homogeneous labour pool to meet production needs. Nonetheless, patterns of labour recruitment and composition were influenced by local conditions. In India, for example, there was widespread internal migration which functioned as a type of demographic safety valve: ‘A certain readiness to migrate was the most effective means the weavers possessed to deal with natural and manmade upheavals.’36 The proto-typical European pattern was additionally modified in India by the disproportionate involvement of village leaders or heads of caste in the recruitment of textile workers; this type of recruitment represents a very specialized form of chain migration. It is not surprising to learn that movement of people through the encouragement of local leaders was more of a factor in the south of the country, where the numbers of land holders was low. The entrance of the English East India Company into the country served to solidify this model, as the English continued to deal with the already established avenues of caste and village leadership to recruit workers for textile industries. But, in the late eighteenth century, concomitant with Britain’s expanding commercial/political role, a dramatic change occurred as contractors engaged to ‘recruit’ textile workers both in Bengal and the Coromandel were now paid bounties by colonial authorities. This pattern suggests a very different level of organization, as well as an increase in long-distance migration.37 Under these circumstances textile production was rife with coercion and abuse. To the extent that the process was driven by ethnic immigrants (that is the British), we can speak of an ethnic divide in which the owners of textile plants were foreigners, while the workforce remained ethnically homogeneous. In China, contemporaneously, an ethnically homogeneous population involved in textile production also conformed to European normative patterns. Until the nineteenth century the rural household was the locus for production; this was especially true of cotton textiles. But, at the same time, silk production spread into urban areas, where it was produced by wage labourers. To the extent that wage workers represent a different economic class from farmers there was internal differentiation in the labour force, but it was not along ethnic lines. Ethnic

35 Ibid. 36 Roy, Chapter 10, this volume, ‘India’. 37 Ibid. 656

How will we get our workers? distinctions (and discrimination) among textile workers would not be an important facet of the industry until the twentieth century.38 Finally, in Japan, during Stage I, cotton and silk cloth production by ethnically undifferentiated workers initially tended to be largely urban-based, although, contrary to the Chinese experience, rural families also produced silk. Migration was limited, as workers tended not to relocate once they had found their niche (as, for example, the silk weavers of Kyoto). However, by the eighteenth century this pattern had been reversed, with ‘textile production shifted increasingly to rural areas. Guild restrictions in the towns hindered the development of new technologies and raised the costs of production, so new producers seeking to satisfy the growing demand located their operations in rural areas where labour was cheaper.’39

Cape Verde and the Americas Up to this point, very little differentiation in worker/recruitment patterns or textile labour force composition has been observed. However, the picture changes dramatically when Africa and especially the Americas are considered for the same time frame. Even though cloth was produced in home settings in both these regions during Stage I, much cloth was produced simultaneously in commercial (proto-industrial) workshops, where the labour force was largely composed of 1) coerced, 2) ethnically differentiated and 3) non-wage-earning labourers who were relocated to these regions often from great distances. These nonvoluntary, non-paid workers are clearly distinguishable from the ethnically homogenous, peasant-turned-wage-earner textile producer who travelled short distances to contemporaneous European proto-industrial settings. Indeed, when American patterns of ethnicity and migration are contrasted with the European proto-industrialized mode of production, the resemblance is only superficial. A primary example of commercialized textile production dependent on coerced labour occurred in the Cape Verde Islands. It was in these Atlantic islands that the renowned and widely sought-after barafula fabrics were created, ranging from plain white to elaborate striped and patterned cloths. The barafula textiles, however, were not produced by freemen who chose to work in the industry in order to supplement income, but rather by non-paid slaves who had no option but to labour all day in workshops established on Portuguese-owned plantations. These planters had initially concentrated resources on the growing of raw cotton for European markets, but later switched gears opting for the manufacture and export of cotton cloth; the latter quickly supplanted raw product as the region’s principal export commodity. Slaves were transformed into spinners (mostly women) and weavers (mostly men)40 and required to spend long hours in workshops on the plantations 38 Cliver, Chapter 5, this volume, ‘China’. 39 Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’. 40 T. Bentley Duncan, Atlantic Islands: Madeira, the Azores, and the Cape Verdes in SeventeenthCentury Commerce and Navigation (Chicago, 1972), p. 219. 657

The Ashgate Companion to the History of Textile Workers turning out the desired cloth. Such assignments were the primary responsibility of these slaves, rather than agricultural work. Ultimately, the output of these bondsmen was so successful that cloth became a widely used proxy for currency, except in those situations when the white population still engaged in commercial transactions with Europe.41 Yet, ironically, even in that context, rolls of cloth became indicators of the relative success of the planter, as wealth (and influence) was calculated by ever-increasing stockpiles of finished fabrics awaiting exportation in warehouses. Indeed, the scale of this semi-rural production was such that by mid-seventeenth century the barafulas ‘drove out Indian cottons and European cloth’42 on the African coast where they simultaneously served as currency for purchasing other slaves.43 The Cape Verde Islands represent a textile-producing context which did not require migration (as the slaves were local to the islands), but did show evidence of a vast ethnic divide between the owners of the workshops and the labourers. Turning to the Americas, we find that the model of coerced worker/protoindustrialized textile production was even more entrenched. Of course, here too, cloth was produced at home by colonial settlers, who were sometimes aided by enslaved workers. But, with regard to proto-industrial organization, unlike the European model in which workers freely chose to either seasonally or permanently devote themselves to production of cloth, in the Americas (North and South) it was slaves, either African or Creole (born in the Americas) bondsmen who were forced to produce textiles on a macro scale. Cloth was also produced in proto-industrialized workshops by semi-free labourers (that is indigenes) who were, nevertheless, under obligation to work full time for Europeans (see below). The breadth of this textile production was quite impressive; it included not only cottons but wool, and some silk as well. Even while cloth continued to be imported from the metropole, the chapters on Brazil (which cites thousands of yards of fabric produced in slave workshops),44 Argentina45 and Mexico46 clearly demonstrate that considerable amounts of cloth were produced commercially in the Americas, for American consumption. This held true whether production was in the form of home preparation47 or took place in proto-industrial textile workshops by non-free labourers. With regard to migration issues, obviously all of these textile producers (with the exception of the indigenes) were initially international or external migrants 41 Ibid., p. 223. 42 Ibid., p. 220. 43 Ibid. The Azores also had a textile industry of substantial proportions. Linen produced there was sold to European markets, including England (p. 110). 44 Delson, Chapter 4, this volume, ‘Brazil’. One workshop in Rio de Janeiro in 1785 produced over 50,000 yards of fabric in a single year; while using slave labour the single province of Minas Gerais for the year 1827–1828 yielded an astounding figure of more than 8 million yards of fabric for local use with the rest sold throughout Brazil. 45 Lobato, Chapter 2, this volume, ‘Argentina’. 46 Bortz, Chapter 13, this volume, ‘Mexico’. 47 On this point, see Blewett, Chapter 21, this volume, ‘USA’. 658

How will we get our workers? from Europe or Africa who, over time, became naturalized or ‘creolized’. For the purposes of this discussion, however, I have considered African and Creole slaves heuristically as ‘migrants’ (albeit on a non-voluntary basis) largely because their numbers were constantly being replenished. On the other hand, their ethnically distinct owners may be considered as non-migrants, since white settlers tended to become permanent residents in a given colony. The divide between the ethnic groups was enormous even though in some settings (Brazil, for example) slave textile workers laboured alongside their masters as a matter of course.48 It is important to remember that textile workers who were enslaved (no matter where they were born) received no remuneration for work, and were assigned production tasks in a process controlled exclusively by their owners. Although some African slaves brought to the Americas may have also been textile artisans, there is no evidence to suggest that skilled African slaves were specifically brought to the Americas for the purpose of working in textile production. Rather, African and Creole slaves were assigned to textile production for the market because they constituted the backbone of the available workforce. Moreover, this phenomenon did not disappear with independence, even though the institution of slavery itself is largely associated with colonialism. Thus, even in the nineteenth century, in those former colonies which perpetuated the slave system, bondsmen could be found working in mechanized textile mills. In Brazil, this meant that slaves continued to work in textile mills until the end of slavery in 188849 while in the US South, slaves produced textile on plantations long after political ties to Britain had been severed.50 In the southern colonies of the United States, textile production was women’s work; wherever cotton was grown or wool sheared, slave women were expected to work on textile production in the evenings, after their field work had ended. Jacqueline Jones reports that there were ‘rigid spinning and weaving quotas on women who worked in the fields all day’,51 and further finds that many workers were required to spin about three hundred yards of thread a night, and five times that amount on rainy days (that is when no field work could be done)! With the click of the weaving loom heard all evening, the output must have been impressive. What is important to underscore here is that first, there was no choice of assignment and second, the production quotas were high enough to suggest that the local needs of the plantation were exceeded (in turn, suggesting re-sale in the market). Protoindustrial workshops appeared about the time of the war of independence against Britain, when obtaining imported cloth became problematic. One South Carolina planter, for example, had 30 slaves from his plantation constantly producing cloth. By contrast, in the northern colonies cloth was primarily produced in homes or 48 Delson, ‘Brazil’. 49 Stanley Stein, The Brazilian Cotton Manufacture: Textile Enterprise in an Underdeveloped Area, 1850–1950 (Cambridge, Mass., 1957), p. 51. 50 Jacqueline Jones, Labor of Love, Labor of Sorrow: Black Women, Work, and the Family from Slavery to the Present (New York, 1985), pp. 30–31. 51 Ibid., p. 30. 659

The Ashgate Companion to the History of Textile Workers alternatively imported, until the organization of textile mills in the late eighteenth century began to make up for importation losses engendered by the war.52 One imagines that where cotton was grown in the British Caribbean (for example the Leewards)53 similar quotas for slave production existed, although documentation is scarce. In Mexico, cloth production was initially assigned as a task to indigenous workers, and later to slaves of African origin. These slaves, who were assigned to obrajes (workshops), accounted for as much as 40 per cent of the textile labour force in some parts of the colony, a condition that prevailed as late as 1800 (before slavery was abolished, but while Mexico was still under Spanish control).54 In sum, non-wage-earning, enslaved labourers in the Americas were allocated to textile production on plantations, in proto-industrial workshops (sometimes located on the plantations) and later in factories.55 Unfortunately, we do not know empirically if the fact that Mexican and Brazilian slaves produced cloth for their masters in obrajes, or later in factories, resulted in any social mobility or elevated their standing. Certainly in plantation settings where slaves may have worked alongside their mistresses (in Minas Gerais, Brazil, for example), this proximity may have somewhat relaxed the fundamental hegemonic order. However, where slaves produced textiles isolated from their owners (as in the southern United States), one would not expect to see an amelioration of relationships. Again, the data are missing to either confirm or deny the possibility that there was more clout attached to workshop assignments. It wasn’t just African or Creole slaves who constituted a non-voluntary textile labour force for Europeans. The papers on Mexico, Argentina and Brazil indicate that indigenous workers (especially those who had expertise in spinning and weaving) were also pressed into service producing textiles. As we do not know the extent to which such workers were geographically shifted in order to be available as a labour force it is somewhat of a conundrum deciding whether or not they should be considered migrant workers. Technically, native workers were not chattel slaves, but were rather ‘allotted’ to Europeans exclusively for the purpose of rendering work. Such workers were, therefore, not ‘owned’ as property by the Europeans, although they were nonetheless assigned specific work tasks via a system known as encomienda (or entrustment). In its most benign rendering, the encomienda system was conceptualized as a reciprocal arrangement; for example, Spaniards would get the ‘gift’ of Indian labour for a certain specified period of time in return for instructing labourers in European values and religion. In reality, no matter how well-intentioned the system might 52 The Carolina slaves were instructed by white spinners and weavers in the art of producing cloth. See Philip D. Morgan, Slave Counterpoint: Black Culture in the Eighteenthcentury Chesapeake and Lowcountry (Chapel Hill, NC, 1998), p. 248. On the organization of mills in the northern United States, see Blewett, ‘USA’. 53 Douglas Hall, Five of the Leewards, 1834–1870 (Barbados, 1971), p. 4. 54 Bortz, ‘Mexico’. 55 Delson, ‘Brazil’. 660

How will we get our workers? have been, the encomienda system was rife with abuses, the indigenes had little say over their own destinies, and conditions of work frequently approached outright slavery. In Argentina, skilled textile-producing indigenes made ‘coarse woollen clothes …, linen, stockings, blankets and hats’ for their encomenderos.56 Eventually, this system evolved into a more commercialized pattern in which native labourers were supplied by encomenderos to work in obrajes, which were, in turn, overseen by Spanish migrant craftsmen who had been specifically recruited for the position.57 Assuming that indigenous workers had any status at all as textile producers, in the obraje system they would automatically be considered as beneath foreign craftsmen in status and, consequently, subject to their will. Such obrajes were also prevalent in Mexico. As noted above, in the 1600s, indigenous obraje workers were largely replaced by African slave labourers. But whether they were indigenes or slaves, most workers considered the obrajes which produced cotton and wool cloth, particularly unpleasant places … Work often began at 4 a.m., always under the direction of supervisors … During his visit to Mexico in 1803, von Humboldt visited some obrajes, noting ‘the unhealthiness of the situation and the bad treatment to which the workmen are exposed … all appear half naked, covered with rags, meagre and deformed. Every workshop resembles a dark prison’.58 Here again, such individuals were clearly at the lowest rungs of the labour hierarchy. In Brazil textiles were also produced on plantations and in workshops, a pattern which intensified in the mid-to-late eighteenth century when Indian workers were resettled into textile-producing villages. Their primary task was to produce the cloth necessary to supply uniforms for the large number of Portuguese military residents in the colony, a feat which was accomplished by mandatory labour in the village workshops for the entire day. Although technically such Indian workers received wages, these ‘salaries’ were computed in measures of cloth, which in turn began to function as a kind of local currency (a feature reminiscent of the Cape Verdes). Technically such ‘wage’ labourers were free, theoretically having entered into the working relationship of their own accord. But, short of running away, they had little say over the work assignments given them by Portuguese administrators. As I pointed out in Chapter 4 on the Brazilian national textile industry, the Portuguese envisioned such enterprise as a way of ‘acculturating’ the Indians and introducing them to European material standards.59 In this context we know that a hierarchy of workers was created: labourers were not only divided by task but by salary differentials, with the master spinner who supervised the workforce receiving the 56 Lobato, ‘Argentina’. 57 Ibid. 58 Bortz, ‘Mexico’. 59 Delson, ‘Brazil’. 661

The Ashgate Companion to the History of Textile Workers highest salary.60 The system created clear managerial/status categories. I consider such Indian workers as non-voluntary migrant labourers, as they were forcibly removed from their native areas and relocated to villages where the textiles were produced. It is also worthwhile mentioning in this discussion of the Americas that the Portuguese planned to transfer skilled textile printers (‘calico’-makers) from the Coromandel Coast of India to Brazil.61 Although this migration was eventually judged infeasible, it is fascinating to observe that had it taken place the Portuguese were quite prepared to receive such ‘recruits’ as a cultural enclave of textile workers. The various configurations of textile production evident in the Cape Verde Islands, the southern United States, Brazil, Mexico and Argentina clearly do not conform to the ‘classic’ European modalities indicated for Stage I. In these African/ American proto-industrialized workshops (and even on the plantations) there is no evidence of an ethnically homogenous workforce composed of local wage-earners; indeed it is the absence of wages, as well as the coercion of long-distance, ethnically diverse migrants into the production of textiles which defines its parameters and distinguishes it so completely from its European counterpart.

Stage II: The modern period Major shifts in the ethnic and migration patterns of textile workers occurred in conjunction with the Age of Industrialization but are largely associated with non-European locales. Within Europe, migration of workers continued to be largely internal and short distance (with exceptions noted below). Moreover, the pattern of seasonal migration of workers to textile facilities did not entirely disappear until the full mechanization of textile industries had occurred (by the mid 1800s). However, European workers also migrated in enormous numbers to the former European colonies of Latin America, North America and Australia. These longdistance migrations have been linked to geographic differences in the supply and demand for labour and, more recently, by world system theory which posits the notion that ‘migration is a natural outgrowth of disruptions and dislocations that inevitably occur in the process of capitalist development’.62 But such a model is limited in its usefulness by virtue of the fact that capitalist dislocations cannot 60 Colin M. Maclachlan, ‘The Indian Labor Structure in the Portuguese Amazon, 1700–1800’, in Dauril Alden (ed.), Colonial Roots of Modern Brazil (Berkeley, 1973), pp. 199–230, at pp. 219–220. 61 Delson, ‘Brazil’. 62 Douglas S. Massey, Joaquín Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino and J. Edward Taylor, ‘Theories of International Migration: A Review and Appraisal’, in Robin Cohen (ed.), Theories of Migration (Cheltenham, 1996), pp. 155–187 at pp. 159, 183. 662

How will we get our workers? explain other major causes of migration in the nineteenth century including political repression (for example the pogroms in Russia) or natural disasters (such as the Irish Potato Famine of 1845). Similarly, ‘network’ (or ‘chain’) migration, in which recruitment is based on contacts within the industry, is not fully encompassed by world system theory.63 This section of the chapter examines specific examples of each of these migration patterns, and focuses on the ethnic composition of the workers by region.

Internal migration Internal migration of textile workers increased in nineteenth-century Europe as textile factories were mechanized and required a large, permanent labour force. In part, this need was fulfilled by recruitment from the ranks of workers who had heretofore produced in either home or proto-industrial settings. In Germany, qualified workers in the factories were primarily recruited from the craft workers whose skills were still valued even though, in general, shop foremen replaced these highly skilled master weavers of earlier times as the highest status workers.64 Much of the migration to urban textile factories in Europe tended to be drawn from surrounding regions. In the case of Roubaix, France, for example, both French and Belgians came to work in the woollen mills from distances rarely exceeding 20 kilometres from the city’s boundaries. At the same time the older pattern of home production continued without interruption, with peasants and artisans engaged in weaving in the surrounding rural villages. Even seasonal migration to centralized textile mills remained an option until the 1860s when the expansion of mechanical looms in that city ‘seriously encroached’ upon rural production.65 Wendy Gordon’s study of the textile mills of Preston documents a similar pattern for Lancashire, England. In Preston, a steady stream of unskilled migrants from surrounding rural areas supplied labour to the cotton mills of that city, with most of those workers being recruited from rural zones within a radius of less than 30 miles from the city;66 a much smaller number of workers chose to relocate from other, nearby, urban areas. Fully a quarter of those short-distance migrants were young, single females (the subject of her study), who migrated (as did young men) for essentially economic reasons. However, Gordon emphasizes that these migration patterns were initiated in response primarily to the needs of the mill employers, rather than the reverse. In other words, relocation was precipitated by 63 Ibid., p. 198. 64 Ebeling et al., Chapter 8, this volume, ‘Germany’. 65 Moch, Moving Europeans, p. 133. Similarly, in the eastern sector of the Kreis region of Westphalia, workers concentrated on linen weaving in the wintertime, when there was little agricultural work to be done. See Jan Lucassen, Migrant Labor in Europe, 1600–1900 (London, 1987), p. 31. 66 Wendy M. Gordon, Mill Girls and Strangers: Single Women’s Independent Migration in England, Scotland, and the United States, 1850–1881 (Albany, NY, 2002), p. 23. 663

The Ashgate Companion to the History of Textile Workers employer demand, rather than employee desire. Many young women who worked in the mills were not necessarily indigent but had come to view such employment as an intermediate stage between their childhood home and marriage. With regard to textile production in Paisley, Scotland, another region of her study, Gordon suggests that the pattern of worker recruitment represents a more cyclical migration. Somewhat atypically, the first stage of migration to Paisley saw workers moving within rural areas and then to the herring fisheries. This process culminated in relocation to Paisley itself, in what Gordon believes was a logical progression to the mills.67 Another pattern saw single women passing from domestic service to work in mills which produced shawls in a variety of fabrics (cottons, wools and silks) and had employment for thousands of workers of varying skills. Those who migrated towards work in these factories tended to use the railway to facilitate their moves. Finally, a much smaller proportion of workers emigrated from Ireland (especially after the potato blight and famine of 1845).68 Stepped-up internal migration to textile mills is also characteristic of nineteenthcentury Indian textile industries. In Chapter 10 Roy explains that there were two broad streams of internal migration, from north to east, and from south to west; clearly, poverty in rural areas was a factor, with migrants coming from both peasant and artisan backgrounds.69 Roy suggests also that this migration was ‘segmented’, by which he means that migrants of specific background moved into specific occupations.70 He contends that poverty was not the only factor leading to migration to textile plants; those who relocated were not necessarily the poorest individuals, but rather the people who owned their own land in villages. Paralleling a situation which we have already seen in Europe during Stage I, these individuals, usually men, left their families at home in the villages while they went to Bombay to work in mills. To the extent that the entire family did not relocate, the author says that migration may be seen as ‘not a total break from, nor a clear alternative to, rural work’.71 Notwithstanding, the transition from peasant to urban labourer could not have been an easy one and those who migrated to cities like Calcutta or Bombay ‘faced very different work, workplace, habitat and social milieu’.72 Towards the end of the century, women increasingly followed their husbands into work in the textile mills, doing low level jobs associated with hand-weaving, such as winding and sizing. While it would appear that the promise of a higher wage was the deciding factor in driving people to migrate to urban textile works, the fact that merely living in the city could eliminate the constant threat of famine which faced farmers, as well as ease the restrictions imposed by caste, tends to undermine the primacy of wages as a factor for migration.73 67 Ibid., p. 102. 68 Gordon, Mill Girls, p. 106. 69 Roy, ‘India’. 70 Ibid. 71 Ibid. 72 Ibid. 73 Ibid. 664

How will we get our workers? A parallel, large-scale internal migration characterized nineteenth-century textile production in the Ottoman Empire, where merchants acted as ‘major agents of recruitment’ in rural as well as urban areas. To fulfil their quotas, such recruiters appear to have relied on imams, priests and rabbis, who personally knew the desired workers and could persuade them of the propriety of entering factories.74 Quataert writes that ‘the religiously homogeneous nature of the carpet-making workforce in a particular region suggests appeals made either through religious authorities or in the name of that religion’.75 What resulted from these recruitment efforts were ethnically/religiously undifferentiated enclaves, which may have produced specialty textile items. In Yemen, for example, woven linen belts worn by imams and the Muslim aristocracy were supplied by a monopoly of Jewish weavers.76 On the other hand, in Egypt, Muslims, Copts and Jews were all engaged in silk weaving in Cairo and it appears that there was no differentiation among these workers who were further organized into guilds.77

A mix of internal migration and external/international migration (skilled and unskilled workers) In the Americas, both internal and external migration patterns were in evidence, with foreign labourers often working alongside local recruits. The international migrants, representing skilled and unskilled labourers, came from England, Ireland, Spain and Italy, Central and Eastern Europe to find work in the textile enterprises of the northern United States, Brazil, Argentina and Uruguay. The significance of unskilled workers in the English textile industries (especially the Irish, post-1845)78 is well documented, but they were equally important in the textile industries of the northern United States,79 where foreign workers competed for positions with unskilled workers emanating from the northern states of Vermont, New Hampshire, New York and Maine. Over time the composition of the textile labour force shifted in favour of immigrants. According to Gordon, up to the 1850s, the labourers who were recruited to the important textile-producing Massachusetts city of Lowell were primarily natives of nearby states. However, beginning with the post-Potato Famine period (post-1845), increasing numbers of Irish women came to Lowell, followed by French Canadians and later immigrants from Central and Eastern Europe.80 The pattern of foreign immigration intensified after the US Civil 74 Quataert, ‘Ottoman Empire’. 75 Ibid. 76 A. Klein, ‘Tablet Weaving by the Jews of San’a (Yemen)’, in Cordwell and Schwartz (eds), The Fabrics of Culture, pp. 425–445, at p. 427. 77 Beinin, Chapter 7, this volume, ‘Egypt’. 78 See, for example, Gordon, Mill Girls. 79 Blewett, ‘USA’. 80 Gordon, Mill Girls, p. 60. 665

The Ashgate Companion to the History of Textile Workers War, with the result that recruiters, who were paid bounties by mill owners, had to struggle to entice American workers back to the mills. What with falling wages and a quickened work pace associated with sophisticated technology, Yankee workers were quite reluctant to return to textile work; recruiters were scarcely able to disguise the reality of immigrant textile workers steadily taking jobs at wages that Yankees would not.81 Indeed, the immigrant population not only swelled the ranks of workers in the mills but changed the nature of the city from a predominantly Yankee Protestant city to a Catholic one. Language differences, as well as religious diversity, tended to separate textile workers along ethnic lines. For most of these workers, native-born and foreign, the primary motivation for relocation was wages, but not always. As Gordon points out, the initial Yankee textile workers were women from relatively prosperous families for whom work in the mills served (as it did in Preston, England) as an intermediate step between the home where one grew up and marrying.82 The housing of female workers in dormitories in Lowell was thought to provide a safe environment for a woman to live apart from her family. By contrast, for the legions of Irish workers arriving post-Famine disaster, the motivating factor for migration was simply avoiding starvation. In the southern United States, however, this mixed worker pattern was not characteristic. Mary Blewett explains that instead, in the late nineteenth century, southern textile mill workers tended to be recruited from rural farm families.83 This was a result of shaky growth in agricultural commodity markets during which time rural families (including children), accustomed to long hours of work, had come to view work in the textile mills as a reasonable alternative to the farm. Ironically, African-American workers, the descendants of the slaves who had for generations produced textiles for southern plantation owners, were denied production jobs in the cotton mills, although menial work (such as dyeing and sweeping) could be obtained. The social distance between the white textile worker and the black hired hand was substantial. Yet another mixed ethnic/migration pattern is provided by the example of the silk mills in Paterson, New Jersey. Prior to the Civil War, these mills were staffed by skilled English immigrants who arrived in New Jersey fresh from stints in Britain’s silk industry. They were, in turn, followed by highly skilled Dutch, German and Swiss textile workers who saw in Paterson the opportunity to advance to upper level management (and potential ownership). Nonetheless, even as late as 1880, immigrants accounted for only one-third of the total workforce.84 Almost half of Paterson’s silk workers at this point were unskilled labourers born in New Jersey, while the rest came from other parts of the USA. 81 Ibid., pp. 86–87. 82 Ibid., p. 61. 83 Blewett, ‘USA’. 84 Richard D. Margrave, ‘Technology Diffusion and the Transfer of Skills: NineteenthCentury English Silk Migration to Paterson’, in Philip B. Scranton (ed.), Silk City: Studies on the Paterson Silk Industry, 1860–1940 (Newark, 1985), pp. 9–34, at p. 26. 666

How will we get our workers? The textile industries of nineteenth-century Latin America also employed a mix of labourers ranging from foreign skilled and non-skilled workers, to labourers who migrated internally to urban mills. (As noted above, some mills continued to use enslaved labourers.) In Brazil, foreign labourers were contracted specifically to work for a set number of years in urban textile factories. Such immigrants frequently arrived as a ‘package’, accompanied by the machinery required to start the enterprise. Their skills were to be imparted to rural Brazilians, who were recruited from the countryside to form the bulk of the unskilled labour pool. English textile workers were soon joined by workers from other nations which had ‘advanced’ textile industries. A typical example is provided by the textile factory at Santo Aleixo in Rio de Janeiro which in 1851 was staffed by ‘seventeen Brazilians (fifteen men, two women) five Italians (three men and two women), two Englishmen, two Americans and eighty-three Germans (forty-three men and forty women)’.85 While in some limited instances (for example, in northern Alagoas) immigrants worked alongside locals, it was more often the case that immigrants assumed managerial roles; Stanley Stein suggests that the underlying motive to their employment was the specific charge to handle lesser personnel. Language difficulties and wage differentials must have created friction but, unfortunately, we have no first-hand accounts. By the time of the fiscal crisis of 1891, recruitment of British workers specifically for the mills had ended, largely because local wages were so low as to discourage emigration. In their stead came large numbers of poor, and largely unskilled, Italian, Spanish and German immigrants who poured into Brazil (and Argentina) in the late 1890s; some were recruited to work in textile mills upon arrival at the docks. Coming from poorer backgrounds, and many speaking languages related to Portuguese, such immigrants were considered far more malleable than the English had been (and would thus be more accepting of lower wages). Afro-Brazilians, who as slaves had formed an integral part of Brazil’s early textile labour force, worked in large number of textile factories until Emancipation (1888) when they were precluded from working in the mills. This represents a remarkably parallel situation to the southern United States, where former slaves were also excluded from textile work.86 The pattern of an ethnically mixed labour force, as well as the pivotal role played by skilled European textile workers is also characteristic of early nineteenth-century Egypt, where textile mill owners recruited British technicians to jumpstart their industries. As in Brazil, such workers often arrived with the requisite machines and advisors. Beinin suggests that c.1816–1818 the textile workshop in the Khurunfish neighbourhood of Cairo sought exactly this sort of outside expertise. Later, skilled Tunisian women were recruited to teach spinning and weaving techniques. Such skilled workers, especially the Europeans, received more money than the Egyptians. 85 Stein, Brazilian Cotton Manufacture, pp. 51–52. 86 Albert Fishlow, ‘Origins and Consequences of Import Substitution in Brazil’, in Luis DiMarco (ed.), International Economics and Development (San Diego, 1972), pp. 311–339 suggests that slaves did not enter the ‘modern’ sector. 667

The Ashgate Companion to the History of Textile Workers For the latter, harsh discipline, as well as requirements of army service, suggest a corrosive atmosphere that depended on coercion for its success.87

Paternalistic migration Another textile labour recruitment configuration of the nineteenth century is best termed ‘paternalistic migration’. In order to attract migrants from rural areas to their factories, textile owners held out the promise of housing and other inducements, including food and medicine. One of the earliest experiments in the creation of ‘utopian socialist’ mill towns is, of course, famously associated with Robert Owen and his mill at New Lanark, Scotland,88 but variations on his scheme quickly spread to other areas of the world, albeit not all with the same concern for the worker’s benefit. In Spain, for example, company towns were constructed in rural regions outside the major textile centre of Barcelona; this had the effect of not only supplying mill owners with workers willing to take lower wages, but allowing them to claim as fiscal exemptions any additional expenses, such as those incurred in the building of churches and shops. Such rural textile mill towns were characterized by a mixture of industrial and peasant worker populations, a condition which worked primarily to the industrialist’s advantage.89 Paternalistic migration also encouraged workers to migrate to mill dormitories in pre-Civil War Lowell, Massachusetts, mid-nineteenth-century Brazil and late nineteenth-century Mexico. In the 1850s, the Todos os Santos textile mill of Bahia, Brazil, provided not only remuneration to labourers but a school where they could learn to read, make music and sew.90 Such enlightened dormitory arrangements eventually gave way to more management-profitable vilas operarias, in which the textile company could charge rent, rather than provide housing for its workers. Similar paternalistic impulses initially characterized textile production in nineteenth-century Mexico where huge textile enterprises created mill towns; by the late nineteenth century, however, mill owners so tended to dominate the social landscape that they could ‘dictate the laws, make them obeyed, punish the transgressors, judge delinquents’.91 The Japanese textile industry of the nineteenth century was also characterized by paternalistic migration. As textile production in urban areas grew increasingly expensive, a phenomenon largely attributable to restrictions imposed by guilds, the emphasis shifted to rural zones. Reversing the usual rural–urban migration, rural workers were drawn to newly created rural textile mills by the promise of a unique set of interpersonal relationships. ‘Employment relations in commercial and 87 Beinin, ‘Egypt’. 88 See William Ashworth, The Genesis of Modern British Town Planning: A Study in Economic and Social History of the Nineteenth and Twentieth Centuries (London, 1954), pp. 120–121. 89 Smith et al., ‘Spain’. 90 Stein, Brazilian Cotton Manufacture, p. 57. 91 Bortz, ‘Mexico’. 668

How will we get our workers? manufacturing enterprises in this period tended to be categorized by the principle of oyabun-kobun, a fictive parent–child relationship established for mutual economic benefit. The oyabun, the parent figure, was responsible for the welfare and guidance of the kobun, who in return had a duty of obligation and obedience.’92

Absentee ownership Yet another characteristic of nineteenth-century textile industries was absentee or foreign ownership of mills, which produced a class of owners ethnically distinct from their workers. Usually such foreign owners relied on fellow countrymen (also migrants) as the ‘go-betweens’ (in their roles as upper-level managers) to a predominantly local and ethnically homogeneous labour force. While this configuration has already been seen for eighteenth-century India, and is largely associated with colonialism, the pattern also figures largely in the economic landscape of post-Independence Latin America. In Mexico, for example, French and Spanish mill owners dominated during the period of the Porfiriato (1876–1910), when the dictator Porifírio Díaz courted foreign capital. These French textile entrepreneurs developed some of the country’s largest mills,93 only to take home with them their profits, which were later ploughed into elegant mansions constructed in Haute-Provence.94

Network migration (chain migration) A final facet of textile labour recruitment in the nineteenth century was the phenomenon of ‘network migration’. Network, or ‘chain migration’, theory holds that ‘migrant networks are sets of interpersonal ties that connect migrants, former migrants and nonmigrants in origin and destination areas through ties of kinship, friendship and shared community origin’.95 This type of networking was typical of the Ottoman Empire in the nineteenth century. Donald Quataert argues that ‘The lack of a particular gender or ethnic profile of the labour force speaks to labour recruitment issues. Overall it seems that some kind of personal familiarity – either through family, acquaintance or place of origin – was a crucial aspect of recruitment.’ Eventually this type of chain recruitment might lead to the creation of ethnic enclaves; in the embroidery works around Istanbul, as well as in the spinning mills of Salonica, Jewish girls comprised 80 per cent of the workforce, while primarily ethnic Greek girls worked at Niausta and Bulgarian girls laboured at Wodena. Apparently the presence of a relative or friend was required to act as 92 Hunter and Macnaughtan, ‘Japan’. 93 Bortz, ‘Mexico’. 94 Laurence Martin, ‘Retour du Mexique’, Maisons & Decors Méditerranée 183 (Aout–Sept., 2004): 37–40. 95 Massey et al., ‘Theories’, p. 198. 669

The Ashgate Companion to the History of Textile Workers an intermediary between the potential worker and the merchant. ‘For girls and women, the relative or friend likely helped to teach the craft as well.’ Quataert further suggests that In some towns and cities, chain migration played an important role in labour recruitment. In one study of an Istanbul neighbourhood, workers in a particular shop most often hailed from the same village in Anatolia. More generally, certain occupations in the capital city were the monopoly of migrants from particular towns and villages in Anatolia and the Balkans.96 Of course, having contacts in the Americas also played its role in stimulating the mass migrations of thousands of Europeans to the mills of North and South America.

Overview of the twentieth century to the present This period has been marked by the continuation of internal migration of workers to textile facilities on a permanent basis and, as a corollary, an end to the pattern of seasonal migration to textile mills (with the exceptions noted below). It has also been marked in some countries by relocation of mills to the countryside, by the waning of paternalistic migration (especially in the face of unionization) and the lessening of chain or network migration (except in the case of ‘guest-worker’ programmes). The exodus of both skilled and unskilled European textile labourers to the Americas, and the accompanying mixed textile labour pools in the United States, Argentina and Brazil yielded to the process of naturalization, as long-distance textile worker migrants and their descendants became citizens. Finally, there have been expressions of nationalism which have appeared in specific configurations of textile worker employment as well as factory owners.

The end of paternalistic migration Paternalistic migration, which has been shown to have been such an important force in the recruitment of textile workers in the nineteenth century collapsed in the twentieth, the result of unionization of workers and/or cataclysmic events. In Mexico, for example, prior to the outbreak of civil war in 1910, network migration had resulted in several generations of family members working for the same textile mill, and the relationship of worker to owner was marked by paternalism and clientalism. The pattern changed irrevocably after the Revolution of 1910 when workers were empowered by newly formed unions, and could now make demands in large numbers through union leaders, rather than individually. As a direct consequence of union activity, rather than the paternalism of the boss, Mexican 96 Quataert, ‘Ottoman Empire’. 670

How will we get our workers? textile workers were now rewarded with housing, education, medical care and tenure on the job on their own terms.97 Unfortunately, the benefits did not outweigh sagging wages in the second half of the century. More recently Mexican textile production has shifted in an ominous direction, with the passage of NAFTA (North American Free Trade Agreement) in the 1990s. Under the terms of the new agreement it became legal for the firms and managers (largely from the United States) who took over Mexican factories to hire non-unionized labour. Locating their factories close to the Mexican–US border, the new factory owners recruited poor workers from all over the country who migrated to low-paying, low status jobs. Bortz suggests that as a result the once proud textile workers have now become ‘the most unfortunate class’.98 In Chile, the nineteenth-century pattern of foreign mill ownership remained a prominent feature of the textile industry until well into the 1950s, even as the paternalistic system of labour recruitment declined. At the vast Yarur mills in the capital city of Santiago, the Palestinian-born founder of the factory (in the 1920s) eschewed urban labourers in that city as troublesome and sympathetic to unionizing rhetoric, preferring instead to hire peasant workers who migrated to the capital from the south. He was especially sympathetic to Mapuche Indian workers, whom he believed would be more docile than the typical Santiago workers. To bridge the distance between himself and his migrant workforce, Don Juan Yarur, who was known to have an eye for pretty workers, would dole out favours to his workers if they cooperated with him. However, this thinly disguised paternalistic system waned as even indigene workers at the Yarur mills became unionized. With Yarur’s celebrated beneficence a thing of the past, and Don Juan’s descendents extremely wary of the unions, relations between management and workers became testy. It is no wonder that these textile workers were at the forefront of Chile’s Revolution of 1971.99 Elsewhere, unionization not only ended paternalism but created further gaps between textile workers who now split along ethnic lines. Adding tension to the already firmly drawn lines between immigrant and native-born labourers, European immigrants in the United States joined unions en masse. In Lowell, Massachusetts, Italian, French-Canadian and later southern and Eastern European immigrants swelled the ranks of strikers in the 1912 textile strike,100 but in Paterson, New Jersey, the immigrants split into two opposing camps, with German and Italians viewed as anarchists while the opposing socialist factions were associated primarily with Jewish broad-silk weavers. Many of the latter had fled Łódź, Poland after an unsuccessful series of strikes and retaliatory pogroms in 1905–1906.101 97 Bortz, ‘Mexico’. 98 Ibid. 99 Peter Winn, Weavers of Revolution: The Yarur Workers and Chile’s Road to Socialism (New York, 1986), p. 35. 100 Blewett, ‘USA’. 101 David J. Goldberg, ‘The Battle for Labor Supremacy in Paterson, 1916–1922’, in Scranton (ed.) Silk City, pp. 107–134, at p. 109. See also Jane Wallerstein, Voices of America: Voices 671

The Ashgate Companion to the History of Textile Workers The phasing out of network migration In Europe, network migration was perpetuated in some locales (for example, in the Netherlands as late as post-Second World War, where you needed an introduction by a family member to get a job in some mills),102 while in others it waned, as personal contacts could not counter the increasing practice of blacklisting textile workers for union activity. This was especially the case in Catalonia, Spain where past participation in strikes would immediately discredit any applicant for a position. This condition was exacerbated by the fact that in Catalonia just two towns dominated the production of woollens (which by the late nineteenth century accounted for more than two-thirds of the entire Spanish national production); if a striker was known in one locale the mill owners of the other city were sure to find out. For this reason many ‘blacklisted’ textile workers from these towns changed their names and fled to America.103 Many such black-listed workers wound up in Argentina, where at the beginning of the twentieth century Italian- and French-owned textile companies recruited a steady supply of their fellow immigrants mainly from Italy and Spain. This type of chain migration was phased out, over time, as immigrant populations became naturalized. Figures from 1936 reveal that ‘60 per cent of the workers (men and women) were Argentinean, about 11 per cent were Spanish and Italian’ while the rest were Polish, German and Yugoslavian. As immigration slackened, and the next generation began to think of themselves as Argentineans, foreign workers vanished from the activity.104 In contrast with Argentina, the major textile factory La Industrial in neighbouring Uruguay operated initially with a predominantly local workforce. When occasionally recruited, foreign migrant workers tended to be highly skilled and either Catalonian or Italian. Often they were recruited by representatives from the plant when they first arrived in Montevideo or Buenos Aires. Such workers received higher wages than local Uruguayans, presumably on the basis of their skill.105

Internal migration Another trend in the twentieth century to the present is the continuation of internal migration of textile workers to production centres within their own countries; gravitation to centralized mills is especially prevalent in developing nations. from the Paterson Silk Mills (Charleston, SC, 2000). 102 Van Nederveen Meerkerk, Heerma van Voss and Els Hiemstra-Kuperus, ‘Netherlands’. 103 See the personal account of Baltasar Porcel, La Revuela Permanente (Barcelona, 1978). I am indebted to Carles Enrech for this reference. 104 Lobato, ‘Argentina’. 105 Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’. 672

How will we get our workers? Indian textile factories which still have a handloom component (in the western parts of the country) continue to attract large numbers of migrant workers. From 1900 to1950 a hierarchy of such workers featured the most skilled migrants at the top of the wage-earning scale, while local-based, rural artisans were at the bottom. More traditional textile centres such as Bombay and Ahmedabad acted as magnets for ‘long-distance migrants’.106 However, post-1985 to the present, this pattern has been reversed with the trend now being to set up factories in semi-industrialized areas, outside of unionized urban areas; in a rare instance of the continuation of seasonal migration, these factories continue to employ peasants from impoverished rural areas on a part-time basis.

Nationalism and textile labour recruitment In Czechoslovakia, concern with the ethnic composition of textile plant owners resulted in the creation of a more localized plant ownership profile in the 1920s. A programme of nationalization implemented in that period aimed to ‘replace German and Austrian owners with Czechs or at least with West Europeans’. But by the 1930s the tide had turned once again, and the 1920s replacement programme was a cause for antagonism with the increasingly powerful Nazi government in Germany. After the Anschluss the policy was reversed as the ‘major share of Czech textile regions were thus integrated into Germany which supplanted Jewish with German owners’.107 Other recruitment migration patterns of the twentieth century are represented by German ‘guest-worker’ programmes in which primarily Turkish worker enclaves staff textile and clothing mills. Such workers also have a presence in Paris (where they started out in auto or construction jobs before employment in textile factories), and did as well in Amsterdam until the government closed opportunities for such employment in the 1990s and the Turkish nationals resident there returned home.108 In China, the textile labour force has remained Chinese, but early twentieth century textile workers began to be ‘differentiated by gender, place of origin, value and skill’. Subei women, who spoke a different dialect, were looked down upon by their Shanghai and Jiangnan bosses as well as their co-workers. They were also frequently given the most arduous and dirtiest tasks. Apparently the discrimination was so intense that during the 1930s some Subei women preferred to work for Japanese bosses at their mills in China, although doing so resulted in further discrimination from other ethnic Chinese. One especially unpleasant feature of textile labour patterns in the twentieth century was the appearance of coerced migration in China. Before the Second World War, recruitment of rural female workers for textile industries was accomplished 106 Roy, India’. 107 Komlosy, ‘Habsburg Monarchy’. 108 Seidman, Chapter 19, this volume, ‘Turkey’. 673

The Ashgate Companion to the History of Textile Workers via contractors who ‘bought’ recruits from families, usually without the approval of the woman. As seen in earlier forms of this kind of non-voluntary migration, there was little chance of escaping this situation, and the arrangement was clearly non-voluntary. Although such women were paid nominal wages, a portion of their earning nonetheless was taken by the so-called ‘Number Ones’, boss-like figures who controlled the mill labour force. They also commanded such authority that they could mobilize thousands of workers, if needed, to oppose any mill owner’s effort to gain control of the hiring process.109 Finally, in Japan, internal migration of ethnic Japanese to textile mills has also been the rule (although forced Korean immigration to cotton factories was significant in the interwar period). Workers came from throughout the Japanese islands, but recruitment was focused in particular on specific areas such as the island of Kyūshū in the southwest and the Japan Sea coastal and inland central areas. Extensive internal migration led to cotton workers from the furthest reaches of the archipelago being concentrated in the urban areas around Ōsaka, Nagoya and Tōkyō, and silk reelers moving into the mountainous central areas from neighbouring prefectures. Patterns of migration were shaped by a range of factors, including the activities of individual recruiters and localized transport developments.110 Countering the disappearance of the pattern elsewhere, paternalistic policies have continued to be important in textile labour recruitment in Japan where ‘dormitories and education remained the two central strategies for managing young female workers’.111

Conclusions This discussion has examined the ethnic composition of textile labour forces and the recruitment of workers from the sixteenth century to the present. By focusing on the issues of migration and ethnicity, both synchronically and diachronically, the discussion has been able to challenge the presumed ubiquity of European modalities. Five observations can be summarized. 1) For the sixteenth to the eighteenth century most historians acknowledge three basic configurations of textile labourers involved in fabric production: workers were involved in rural/home production or they were involved in proto-industry, working from home to earn additional capital, or they migrated seasonally to emerging textile centres. There were additionally those textile centres (Leiden, for 109 Cliver, ‘China’. 110 Hunter and Macnaughtan, ‘Japan’. 111 Ibid. 674

How will we get our workers? example) that exerted a more modern ‘pull’ resulting in a ‘network’ of employees, in which the desire to secure skilled labour was fundamental. In some instances, religious and political migrants were also absorbed into major textile cities, albeit at the lowest income levels. All three of these patterns may be found coevally and globally, and have been noted for the Ottoman Empire, India, China and Japan. Everywhere, these patterns would eventually be supplanted by mechanized production in mills. However, these basic patterns also coexisted with a fourth configuration of textile labourers. This was a phenomenon closely associated with colonialism in Africa and the Americas, and characterized by the use of non-voluntary labourers (outright slaves brought from Africa, or those held in obligatory arrangements). In this configuration, workers did not elect to enter the textile industry to earn additional capital, but were obliged to produce cloth on a daily basis, usually as their primary assignment, in settings which ranged from plantations to Europeandirected obrajes or workshops. Despite the similarity of chronological time frames, superficial resemblances to European proto-industry must be discounted, as the workers in these coerced textile enterprises exhibited none of the characteristics associated with that modality; that is they were not ethnically homogeneous, short-distance or seasonal migrants, nor were they wage-earners. Furthermore, the differences between the European form of proto-industrialization and the American coerced-proto-industrialization cannot be minimized by referring to economies of scale, as American workshops were not minor enterprises but produced thousands of yards of fabric sold in local markets and involved large numbers of workers. This fourth configuration of textile workers based on coerced labour is consistent throughout the Americas, but there are also echoes in Africa, in eighteenth-century India under British rule, as well as in Eastern Europe where the serf system, or its lingering characteristics, remained. The widespread global presence of coercedproto-industrialization suggests that such a modality was hardly an aberrancy and that it must be acknowledged as an alternative form of production. 2) With regard to migration, there is also a distinct difference between European patterns and what is seen elsewhere. The migration of textile workers within Europe has been largely internal (within national borders) and thus largely ethnically homogeneous throughout Stages I and II (the sixteenth–eighteenth and nineteenth– twentieth centuries, respectively). Although there has been constant relocation of workers from rural to urban or semi-urban settings where ‘proto-industries’ or workshops were located, the majority of these workers migrated relatively short distances to employment opportunities (notwithstanding that the long-distance migration of ethnically diverse workers was a feature of the Hapsburg Empire). This characterization holds even for Stage II when the application of mechanization to the urban textile industries eclipsed the output of rural producers who, in turn, responded by migrating to mills in nearby urban centres.112 Although examples of long-distance or international migration are not unknown (for example the Irish workers in nineteenth-century Lancashire or Turkish ‘guest-workers’ in the 112 Moch, Moving Europeans, p. 132. 675

The Ashgate Companion to the History of Textile Workers twentieth), local (internal) and short-distance migration/recruitment has remained the dominant pattern to the present. The same may be observed for India, where there was mostly internal migration from rural to urban or semi-urban settings, resulting in an ethnically homogeneous workforce. It is equally the case for the textile industries of Japan and China up to the twentieth century, although recent locations of industries to rural regions in both countries (as well as in India) may involve greater migration distances. Not surprisingly, given the breadth of the Ottoman Empire in the nineteenth century, configurations of textile labour forces varied substantially, ranging from ethnic enclave labour recruitment in textile production to localized, internal migration of ethnically homogeneous workers. By contrast, in the Americas during Stage I, long-distance migration of workers was the norm. While it is true that migration of local, indigenous workers did occur, the larger ‘recruitment’ of textile workers took place not in the context of local labour pools, but reflected international migration. Moreover, even where labourers were of indigenous stock (Argentina, Mexico and Brazil, Cape Verdes) they still represented a distinctly different ethnicity from the owners of the textile works. This diversified character of textile production has been perpetuated into Stage II in the Americas, when local workers laboured in mills alongside foreign immigrants. Other geographic regions have exhibited relatively limited diversity in the composition of textile labour forces (with the notable exceptions of the skilled English textile workers in Egypt, the Irish workers in Lancashire and the Turkish workers in France and Germany up to the present) and broadly speaking nowhere else does the degree of migration and ethnic diversity of textile workers occur on as great a scale as in the Americas. 3) Migration of skilled labourers represents a specialized category within the broader range of textile workers. Some prime examples include the textile-worker migration system in Leiden, the Hapsburg Empire (Stage I) and in Egypt and the Americas (Stage II). Nevertheless, the migration of unskilled workers to textile industries is undoubtedly the statistically more significant phenomenon. 4) Ethnic differences between workers and owners of plants, or preferences for certain types of migrant labour may have been an important factor initially in some textile industries, but disappear over time as such migrant textile workers (and sometimes owners) become absorbed and/or naturalized into the local population. While discrimination against some ethnic groups of textile workers has been documented into the twentieth century (for example the Subei women in China and Korean workers in the interwar period in Japan), and waves of immigrants undoubtedly jockeyed for status in factories, the bottom line for the employer was having sufficient labour available. Accommodation was also the byword from the immigrant’s point of view and, as Nancy Green has suggested with regard to the related US garment industry, ‘each new group is compared to the previous one,

676

How will we get our workers? whose decried self exploitation at the time of their arrival starts to look good as the next train or boat-load arrives’.113 5) Finally, protectionism, globalization and/or free trade agreements may change forever the pattern of migration and ethnic composition of textile workers. The recruitment of Turkish ‘guest-workers’ for work in textile factories is still an important component of the industry in France and Germany, but has been discontinued in the Netherlands. As textile production shifts towards China, reflecting recent WTO decisions, and China demonstrates unwillingness to be constrained by quotas, textile workers globally may find their opportunities for employment have all but disappeared, regardless of their willingness to relocate. Seen synchronically and diachronically in global perspective, distinct trajectories of migration and ethnicity of textile workers have emerged over the course of five hundred years. Using these variables to consider the history of global textile workers has allowed the discussion to advance beyond the parameters of the European experience. Had the discussion been limited to modes of production, without consideration of the workers themselves, these distinctions would not have been evident. Clearly colonialism (especially in the Americas) and its lingering influence globally, more than anything else, is responsible for deviations from European norms. To understand textile industries, therefore, at any point in time, one must examine not only the production mode but the labour structure and the particular circumstances in which the industry operated. As in the past, what lies ahead for the future of textile workers will likely be determined by globalization and the continuing cycle of eclipse of traditional production zones.

113 Nancy L. Green, Ready to Wear and Ready to Work: A Century of Industry and Immigrants in Paris and New York (Durham, 1997), p. 196. 677

This page has been left blank intentionally

27 Work floors under tension: working conditions and international competition in textiles Peter Scholliers

Through time and space working conditions in textiles varied tremendously according to production organization, customs, power relations, fibre, or technology. Take on the one hand a young girl who had actually been sold by her parents to a labour contractor for working during a given period in a Shanghai mill in the 1920s. She received hardly any pay, and worked for long hours in harsh conditions, living under total control of the ‘boss’ in crowded rooms, with no possessions. Then, on the other hand, take a Lancashire mule spinner in the same period, an ‘aristocrat’ with high income, who directed his team and could afford to take his family to a seaside resort. Both workers lived in entirely different conditions and yet, they both produced cotton yarn, that eventually competed in one market. Very dissimilar production modes existed (and still exist) alongside each other. Sometimes they cooperated within one production line (carding and spinning in modern mills, weaving via domestic production and finishing in specialized, small workshops). More often, however, the various modes of production throughout the world competed with each other, which was particularly the case from the moment the modern mill emerged around 1800. This competition influenced the work effort, working hours, organization, technology, shop floor relations and,  I sincerely thank the editors, as well as the contributors to the debate during the conference, for the many useful comments and tips.  The ‘Introduction to Part IV’ of D. Jenkins (ed.), The Cambridge History of Western Textiles, II (Cambridge, 2003), stresses largely the importance of ‘comparative advantages’ in the nineteenth century, for instance with the following phrase ‘Unprecedented levels of international competition in all the main textile areas resulted’ (p. 718).

The Ashgate Companion to the History of Textile Workers most directly, pay. Moving beyond production, this competition also affected the household, housing conditions, identity formation and community life. This chapter tries to capture some consequences on the work floor of the changes in the modes of production throughout time and space. However, working conditions, work floor relations and general traditions in turn also influenced the factors that drove this competition. For example, resistance to wage cuts may have led to the introduction of new machines, or premium systems may have increased productivity and lowered the sales price. The present chapter may be read in order to understand and explain technological innovations, the size of the mills, (gender) composition of the workforce, and identity formation. Hence, the consideration of working conditions has a mediating and central position between the economic (technology, organization, size of mills, output) and the politico-cultural (trade unions, family life, identity). Nonetheless, and most prominently, this chapter undoubtedly has concerns of its own: it explores the working conditions with particular interest in their changes in relation to the increasing international competition. What happened to traditional working conditions when confronted with production or importation from modern textile mills? How did working conditions in modern mills change when confronted with constantly diminishing production cost in putting-out systems? Which worker fared better; the one who worked under the Kaufsystem, the one who toiled under the putting-out system, or a modern mill worker? Naturally, this is hugely ambitious, and therefore some limitations should be set. It would be easily conceivable to distinguish between working conditions in a pre-factory and a factory period. The pre-factory period would run to the moment when the factory (a central production unit – whether a workshop, a manufacture or a modern mill) with machines driven or not by mechanical power and operated by wage workers, replaced home production of yarn and cloth. Within the prefactory period a distinction could be made between the putting-out system and the Kaufsystem, according to the level of dependence of the worker with regard to the supply of raw material, tools and machines. This division, however, may not mean a big difference with regard to actual working conditions. Clearly, working conditions in the pre-factory world (at home, with the family, often combined with agricultural labour, flexible hours, own work rhythm) differed greatly from those in the world of the factory (separation between home and workplace, full-time employment, strict control, centrality of the wage, machine-regulated work pace).  M.B. Rose, ‘International Competition and Strategic Response in the Textile Industries since 1870’, Business History 32 (4) (1990): 1–8, stresses particularly the striving for low(er) labour cost in response to increasing international competition (although of course this was not the only tactic).  W.O. Henderson, ‘The Labour Force in the Textile Industries’, Archiv für Sozialgeschichte 16 (1976): 321–322.  Following the national overviews in Part I, my chapter only considers production for the market. In rare occasions the reports mention non-market textile production, but then nothing is said about working conditions. 680

Work floors under tension The advent of a modern mode of production alongside a traditional one occurred in all countries under consideration. Most of the national overviews in this volume address the fate of domestic workers after the emergence of modern mills or the arrival of their products from foreign countries after c.1800. The outcome of this confrontation for the local producers was ambiguous and is at present debated. Often, it meant the disappearance of traditional production. In India the coming of mechanically produced yarn from England led to the ‘loss of industrial employment to the extent of 4–5 million persons’, in Egypt ‘[European imports] resulted in the decline and even disappearance of certain crafts’, in Mexico ‘England’s cheap exports … combined to hammer Mexican textiles’, in Russia the British technological innovations made ‘several factories close’, and in China ‘European textiles … began a fundamental, and initially very painful transformation of cotton textile production’. Obviously, European exports had a worldwide devastating effect. However, competition with modern products could equally unleash ‘positive forces on handloom cloth demand’ as in India, lead directly to the expanded use of the putting-out system as in Egypt, or could extend rapidly due to higher tariffs on imported yarn, as in Russia. Such mixed consequences of the clash between ‘old’ and ‘new’ are also reported for Argentina, Austria, China and the Ottoman Empire. I may thus suggest that the double reaction to modern millwork was a general phenomenon, albeit with different rhythm and size, and with, in the end, the prevalence of modern millwork. These ambiguous consequences are (still) in need of academic attention. The fate of domestic workers after the confrontation with centralized modes of production points at the problem with regard to a clear chronology about ‘old’ and ‘new’ work conditions. Such a partition carries in fact a triple problem. Firstly, the coming of more modern production units or of imported goods did (and does) not mean the end of traditional production modes. Secondly, the timing and rhythm of modernization differ highly from one place to another and, thirdly, the pace of modernization depended on the type of fibre. Often, ‘old’ and ‘new’ production modes existed alongside each other, and sometimes encouraged each other, as was the case in Lancashire, the heartland of textile innovation in the late eighteenth century. This makes a simple, overall chronological division of the global history of working conditions in textiles nigh impossible. I therefore shall consider diverse themes rather than apply chronological divisions, although a distinction between the modern mill (a mechanized production unit driven by steam power), manufactories, workshops and domestic production (in this order) shall be retained here. The themes under consideration are of course guided by the attention of the national overviews. I shall thus limit my chapter to four main topics: the condition  Roy, Chapter 10, this volume, ‘India’; Beinin, Chapter 7, this volume, ‘Egypt’; Bortz, Chapter 13, this volume, ‘Mexico’; Cliver, Chapter 5, this volume, ‘China’; Pretty, Chapter 16, this volume, ‘Russia’.  Roy, ‘India’; Beinin, ‘Egypt’; Pretty, ‘Russia’. 681

The Ashgate Companion to the History of Textile Workers of the workspace, the practice of the labour market, work control and working hours, and wages. This implies the disregard of themes such as rest breaks, the actual measuring of the daily or weekly output per worker and the conflicts about it, the way the workers experienced and valued their work, the cyclical or structural unemployment, and the mills’ rules with regard to eating, talking, smoking and firing. Finally, I wish to stress that, to avoid superficial comparisons, this chapter could hardly be read and comprehended without keeping in mind the national overviews with their often-implicit wider historical contexts.

Room to work In many cases the coming of modern textile mills since the last quarter of the eighteenth century meant the start of the definite division between living place and working place. Only rarely do the national overviews mention characteristics of modern mills (building style, windows, shape, size, space, ventilation, development and so on). Did all modern textile mills in the world have a similar appearance, copying more or less the pioneering ‘Lancashire style’? Did they offer sufficient and bright space? Did they have sanitary equipment, cooking facilities or, later, cafeterias? Were they established in areas with a textile tradition, making use of existing buildings (as in Belgium around 1800, where nationalized abbeys and monasteries were converted into textile factories)? Did the new machines, including huge water or steam power equipment, readily fit into these buildings or was new construction necessary? At least textile mills in Mexico had a particular look: they were established in old haciendas, and newly erected buildings kept some characteristics of the hacienda. In general, the interior conditions of modern textile mills were bad, which changed only gradually in the course of the twentieth century. This became known through enquiries of individual witnesses, unions and official institutions, which started in the 1840s and multiplied rapidly. Almost all of the national overviews stress the bad interior conditions of textile mills into the twentieth century. A distinction may be made according to the size of the mills: conditions in bigger mills seemed generally to be better than those in smaller ones (as in Argentina in 

Bortz, ‘Mexico’. In the case of one Uruguayan mill, the Campomar & Soulas (1907), some information is provided about the mill’s architectural style (Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’). A vast literature exists about industrial architecture, for example C. Bertsch, Fabriksarchitektur (Braunschweig, 1981), that includes modern textile buildings, but some authors devote particular attention to textiles. See M. Doriez et al., Architecture Textile (Paris, 1990), or C. Giles and I. Goodall, Yorkshire Textile Mills: The Buildings of the Yorkshire Textile Industry 1770–1930 (London, 1992).  See, for example, L. Villermé, Tableau de l’état physique et moral des ouvriers employés dans les manufactures de coton, de laine et de soie (Paris, 1840); N. Briavoinne (ed.), Enquête sur l’industrie linière (Brussels, 1841) and (of course) F. Engels, Die Lage der arbeitenden Klasse in England (Leipzig, 1845). 682

Work floors under tension the 1930s). In both, however, ‘heat, humidity and cotton particles floating in the air were health hazards’.10 The latter was also reported in German and Dutch cotton mills, where the tremendous noise was emphasized. In China, cotton mills were described as ‘darker and dirtier than the Japanese mills [in China], but all cotton mills were noisy places, filled with dust and fluff’.11 In cotton mills in Japan, ‘health and safety provision was inadequate’.12 In New England and Dutch cotton mills of the 1900s heat, humidity and drafts resulted in poor health conditions. In Russia, working conditions were designated as ‘miserable’ or ‘harsh’, and, moreover, safety concerns were absent, which led to frequent injuries and mutilations. Thus, modern textile mills offered dusty, wet, noisy, dark, dirty and risky workplaces: bad, indeed. The number of employees in the new mills shows that nineteenth- and twentieth-century modernization was very sudden and extreme, with mills that came out of the blue and housed several hundreds of workers right from their creation. Modern mills, thus, rarely developed gradually but appeared fully equipped, which necessitated big investments. This, in turn, led to generally hard and stubborn attitudes from the employers. Examples of the rapid expansion of mills may be found all over the world. In New England in the 1830s eight mills totalled 6,000 workers, in Bombay’s modern mills of the 1850s the employment scale was impressive, the first Mexican modern mills ‘were much larger than the old workshops’ and in Russia modern spinning factories in the 1850s had an average of 700 workers.13 In most countries after 1870 the bulk of employees worked in factories that grew even bigger, as shown by the examples of Spain, Denmark, Russia, Japan, and Mexico. It is not certain whether bigger workplaces imply better working conditions. Moreover, modernization did not automatically mean ever-growing mills. In Japan, for example, factory silk reeling remained relatively small-sized. ‘Globalization’ and neo-liberal politics from the 1970s onwards led to fast decreasing employment in the older production centres as a result of competition with low-wage regions,14 but it is not clear to what extent this affected the average size of mills (for example as a consequence of specialization). Very generally, hygiene and security conditions in textile mills improved in the course of the twentieth century. This progress may be linked to the extension of general social legislation that aimed at the length of the working day (see below), to the creation of medical care, employment and wage protection or, as suggested above with regard to Argentina, to the fact that mills became bigger. Legislation may have resulted from paternalism, revolutions or strike movements, 10 Lobato, Chapter 2, this volume, ‘Argentina’. 11 Cliver, ‘China’. He also refers to working and health conditions in the 1950s. 12 Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’. 13 Blewett, Chapter 21, this volume, USA; Roy, ‘India’; Bortz, ‘Mexico’, Pretty, ‘Russia’. 14 For example Lobato, ‘Argentina’, Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’ and Fowler, Chapter 9, this volume, ‘Great Britain’ point at the fact that the industry ‘practically disappeared’ in these countries, while Cliver, ‘China’ reports the move towards cheaper provincial towns within the country. 683

The Ashgate Companion to the History of Textile Workers but mostly it emerged through democratic governments and/or institutionalized labour relations. Undoubtedly, the bad conditions in textile mills, and particularly the high labour participation of women, adolescents and children, contributed to the setting up of general social legislation. In some countries, hygiene conditions started to improve before the First World War, as was the case in Denmark. In other countries, however, notable legislation only appeared in the 1990s, as was the case in Egypt. The coming of big factories after about 1800 did not necessarily mean the very onset of the separation between home and work. In many countries prior to 1800 (and indeed up to now), textile production had been organized within workshops and manufactories (the latter being generally larger than the former) that could bring together hundreds of workers. Thus, Europeans established workshops in Argentina in the last decade of the sixteenth century (obrajes), where the raw material arrived and came out as dyed cloth. In Russia from 1840 onward, svetelki (buildings that could contain up to 15 handlooms) were set up, which were run by groups of peasants on a communal basis and, more frequently, by well-off peasants who supervised workers and took a commission on sales. In Twente (the Netherlands) in the 1800s, sheds were set up to install several looms. Because of the capital investment (that was often huge due to the buildings, stocks and general costs) and because of expected higher tax revenues, these manufactories were often established, directed and/or promoted by the state. Such was the case in Austria in the eighteenth century, where ‘manufactories were huge enterprises employing hundreds of workers in the central manufacture and thousands of putting-out workers’.15 Such was the case also in Cairo in the late 1810s, where up to 2,000 workers were employed at the Khurunfish manufacture. Colonial powers in particular erected big factories, as the Portuguese did in eighteenth-century Brazil. In many countries, thus, the regime interfered directly in textile production, which could have affected working conditions (see below, with regard to wages). The outer appearance and interior conditions of workshops and manufactories varied, but in general they were bad (as in the modern mills up to the early twentieth century). In seventeenth-century Holland, ‘[workers] worked … in the same physical position in dark, badly ventilated rooms’.16 Mexican workshops around 1800 were described as ‘resembling a dark prison’.17 And in Cairo around 1900, weavers worked in ‘quarters where most of the buildings are ramshackle, at the back of courtyards, in sheds without roofs … in long, narrow, and badly lit spaces’.18 Domestic spinning and weaving represent the work form that has been the most common throughout time and space. As I emphasized earlier, the advent of other production forms did not necessarily threaten the existence of domestic work. On the 15 Komlosy, ‘Habsburg Monarchy’. 16 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’. 17 Bortz, ‘Mexico’. 18 Beinin, ‘Egypt’. 684

Work floors under tension contrary, domestic work formed an integral part of the organization of production in workshops and modern mills. Very little is reported by the national overviews on the domestic working environment. Would it have made a difference, in terms of work conditions and workspace, whether workers were employed within a Kauf- or a putting-out system? In England in the eighteenth century, for example, a difference existed between worsted (fine) and woollen textiles. Workers of the woollen textiles owned the tools and raw materials, and sold their production on the local market. But workers of the worsted textiles received the raw material from a ‘big clothier’ who sold the production, that is, a more capitalistic production mode. The former often was an agricultural worker, the latter a full-time wage worker. Would this have made a difference concerning the actual work conditions in the house? Probably, it did not. This divide between the so-called Kaufsystem (with a modest role of a merchant) and the putting-out system (with a more prominent role of a merchant or other intermediary) existed also in other countries, and the two systems could operate alongside each other. In eighteenth-century China, for example, the baomaizhuzhi (a putting-out system) emerged, which responded to increasing demand for cotton yarn and cloth, although from early times rural weavers who grew their own cotton and possessed their tools, directly sold their product to merchants. Often, colonial powers established a form of putting-out work, like the encomienda (a concession system) in Argentina. Although domestic textile labour is often associated with the countryside, agricultural work and seasonal occupation (as, for example it is very prominent in so-called proto-industrialization), homework in textiles was largely embedded in towns and cities too. Thus, in Denmark, Egypt, India, Japan and the Netherlands up to the eighteenth century, domestic weavers worked in villages and towns, where guilds commonly organized the work.19 A particular form of working conditions came along with slavery. In seventeenthcentury Brazil the fazenda existed, which was small-scale domestic textile production that was mostly meant for internal use and where slaves were put to work. Modern Brazilian factories also employed slaves up to the 1880s. Such labour was applied for a short while in the above-mentioned Khurunfish factory (Cairo, 1816), where black Sudanese slaves were employed before free labour was recruited. The same occurred in the Russian cotton industry of the first decades of the nineteenth century, where serfs made up to 5 and even 15 per cent of the workforce. However, in this country, freely hired workers were often bound one way or another (for example, to pay off feudal indebtedness), which questions their legal status. A particular form of slavery, although it was not labelled as such, was the contract worker system in China. Parents sold daughters for a certain period to labour contractors who put 19 The literature on craft guilds is (again) abounding after the reassessment of their cultural significance, political influence and, especially, their role in economic development. The national overviews do not tackle these issues with particular interest, though. On the role of guilds, see H. Soly, ‘The Political Economy of European Craft Guilds: Power Relations and Economic Strategies of Merchants and Master Artisans in the Medieval and Early Modern Textile Industries’, International Review of Social History, 53: supplement (2008): 45–72. 685

The Ashgate Companion to the History of Textile Workers the children to work, housed and fed them, and controlled virtually every aspect of their life. The borderline between free and unfree seemed to have been thin at times. For example, black people and castas (poor Spaniards) in Mexico worked in a mix of free (wage) and unfree labour. Clearly, conditions of the workspace have differed widely according to place and time. These conditions were the outcome of custom, labour relations, state interference, investments, organization of the trade, size of the mills, technology and so on, all embedded in local, regional and national widely varying contexts. Despite these very distinct contexts, one common characteristic ought to be mentioned: whether at home, at the workshop, the manufacture or the modern mill, the workspace generated harsh working conditions, which started to change in some countries only in the late nineteenth century. In general, the (re)organization of the production in one place (whether under the influence of so-called proto-industrialization, colonialism or the industrial ‘revolution’) did affect the way the workspace was shaped and arranged in other places.

Training, hiring and hierarchies The modern mills of the nineteenth and twentieth centuries necessitated an abundant number of workers, which led to new forms of recruitment, as well as to problems of retaining (experienced) workers. Factory labour poured into the mills via two channels: local workers and immigrant workers. Employment of local workers occurred in areas with a textile tradition, as in late eighteenth-century Lancashire and Catalonia, mid-nineteenth-century Bahia, and late nineteenth-century Japan. In general, however, immigrants from the surrounding countryside made up the bulk of the new workforce. This was the case, for example, in New England in the 1820s and in US southern states in the 1880s, in Bombay’s and Calcutta’s cotton factories in the 1860s, in Russia’s modern spinning mills in the nineteenth century, and in Chinese and Japanese mills in the early twentieth century. Workers from the countryside may have functioned in very different systems of labour markets, though. In general, two forms may be distinguished: the market and the family-based system.20 In the latter, close agricultural bonds and the nature of the family (its composition and values, members with a job, the income level) were much more decisive. Such distinction had implications on, for example, the housing conditions (boarding houses against living-in with the household in factory villages or mill towns; see below). Japan offers an example of a family-based labour system; Catalonia offers an example of a market system. However, such a division was never strict. The family did play a role in the market system, as is shown by 20 This is based on the comparison of labour markets in the South of the USA and in Japan in different periods. In fact, three systems may be distinguished, including a ‘total market’ system (G. Saxonhouse and G. Wright, ‘Two Forms of Cheap Labor in Textile History’, Research in Economic History, Supplement 3 (1984): 3–31). 686

Work floors under tension research on modern millwork in cities, pointing at the relationship between labour participation of women and children, and work and income of the husband.21 In some regions, immigration meant long-distance journeys, as for example in the modern spinning factories of St Petersburg and Moscow, where workers came from villages that were hundreds of miles away. The fact that the great mass of workers originated from the countryside could imply the persistence of a particular bond between their new life in town and their old life in the country. Even second-generation migrants (born in the city) often wished to keep such a bond. A survey of workers in a printing factory in Moscow (1880s) showed that no less than 90 per cent of the workers had an allotment of land in the villages of their (fore)fathers. This combination of loom and plough guaranteed the best chance to do well. A similar tactic may be observed in other countries too, as in the Netherlands. Many mill workers in the town of Twente cultivated a small plot of land, about which a weaver testified in 1890 ‘I put a lot of effort in my little farm, so I will always have something to eat, even if I am fired from the factory.’22 The bond between the factory work and agriculture undoubtedly had consequences for the general wage level in these countries (see below). Some workers, however, travelled over very huge distances, which was particularly true for skilled workers and engineers. In Argentina, textile workers came from Italy and Spain in the 1890s, in New England British textile workers found jobs in the 1820s and 1830s, and in Brazil after 1870 British engineers held managerial jobs. Most likely, the hope for better income drove skilled workers to migrate, as was shown by Lancashire weavers and spinners who migrated to the USA, after a wage cut of 10 per cent in England. Recruitment of mill workers differed according to skill, gender and local custom. Traditional forms of apprenticeship (guilds, strict rules, learning money, and so on) hardly existed in modern mills.23 Abilities and skills were acquired within the factory. Some tasks necessitated a training of several months and even years, whereas for other tasks a week or two would suffice. In modern Japanese textile mills workers were recruited via billboards, personal contacts and by word of mouth. Very likely, such was the case in most textile regions around the world. Personal contacts seemed to be most common. Networks of workers’ families and friends were reported in the Catalan, Uruguayan, Russian and Chinese textile mills of the nineteenth and twentieth centuries, thus providing 21 See, for example, the research on family income in two textile centres of Europe, C. Borderias and P. Lopez, ‘A Gendered View of Family Budgets in Mid-Nineteenth Century Barcelona’, Histoire & Mesure 18 (2003): 113–145, and P. Van den Eeckhout, ‘Family Income of Ghent Working-Class Families Circa 1900’, Journal of Family History 18 (1993): 87–110. 22 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 23 In Spain ‘systems of apprenticeship’ were adopted within the new modern mills (Smith et al., Chapter 17, this volume, ‘Spain’), which refer to spinners who trained piecers (as in Lancashire). This, however, had no link with traditional apprenticeship. In the modern silk factories of Shanghai in the 1920s, new workers received training as apprentices (Cliver, ‘China’). 687

The Ashgate Companion to the History of Textile Workers a relatively smooth functioning of the labour market. In Russia, peasants found it easy to get a job (and housing) through family networks. A complex system operated in Chinese cotton mills. The ‘Number One’, often a female worker with long experience, served as go-between for workers (and their families) and employers. This ‘Number One’ received small gifts and offered protection to ‘her’ workers.24 In Indian cotton mills, the ‘labour contractor’ may be compared to the Chinese go-between: female recruiters relied on connections (often in the countryside) for supplying labour. In late nineteenth-century Japan, mills’ recruiting agents scoured particular regions. It may be asked whether such intermediaries were necessary in areas with abundant workforce and/or family networks. There is no straightforward answer to this question. Some recruiters were looking for a cheap and docile workforce, and hence supplied impoverished migrants from low-wage areas. In some countries mill owners searched very actively for workers they thought best suitable (and wishing to avoid intermediaries). For example, some Japanese mill owners sent out employees to enlist workers. On the other hand, some migrants searched for work for a short while, thus trying to diversify income. In Japan around 1900, employment in cotton mills was considered as providing an education as well as additional income. Early nineteenth-century mule spinners held a special position in the labour market. Although they themselves were wage workers, they enlisted and paid ‘little’ and ‘big’ piecers, who formed a pool of labour from which spinners were recruited. After the First World War, labour supply and demand was organized more formally. Labour exchange offices appeared (as in Japan and many European towns), while unions were in a position to influence recruitment (as in Mexico, where the Puebla code made unions responsible for hiring workers). Textile mills were set up in towns and in the countryside, and in both cases the mills often supplied housing to the workers. This was done to secure sufficient and tranquil labour. One factor that explains this initiative was that labour turnover was quite high, as the examples of China, Mexico, Japan and Egypt around 1900 show. In Austria and the Czech lands in the first half of the nineteenth century, factory villages were erected where traditional social norms were absent, and where ‘workers were shut up in the factories, which they only were allowed to leave on Sundays’.25 In New England in the 1830s, workers, whose ‘behaviour was regulated and controlled’, slept in boarding houses.26 The latter was also the case in late nineteenth-century Brazil where vilas operarias replaced crude dormitory arrangements (implying a shift from a family-based to a market system of labour recruitment?). In Shanghai local women were dissatisfied with the room and board offered by the mills, in Japan ‘early dormitories were often crowded and unsanitary’

24 Cliver, ‘China’. In Shanghai in the 1920s, criminal gangs (‘Number Ones’, as well as ‘labour contractors’) controlled the labour market. 25 Komlosy, ‘Habsburg Monarchy’. 26 Blewett, ‘USA’. 688

Work floors under tension and in Russia ‘large barracks or apartments for families’ were provided.27 In Mexico, genuine mill towns appeared where paternalism was commingled with Catholicism. In Brazil, for a short period, mill owners provided food and education next to housing. In all cases mentioned here, living conditions in the boarding houses were initially hard. The fact that dormitories and boarding houses primarily lodged girls and young women was used to justify the severe rules. However, alongside such rules, the dwellers were victims of harassment of diverse kind. For example, within the Chinese contract-work system, girls were not allowed to keep their wages, they had to live in crowded rooms, had no possessions, had poor meals, often fell ill and were weak, tired and apathetic. Living conditions eventually improved under paternalist sentiments, workers’ struggle, or legislation. In Japan, for example, the initial bad housing conditions ameliorated considerably over time, and large mills in particular extended paternalist policy, which included educational and recreational facilities. From the 1950s onwards, dormitories were even seen as an important advantage by young female workers because of the opportunity to save or to spend money on more and luxury goods. Housing conditions of textile workers who did not live in dwellings owned by the mill and who undoubtedly represented a large part of the textile workforce in most European countries are hardly dealt with in the national overviews. For late nineteenth-century Barcelona, for example, it was reported that real wages were low which resulted in poor housing accommodation, gross overcrowding and a lack of sanitation, and death rates that outstripped birth rates. Russian cotton workers in the late nineteenth century rented quarters, ‘usually just a corner of a room’.28 Similar observations may be read with regard to (individual) textile dwellings in seventeenth- and eighteenth-century Holland, where domestic spinners lived ‘in basements, renting small rooms from richer citizens or [lived] in special houses for the poor’.29 In general, sixteenth-to-eighteenth-century workshops and manufactories hired workers in similar ways to modern textile mills. Relatives and friends were important for entering a workshop, and in some cases chain migration of textile workers appeared. Merchants could play the role of the recruiter, as was the case in the rural and urban silk industry of the Ottoman Empire. A peculiarity seemed to have been the fact that religious dignitaries may have been involved in the hiring process. Silk weaving workshops in China used the labour of skilled workers who gathered at specific places under guild auspices. Another resemblance to modern mills was that manufactories provided dormitories for the (female) workers. Big manufactories not only recruited local workers, but sometimes also foreign workers, as was the case in early eighteenth-century Austria and Denmark. With regard to domestic workers a distinction may be made between rural and urban labour. In the towns, whether in Latin America, Asia, Europe or Northern Africa, artisans’ workshops operated within the guild system, with masters 27 Delson, ‘Brazil’; Cliver, ‘China’; Hunter and Macnaughtan, ‘Japan’; Pretty, ‘Russia’. 28 Pretty, ‘Russia’. 29 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 689

The Ashgate Companion to the History of Textile Workers accepting and training apprentices who often lived with their master. In villages, however, weavers were allowed to stay outside the guild system, as was the case in eighteenth-century Denmark. Guilds’ rules and wage norms were seen as restricting the development of textile production in towns, and therefore merchants searched for work in the countryside (of which seventeenth-century Japan and the eighteenthcentury Ottoman Empire provide examples). Unskilled domestic labour was mostly recruited in rural areas, where the needs of the workshops and merchants were confronted with long-term established (feudal) rules. In the Habsburg Empire, a set of laws abolished serfdom and liberated marriage, travelling and choice of profession, thus setting free restrictions on work, which allowed thousands of rural workers to engage, willingly or not, in textile production.

Work control and effort It is a commonplace to refer to the fact that modern mills were established to make profit, and that efficient work by a steadfast workforce was required to guarantee that. The creation of a suitable and disciplined workforce was a problem in early nineteenthcentury Lancashire,30 just as it was in twentieth-century Brazil or Japan. Most modern mills had house rules, and whether or not these were written down in an individual contract31 or put on posters on the mill’s wall, workers knew about work rules. Very generally, work control was particularly severe when modern mills were started up. This barely changed for a long time. In some mills in the 1930s, such as the Egyptian Misr Spinning and Weaving, control by harsh discipline was still considered as ‘economically rational in the short term’. In the course of the twentieth century, however, crude hierarchical relations gave way to more balanced relations, which may be linked to the emergence and growth of unions and to political interference. This appeared, for example, in many European countries with the emergence of collective agreements around 1900, in Japan during the 1950s and, more radically, in the People’s Republic of China. Paternalist sentiments may also have softened the crudest systems of control. However, political conditions also enforced severe work control, as was the case during dictatorship in Argentina in the 1970s and 1980s. Control was exerted by various means. First and foremost, the cadence of machines regulated work, which made the difference with domestic spinning and weaving. The introduction of new machines, or more machines for one worker, meant the increase of the work rhythm, as, for example, in Argentinean and US cotton spinning in the 1930s. Often, weavers complained about the fact that this made their job ‘too nervous’ and too arduous, as, for example, in the Netherlands 30 W. Mass and W. Lazonick, ‘The British Cotton Industry and International Competitive Advantage: The State of the Debates’, Business History 32 (4) (1990): 9–56, at p. 11. 31 Formal, written contracts appeared in the twentieth century, but were not widespread (as in Japan, see Hunter and Macnaughtan, ‘Japan’). 690

Work floors under tension in the 1900s. Yet, more modern machines did not always lead to a faster work rhythm. Automatic weaving looms that appeared after 1900 ran at a slower pace than power-looms. Yet, the automatic looms were used in the 1930s to further the multiple loom system (with one worker operating six, eight or more looms at once). This, of course, increased work effort as well as productivity. Secondly, the clock – or time in general – became a cogent factor of control because of the way workers were paid (see below). Thirdly, supervisors continuously controlled the work rhythm and production.32 They did so in an authoritarian way (as in the US southern cotton mills prior to 1900), by using violence (as recalled by an Egyptian worker in the 1930s, or reported by an observer of Chinese cotton mills in the 1920s) or by being capricious and intimidating female workers (as in Mexican cotton mills of the 1870s). More complex forms of control were exerted according to the way workers had been enlisted. European mule spinners of the first half of the nineteenth century directed their team (the piecers, sometimes including their own sons) with a strong hand, because the spinners’ income directly depended on the team’s effort. Chinese ‘Number Ones’ and Indian contractors established personal ties with workers, which were ‘exploitative and very complex’.33 Combined with harsh control in dormitories and boarding houses, control in and out of modern mills was oppressive: in Japan, for example, it was reported that the ‘integration of residence with the production process … offered greater possibilities for labour control’.34 Work control in manufactories and workshops did not differ much from that in modern mills, despite the much higher investment and hopes of prompt returns of the latter. In the Khurunfish factory (Cairo, 1820s) work discipline was maintained by three overlapping control systems: masters supervised the guild members, government officials controlled all aspects of production, and soldiers were present to enforce discipline. Such mixed control forms, whether instigated by entrepreneurs or state administrators, appeared in other countries too. These were sometimes supplanted by yet other control, as in Japanese manufactories, where a ‘fictive parent–child relationship’ existed between older and younger workers, with the ‘parent’ offering guidance and welfare and the ‘child’ returning obedience and obligation.35 Work control in a domestic context differed of course from that in workshops or mills. This does not mean, however, that work would be more independent, although domestic workers sometimes were self-employed. In eighteenth-century Austria, for example, domestic workers depended on putting-out agents who wished to lower the price and, hence, the wage. They thus immediately affected the work effort, which had to increase if income were to remain stable. 32 Information on work relations is to be found in P. Van den Eeckhout (ed.), Supervision and Authority in Industry. Western European Experiences, 1830–1939 (New York and Oxford, 2009), and particularly the chapters on the Barcelonese and Ghent textile industries. 33 Cliver, ‘China’; Roy, ‘India’. 34 Hunter and Macnaughtan, ‘Japan’. 35 Ibid. Such particular form of control also existed with the ‘Number Ones’ and labour contractors in China and India. 691

The Ashgate Companion to the History of Textile Workers Work effort was directly linked to the length of the working day. In all countries under consideration, the working day was long and seemed to have been extended up to the late nineteenth century. This includes all types of work organization (mills, workshops and domestic work). Long and extremely long working hours were reported for modern mills all over the world. In the textile mills of the Northeast USA in the 1850s, workers protested against long working hours (14 hours per day), and in the Habsburg Empire in the 1880s spinners in modern mills worked ‘as many hours as possible until they were replaced by newcomers’.36 In Russia’s early cotton mills the working day reached 14 hours, which continued well into the 1890s. In Dutch cotton factories, work lasted for 12 to 13 hours per day, leading to a total of 75 actual working hours per week since work continued on Saturday morning. In order to lengthen the working time, customary Catholic holidays were converted to working days in the 1870s. In Germany’s spinning and weaving mills up to the 1860s, working days lasted between 14 and 16 hours, which necessitated that workers sleep in the mill and which threatened their health. In Brazil’s late nineteenth-century factories, long hours were standard and the installation of electric light made night labour possible, leading to working days of 14 to 17 hours. Night work and night shifts were widespread throughout the world from the late nineteenth century, and were vehemently combated by workers and their associations.37 In Chinese cotton mills of the 1920s, women ‘started their days early and ended late’, with most working 11 to 12 hours per day, with half an hour rest for lunch. Night and day shifts implied that workers would share the same sleeping space.38 In the Japanese cotton and silk mills around 1900, ‘working hours were extremely long’, ranging between 12 and 16 hours, which included night shifts. Together with bad hygiene and safety conditions, the long hours were considered ‘a national disgrace’.39 In Barcelona’s cotton mills in the last quarter of the nineteenth century, 11 to 12 hours per day were common, but mills in smaller villages enforced a ‘working week between four and eight hours longer than in the Catalan capital’.40 The national overviews rarely make comparisons with other trades, but it seems that textiles all over the world had longer working hours than most other trades. This was explicitly mentioned in the case of the Barcelona cotton mills in the 1910s where women worked for 11 hours per day, as against a usual working day of 8 to 10 hours in most other trades. Working hours in workshops and manufactories did not differ much from those in modern textile mills. In Cairo’s workshops around 1900 spinners and weavers ‘worked an extra half shift until midnight or 1 a.m. to earn an extra 75 centimes 36 Blewett, ‘USA’; Komlosy, ‘Habsburg Monarchy’; Pretty, ‘Russia’. 37 This led to legal restriction of night work for women and children in many countries, as in Britain in the nineteenth century (Fowler, ‘Great Britain’), Germany around 1900 (Ebeling et al., Chapter 8, this volume, ‘Germany’), Japan in the 1930s (Hunter and Macnaughtan, ‘Japan’) and India in the 1920s (Roy, ‘India’). 38 Cliver, ‘China’. 39 Hunter and Macnaughtan, ‘Japan’. 40 Smith et al., ‘Spain’. 692

Work floors under tension or a franc’, and in Brazil’s manufactories around 1800 ‘Indian workers put in long hours’.41 In the workshops of Leiden and Den Bosch (the Netherlands) in the seventeenth century, weavers ‘worked long hours’.42 In the Macedonian workshops around 1850, the ‘workday averaged 15 hours in the summer and 10 in the winter, with a 35-minute break for dinner and none at all for breakfast’. Working days of 14 hours are reported in the silk workshops of Salonika during the 1840s, of 13.5 (summer) and 10 hours per day (winter) in Bursa, and ‘from sunrise to sunset’, six days a week in Damascus.43 Working time in a domestic environment is badly documented. Most likely, other tasks (work on the field, child-minding, food preparation and so on) hampered the continuous work on the loom or the spinning wheel. Often, textile work occurred only in winter, alternating with agricultural work during summer. It may be asked whether work control, effort and working hours differed for domestic weavers and spinners, wage labourers and independent workers, woollen, silk or cotton workers, men and women. There is hardly any research done in this field, and at best, national overviews mention that ‘workers at home could decide how much time they spent spinning or weaving’.44 I may suggest that working hours and effort were linked to the income evolution. This implies that work control, effort and the length of the workday were not necessarily harder for wage workers visà-vis independent workers, and that ‘self-exploitation’ of independent workers may have led to long working hours, severe self-discipline and harsh control of the ‘helpers’, usually family members.45 In fact, the fate of domestic workers cannot be judged on the basis of textile activity alone, but should be seen within the agricultural context of work, family and income. Working in a Kaufsystem (with the moderate influence of a merchant) involved agricultural work and income, and the spreading of risks. On the other hand, workers did run risks when buying tools and raw materials. The latter risk could diminish somewhat when a clothier or entrepreneur was involved, who supplied materials on credit. This credit could lead to higher dependence of the worker on the entrepreneur and thus the need to put in unrelenting effort.46 Working within the putting-out system, thus, implied less direct bonds with agricultural work and income, often leaving the worker with full-time textile work and greater dependence on price fluctuations of the produce. Action by workers, unions and politicians against long working hours appeared all over the world from the 1880s onwards, as, for example, in Spain in the 1880s, 41 Beinin, ‘Egypt’; Delson, ‘Brazil’. 42 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 43 Quataert, Chapter 18, this volume, ‘Ottoman Empire’. 44 For example, Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 45 Quataert asks whether there was ‘an “industrious revolution” in which people worked harder in order to buy goods’, answering cautiously that ‘indirect evidence indicates this to be the case’. 46 See L.A. Clarkson, ‘Domestic Industry’, in J. Mokyr (ed.), The Oxford Encyclopaedia of Economic History, vol. 2 (Oxford, 2003), pp. 99–102. 693

The Ashgate Companion to the History of Textile Workers in Russia in 1897 and in Mexico in 1912. In many countries, the workweek started to shrink from the 1880s onwards, reaching about 60 hours around 1910. The 1919 factory acts, supported by international agreements, led to a working week of 48 hours. In most countries, however, protest against long working hours had effect only in the 1920s and even the 1950s. During the wave of industrial disputes in Britain and Catalonia in 1918 and 1919, unions were largely successful in reducing working hours (sometimes entailing legal measurements, as with the Catalan Government decree of 1919 which established the eight-hour day). Yet, legal restrictions with regard to working hours had been established earlier in some countries.47 Social laws in Uruguay during the 1920s included a working week of 48 hours. But strikes in Bombay in the 1920s did not lead to shorter working hours. When Egyptian spinners and weavers at Misr struck and demanded the eight-hour day in 1938, they were severely oppressed and did not succeed. After the Second World War, demands for shortening working hours were formulated and often granted (as in Argentina, Egypt and Japan). The difference between harsh and more relaxed ways of work may be stressed. Chinese cotton and silk workers in the 1920s, for example, had the same working hours, but silk weavers enjoyed a more relaxed work environment, allowing them to have tea or even ‘listen to Shaoxing opera before they had to return to their loom’. Cotton workers had to cope with stress by ‘turning off the machine on some pretext, or going to the bathroom’.48 Catalan textile workers in the last quarter of the nineteenth century provide another example of more relaxed ways of working: ‘industrial discipline was not that tight’ and ‘workers were given 5 minutes leeway at the start of work, and were allowed to smoke, drink and eat in the factories’.49 Much of the ‘work comfort’ was linked to methods of pay.

Pay systems, wage gaps and evolution, and the family’s standard of living In 1909 a Copenhagen textile worker testified that the mill had a heating-cabinet for cooking a meal, but that workers lost time and money by using it, because ‘we were paid by a piece-rate’.50 This clearly makes the link between work control, effort and working hours on the one hand, and methods of pay and wage levels on the other. 47 For instance, the 10-hours act in Massachusetts in 1874 (Blewett, ‘USA’). 48 Cliver, ‘China’, original version as presented to the conference, quoted from http:// www.iisg.nl/research/textilenational.php (consulted 1 October 2009); Cliver, ‘China’. 49 Smith et al., ‘Spain’, original version as presented to the conference, quoted from http:// www.iisg.nl/research/textilenational.php (consulted 1 October 2009). 50 Christensen, ‘Denmark’ original version as presented to the conference, quoted from http://www.iisg.nl/research/textilenational.php (consulted 1 October 2009). 694

Work floors under tension Understandably, the national overviews devote much attention to wages. Wages were part of the production cost and, hence, central to employers who wished to lower the wage cost in an increasingly competitive industry. Throughout times and places, keeping low wages was thus central to employers’ policy. Wages were of course equally important to workers, whether as their sole income or as so-called additional income. Defining the wage seems unnecessary: it is the payment for a given amount of work. Yet, it is often unclear how many workers were involved in the production of one piece of cloth, and how long they worked at it. The Lancashire mule spinners of 1800 and the Indian handloom weavers of 1900 received a wage, but the former had to pay their helpers (piecers), and for the latter it is unclear who received precisely what. It is perhaps this complexity that makes (international) wage comparisons difficult. Another feature that adds to the complexity is the form of the wage. The national overviews mention money wages, except for workers in the north of Brazil in the seventeenth century, who were paid in cloth. In other countries, food and (more frequently) housing were provided, as was the case in Russia and China, and these should be seen as a form of payment. Despite such difficulties, some researchers have made international comparisons of textile wages.51 Since the sixteenth century and throughout the world, men and women in spinning and weaving have received piece wages (wages according to production units: kilograms of yarn or quantities of cloth of a precise size and type). However, some workers in spinning and weaving received time wages (a wage per day, week or month). The two payment methods appeared, for example, in the Khurunfish factory of Cairo in the 1810s: ‘there were three categories of workers: daily paid, monthly paid (mostly not artisans or manual workers but bureaucrats), and piecerate workers’.52 In general, those paid by time were commonly skilled male workers (particularly supervisors and engineers), although some women were paid a fixed wage too. In some countries in the twentieth century, a clear gender distinction with regard to pay systems appeared, with women paid according to piece-rates and men monopolizing time rate jobs, as in Chinese cotton mills. This mix of piece and time wages may be encountered in the Brazilian Rio Negro factory in the eighteenth century, Lancashire cotton mills around 1800, Japanese modern mills of 1900, and Danish mills in the twentieth century. Since there were more workers involved in production and paid according to piece-rates than supervisors, engineers or administrators who were paid by time rates, the piece wage was the most common pay system in textiles. This is confirmed by 51 F. Merttens, ‘The Hours and the Cost of Labour in the Cotton Industry at Home and Abroad’, Transactions of the Manchester Statistical Society (1893/4): 125–190; L. Varlez, Les salaires dans l’industrie gantoise. I. Industrie cotonnière (Brussels, 1901), pp. 137–157, 167–228; R. Gibson, Cotton Textile Wages in the United States and Great Britain: A Comparison of Trends 1860–1945 (New York, 1948). Since the late 1920s, the International Labour Organization has conducted comparative wages enquiries (see http://laborsta.ilo.org). 52 Beinin, ‘Egypt’. 695

The Ashgate Companion to the History of Textile Workers seventeenth-century Spain, eighteenth-century Japan, the early nineteenth-century USA, the nineteenth-century Ottoman Empire and late nineteenth-century Russia. Piece wages were also most common in Catalan textile mills of the late nineteenth century, as well as in Spain’s (semi-)illegal workshops of the 1970s. Piece wages, thus, were the rule in domestic textile production, workshops, manufactories and modern mills. Nonetheless, in some regions or workshops time wages were prevalent, as for weavers in Brazilian textile manufactories around 1700. The national overviews do not mention hourly wages, which, for example, would have replaced piece time rates in modern mills in the second half of the nineteenth century (as was the case in Belgian cotton mills).53 At this point it is quite difficult to find out about changes in pay systems and, particularly, about the causes of changes. For instance, did increasing international competition lead to a growing application of piece wages? In some countries, tariff lists were introduced with clear indications of the price for each type of yarn, cloth or whatever product. Famous were the Oldham and Bolton wage lists in the UK. In other countries, workers’ associations strove to obtain such lists, as, for example, in Denmark and Fall River (Northeast USA). In the twentieth century, minimum wages were introduced, as in Mexico and, much later, in Turkey, whereas in other countries wages were linked to the evolution of the retail price index, as in Britain and Belgium in the 1920s. Occasionally, the state intervened directly in wage formation. Prior to 1700, magistrates controlled England’s wool production, which included wage-setting. In late nineteenthcentury Denmark, the state took a more modest role, creating negotiating bodies and launching social legislation, which could influence both wage levels and general working conditions. Such a moderate position seemed to be common to many countries. However, Uruguayan social legislation during the 1920s fixed wages (and working hours), and Chinese legislation after 1949 introduced job security, labour insurance and better work conditions in general. The relatively simple pay systems, whether time or piece-rates, were often complicated by the application of fines, the deduction of specific costs or the introduction of bonuses. Fines were applied in, for example, Mexican mills in the 1860s, Russian factories in the 1890s and Chinese mills in the 1950s. Also, additional costs (for example a part of the wage taken for payment of guardians) could be deducted, as in Japanese mills around 1900. The latter mills used a sort of efficiency wage (‘the principle of relative efficiency’). British mills in the 1960s introduced bonus payments (to attract male workers for night shifts, which failed). A more complex method of pay was reported in Denmark in the 1930s with the Bedaux system that combined Taylorist principles of rationalization and efficiency to payment schemes. Income, work effort and working hours depended directly on the wage system. With time rates, a fixed amount was paid for an expected (negotiated or ‘usual’) quantity of work during a given period (week or month), which implied a relatively 53 P. Scholliers, Wages, Manufacturers and Workers in the Nineteenth-Century Factory: The Voortman Cotton Mill in Ghent (Oxford and Washington, 1996), pp. 126–129. 696

Work floors under tension relaxed way of working (although the work was thoroughly controlled). In general, (male) workers paid by time obtained relatively high wages. A disadvantage was that time wages hardly reflected productivity rises. Piece-rate wages, on the contrary, were directly linked to the worker’s output. The new machines and the reorganization of work, together with the pressure of trade unions since the nineteenth century, led to a gradual increase of piece-rate wages.54 However, and most importantly, the cutting of piece-rates very frequently accompanied the installation of new machines and/or the reorganization of work. Moreover, the diminishing of the piece-rate could be applied to increasing productivity: lower wage rates would lead to higher work effort if workers wished to retain their income level.55 The wage history of textiles since about 1800 reveals a continuous striving to cut piece-rates and workers’ protests against it. In periods of general economic downturn, wage-cutting was the rule, as for example in the Twente factories during the early 1930s. However, in many countries the productivity increase eventually led to a general increase in earnings. Wage-cutting was, of course, not new. Almost all the national overviews mention wage cuts under the influence of competition with foreign imports or, from about 1800 onwards, mechanical production in the home country. Such is noted, for example, for mid-nineteenth-century Austria, the nineteenth-century USA, nineteenth-century Damascus, 1930s Argentina and the Netherlands, and especially with regard to domestic spinning and weaving during the whole period. Most noticeably, ‘globalization’ since the 1970s has led to wage decreases. In some countries, specific political conditions eased the diminishing of wages, as in Uruguay during dictatorship. This downward pressure resulted in generally modest wage levels, as in China, Egypt, Japan and the Netherlands in various periods. However, some places and sectors offered higher wages, which implied rather large pay differentials within the industry. In Britain and China, the term ‘aristocrats’ has been used to refer to particular workers, with Lancashire mule spinners and Chinese silk weavers – both male roles, of course – receiving high pay and having relative autonomy on the work floor. Earnings for domestic workers in the Russian and Chinese countryside between 1800 and 1850, and 1850 and 1900 respectively were rather satisfactory too. In some cases, wages were raised as a premium on retaining skilled workers, as in early nineteenth-century Russia. Wages of textile workers in many countries, both women and men, were raised repeatedly under pressure of strikes, collective agreements and labour market conditions.56

54 This is documented by G.H. Wood’s wage statistics between 1850 and 1914 (Fowler, ‘Great Britain’). 55 Examples of the former are given by Smith et al., ‘Spain’; an example of the latter may be found in Britain in the 1920s (Fowler, ‘Great Britain’) and Austria in the mid-nineteenth century (Komlosy, ‘Habsburg Monarchy’). 56 Examples are given for Britain in the 1820s (Fowler, ‘Great Britain’), China in the late 1940s (Cliver, ‘China’), Mexico in the 1910s and 1920s (Bortz, ‘Mexico’), and Catalonia in the 1880s (Smith et al., ‘Spain’). 697

The Ashgate Companion to the History of Textile Workers The national overviews particularly stress wage gaps between men and women in all times and places. It was taken for granted that women were paid less than men. In Russia, ‘it was a given that women should always receive less pay’. Hence, saving on the wage cost meant often the replacement of men by women and girls (even if this implied the installation of more modern machines).57 Women replaced men, for example, in late nineteenth-century Russian mills, in the 1930s Chinese and Japanese mills, and in interwar Uruguayan mills. Diverse reasons for the structural wage gap between men and women are pointed at. The breadwinner ideology – a man had to earn a ‘family wage’, while (young) women were only earning ‘additional’ income – is mentioned for various times and places.58 This may be interpreted as plain discrimination. Also, explanations for the wage gap referred to physical differences between women and men: the latter could operate bigger machines. This would lead to higher wages for male weavers, despite the fact that piece-rates for male and female weavers were equal. This issue was noted with regard to the Danish textile industry between 1900 and 1940: men were able to get hold of the ‘best’ looms. From the late nineteenth century onwards, workers’ associations aimed at obtaining equal pay for women and men. This was achieved in the second half of the twentieth century, although wage gaps between men and women still remain. Finally, huge gender wage gaps can also be explained by the wide use of children, adolescents and particularly young girls in the labour force. Wage differentials also appear when the type of fibre, various occupations within a region, one mill or section of a mill, the size of the working place or the mills’ geographical location are considered. Thus, wages in Chinese cotton and silk mills differed widely, and the application of the multiple loom system in Denmark in the 1910s led to the increase of wage gaps within one mill. Intra-factory wage differentials grew in a Uruguayan mill in the 1940s when supervisors’ wages started to increase rapidly, and wages in modern mills surrounding Barcelona were about 25 per cent below those of the city’s mills, which was also the case with large towns and the surrounding villages in Germany. The fact that textile industries around the world employed many women and children contributed to the industry’s image of low wages. Yet, this image may also be upheld by the fact that the textile industry searched to establish itself in low-wage areas. The Dutch rural provinces of Twente and Brabant provide an example of this, where workers ‘accepted lower wages’,59 and, hence, the 57 A clear example of this may be found in the case of the Ghent cotton mill Voortman. To cope with the increasing competition in the 1870s, and because this manufacturer lacked sufficient capital to shift to more modern technology, he systematically enlisted young girls for the weaving section. Together with other measures, this lowered the wage bill (Scholliers, Workers, Manufacturers and Wages, pp. 63, 221–222). 58 The ideology of the male breadwinner was, for example, very present in Danish trade union debates in the 1890s, which reflected ‘a popular belief in society as such, among the employers and probably among many textile workers at the time’ (Christensen, ‘Denmark’, original conference version). 59 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 698

Work floors under tension persisting link with agricultural labour. More regularly, workers migrated from the countryside, where, save very rare exceptions, the general wage level was low. The national overviews rarely compare textile wages with those in other industries, which makes it difficult to generalize. Wages of adult male workers seemed to equal and even surpass those in other industries, which is exemplified by the so-called barefoot aristocrats in Lancashire.60 Apart from the mention of such well-paid workers, many national overviews suggest that although wages in textiles were usually low, the income of a textile worker contributed to the wellbeing and even prosperity of the family. This had been the case for Austrian and Balkan handloom weavers before the coming of mechanization. Prior to 1850 in Russia, wages were higher than they would be later in the century, which may be explained by the high profits of the time and by premiums on retaining experienced labour.61 In Chinese cotton mills, with long hours, harsh conditions and low pay for female workers, it was noted that earnings were ‘fairly attractive compared with other forms of employment’. Even Chinese handloom weavers in the 1920s got a ‘decent living for themselves and their families, [who earned] several times the annual wage of an agricultural labourer’.62 The same is noted with regard to Indian cotton factories of the second half of the nineteenth century: ‘the average wage was usually better than agricultural labourers’ earnings, but too little or too insecure to think of growing roots in the city and giving up connections with land’.63 The latter, however, seemed to have been possible for workers at the top of the wage hierarchy, who usually enjoyed a lifestyle better and more secure than that in the villages. In Japan in the late nineteenth century, with low wages, poor conditions and high turnover, women’s work was considered as ‘supplementary’ to the family income. Yet, for some it was possible to earn wages sufficiently high to become independent from the family. The same occurred in Chinese silk factories in the interwar period, where young girls could earn an income that sufficed to support their whole family, or permitted them to move in with other mill workers, ‘wear fancy coiffures or embroidered shoes, and … go shopping or to see an opera with friends’. Young male silk weavers in Shanghai in the 1920s ‘looked upon factory life as a step up in the world, and modelled their dress, customs and politics on urban students their own age’.64 In nineteenth-century Mexico, ‘textile work … represented the possibility of a better life’, and the late 1920s were characterized as a ‘time of extraordinary prosperity for textile workers’.65 60 However, Fowler, ‘Great Britain’, stresses the fact that Lancashire’s relative prosperity was not so much based on individual wages, but on the family’s income pooling to which father, mother and children contributed, and which permitted the family to visit Blackpool, the seaside resort of Lancashire cotton operatives and the only seaside town in the world based on working-class prosperity. 61 Pretty, ‘Russia’. 62 Cliver, ‘China’. 63 Roy, ‘India’. 64 Cliver, ‘China’. 65 Bortz, ‘Mexico’. 699

The Ashgate Companion to the History of Textile Workers Wages may have been ‘decent’ and even ‘satisfactory’, but it should not be forgotten that they were obtained in harsh conditions. Moreover, the national overviews often indicate falling real wages in many countries during many periods. In general, real wages tended to rise in the long run, as it is shown by the example of Indian handloom weavers between 1880 and 1940, Catalan cotton mill workers up to the 1930s and US workers particularly after 1945. However, falling real wages are reported in Britain during the 1930s, Egypt in the 1940s and 1970s, Mexico in the 1940s and 1950s, Spain in the late 1930s, 1940s and 1950s, and in Uruguay in the 1970s. Particularly bad conditions are reported in regions with huge domestic production confronted with the competition of mechanical production, as in nineteenth-century Austria where falling real wages led to poverty.

Connecting textile centres of the world: some conclusions and wishes Competition (between various types of production modes in one region or country, and with modern mills and/or cheap foreign import) modelled working conditions in the textile industry. The wish to achieve low prices and raise productivity led to continuous mechanization, work reorganization, extension of working hours and/or the introduction of night shifts, the search for cheaper labour, stricter work control and demand for increasing work effort. Workers escaping this logic were producers of luxury goods, such as wool in the eighteenth century or silk up to the early twentieth century. The others, the bulk of textile workers, were forced to pool incomes of family members. Some managed not only to survive, but even to enjoy a rather satisfactory standard of living. With regard to wages, two contradictory trends appeared in many countries from the late nineteenth century: one led to diminishing incomes (due to ever-increasing international competition) and the other implied increasing incomes (due to general economic developments and increased productivity). Who got what depended on power relations on the work floor, as well as on manifold conditions outside of the textile industry. Lines may be drawn between the various textile centres of the world, directly connecting changes in one region to working conditions in another. These are most apparent when wages are considered.66 However, such connections should not be exaggerated, because not all changes can be explained by (anxiety about) international competition. Local, regional and national customs and practices continue to play a prominent role in the shaping of work conditions. 66 For example, around 1880 some Ghent cotton barons started to move their weaving sections to the countryside where wages were about 30 per cent below those in the city. They explained this as a reaction to wage cuts in English factories in the 1870s and early 1880s, which in turn were linked (by the Ghent manufacturers) to the expanding cotton industries in Germany, India and the USA. 700

Work floors under tension On the heuristic level, I may suggest further research that would investigate the workplace. Working conditions were (and are) highly determined by the material conditions of work, that is the structure of the building, the various facilities (rest rooms, cooking and eating spaces), the way the machines were conducted and so on. In short, aspects of industrial archaeology and material culture should be incorporated in the study of working conditions. The pay system is another area worthy of further attention. Wages were crucial, and not just with the coming of the modern mill: they determined the work effort, working hours, the number of workers and standards of living. The way workers were paid needs to be explored further. This implies attention to the measurement of daily or weekly production, the frictions around that measurement, negotiations about tariffs, individual and collective wage conflicts, changes with regard to pay systems (time versus piece wage), bonuses and fines, and so on. And finally, more attention may be paid to work conditions in domestic industry. Kauf- and putting-out systems, and other forms of domestic work organization are well known, but the actual work conditions (work effort, working hours, number of workers involved, actual pay, costs and so on) remain to be investigated.

701

This page has been left blank intentionally

28 Gender and the global textile industry Janet Hunter and Helen Macnaughtan

It is strongly emphasized in a number of the national overviews that there was no such thing as a common textile worker experience even within a single national or regional entity. Rather, a diversity of forms of production, workforce composition and economic and commercial environments combined to produce a range of historical experiences across time and space. Some of the most fundamental divisions within and between workforces were related to the use of male and/or female workers, and constructions of gender impacted on, and were influenced by, the coexistence of and competition between different forms of production and the industry’s response to the economic imperatives that it faced. Commercialization and industrialization, often associated with the gradual growth of wage labour and movement of textile production away from the household, had significant implications for the gender identity of textile work, but this process was far from being always the same in different countries and regions. The importance over time of the textile industry globally has meant that employment in textiles has at certain periods, particularly before and during the course of a nation’s industrialization, accounted for a significant share of the national labour force. For example, in the mid-nineteenth century, an estimated 10 per cent of the Japanese population worked in textile production, while in the  For some other recent works on women, gender and the textile industry, see, for example, Emily Honig, Sisters and Strangers: Women in the Shanghai Cotton Mills, 1919–1949 (Stanford, 1986); Miriam Glucksmann, Cottons and Casuals: The Gendered Organisation of Labour in Time and Space (Durham, 2000); Wendy Gordon, Mill Girls and Strangers: Single Women’s Independent Migration in England, Scotland, and the United States, 1850–1881 (New York, 2002); Janet Hunter, Women and the Labour Market in Japan’s Industrialising Economy: The Textile Industry before the Pacific War (London, 2003); Jutta Schwarzkopf, Unpicking Gender: The Social Construction of Gender in the Lancashire Cotton Weaving Industry, 1880–1914 (Aldershot, 2004); Helen Macnaughtan, Women, Work and the Japanese Economic Miracle: The Case of the Cotton Textile Industry, 1945–1975 (London, 2005).  Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’.

The Ashgate Companion to the History of Textile Workers early 1930s some 50 per cent of the industrial workforce in Shanghai was employed by the textile industry. In 1950s Lower Austria and in Uruguay, 16–17 per cent of all industrial workers were in textile production. While the significance of the textile industry’s workforce cannot be denied, what is of interest to this chapter is the gender composition of that workforce and the differences and similarities in gendered aspects of textile work. The chapter will also consider the extent to which the gendering of textile work embraced characteristics that are more widely found in the history of manufacturing as a whole. The textile industry has generally been viewed as an industry suitable for the employment of female workers. In part this has been due to the tradition of textile production (particularly spinning) as ‘women’s work’ in the pre-industrial household economy, and later its classification as a light (rather than heavy) manufacturing industry (prior to advanced mechanization and the rise of synthetic textile production), therefore seen as suitable for the ‘nimble hands’ of female workers. We also find widespread the view held by textile producers that women workers were both a more ‘docile’ and a ‘cheaper’ alternative to male employment and were readily available as an expendable and/or untapped labour pool. In Japan, for example, primarily young females outnumbered males in the industrial workforce up until the 1930s because of the large numbers of ‘daughters’ released from the agricultural household into the growing textile industry, while in the 1960s the industry turned to the employment of ‘housewives’ as an unused labour force within the constraints of a rapidly expanding Japanese economy and growing labour shortage. The gendered nature of textile employment overall is certainly a theme that emerges from many of the national overviews in this study. However, statistical data on the gender composition of textile employment in the past are often difficult to unearth, particularly when considering the pre-industrial industry. The same can be said for the availability of data providing a further breakdown by age and marital status. The national overview on Germany notes that, although overall data are lacking, it can be assumed that women and girls formed the majority of labour in pre-industrial spinning, while the Netherlands chapter on the pre-industrial industry notes that the practice of only registering household heads in official records results in an under-registration of married women and children working in textile production. This is an important point when considering the gender composition of the textile workforce over time, as females may well have been unrecorded or unrepresented in historical sources on textile employment due to their very gender as ‘female’ within a history centred

 Cliver, Chapter 5, this volume, ‘China’.  Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’; Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’.  Hunter and Macnaughtan, ‘Japan’.  Ebeling et al., Chapter 8, this volume, ‘Germany’.  Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’. 704

Gender and the global textile industry around patriarchy. Even in post-Second World War Japan, with its vigorous recording of industry data and compiling of industry histories, data on women workers (particularly ‘older’ married women) are often un(der)-recorded especially when they are viewed as temporary or supplementary workers. It is fair to say that in many of the countries represented in this study, female workers constituted at the very least a significant presence and at the most a majority in the textile workforce. Estimates suggest that 51–92 per cent of spinners in the sixteenth–eighteenth-century Netherlands were female, as were up to 80 per cent of textile workers in eighteenth-century Lower Austria.10 Female workers dominated in the industrialized textile workforces of Europe and in the USA, Japan and Argentina. However, while most of the authors mention significant proportions of female employment in textile production, there are some notable exceptions. Textile mills in nineteenth-century Mexico comprised mainly male workers (though women did work in smaller workshops as seamstresses) and it was not until US companies set up textile plants during the 1960s that female workers were utilized as a main workforce in Mexico.11 In India, through to the 1920s, handloom weavers were primarily male, and it was males who tended to migrate to work in the cotton mills that developed from the second half of the nineteenth century.12 The development of wage labour during the twentieth century de-feminized textile work, and it was only right at the end of the century, with the setting up of new spinning mills in semi-rural areas, that this process began to some extent to be reversed. The presence of women on the factory floor was often determined by social constructions of gender over time in any one nation, and whether women were ‘permitted’ to work in large numbers outside of the home. The composition of the textile workforce by gender was also determined by the industry’s position within the national (and global) economy over time. A core male workforce often predominated when the industry was commercialized in pre-industrial economies, while mass production under industrialization appeared to give way to ‘cheaper’ and ‘easily sourced’ female and young labour. In many industrializing nations the ‘model’ employment structure when setting up modern textile mills appeared to be one which employed small numbers of male technical staff and large numbers of female factory operatives. It would be fair to suggest that during and following the process of industrialization, a significant majority of textile factory operatives around the world were characterized as female, young and initially of rural origin. Many of the countries under consideration were characterized in the pre-modern period by a process of rural industrialization  For general problems in measuring the historical problems of women’s work, see, for example, B. Hill, ‘Women, Work and the Census: A Problem for Historians of Women’, History Workshop 35 (1993): 78–94.  Macnaughtan, Women, Work and the Japanese Economic Miracle, pp.19–20, 22–23. 10 Komlosy, ‘Habsburg Monarchy’; Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 11 Bortz, Chapter 13, this volume, ‘Mexico’. 12 Roy, Chapter 10, this volume, ‘India’. 705

The Ashgate Companion to the History of Textile Workers (in some cases ‘proto-industrialization’). Women became increasingly dominant in Lower Austrian textile districts in the eighteenth century as producers felt the full weight of British competition. That pattern was replicated not just in Austria at a much later date, but in a range of countries through the eighteenth and nineteenth centuries. The US rayon industry in the first half of the twentieth century followed this strategy, as did Indian textile producers in the 1980s.13 The gendered aspects of production varied according to the region and the mode of production. An urban–rural distinction, originating in pre-industrial production, led to male dominance of artisanal production and female dominance in much rural production. This distinction continued to manifest itself in certain respects throughout the period up to the present, and is discussed below. In particular, the ongoing attempts by textile producers in many countries to turn to rural areas for their labour supply was premised on the assumption that rural labour was cheaper than urban, and that rural women were even cheaper. In some cases this pursuit of cheap rural female labour was associated with relocation. In others it brought temporary or permanent migration of workers from rural areas, or the persistence of subcontracting and putting-out, and home-based production, such as in Turkey and Uruguay during the twentieth century.14 This process has only slowly undermined regional distinctions, but it has certainly influenced regional patterns of production, and acted as a driving force in the development of textile production overall across a large number of very diverse producing countries. The gendered nature of the workforce was also determined by the economic position of the industry and any decline in that position. Curtailment of production or economic downturn has meant the industry could be staffed by older women (and men) as was the case in England during the Second World War.15 Globalization in the late twentieth century has again produced a phenomena of a core female workforce, seen as cheap and easily laid-off when textile production loses its glow, as witnessed in 1960s Mexico and 1990s Austria.16 This notion of female labour being ‘supplementary’ to core male labour is a notion that has continued over time, even in an industry such as textiles which has historically had a larger share of female workers than other (manufacturing) industries. This often leads to the female composition of the textile workforce being considered ‘supplementary’ and ‘temporary’ as was the case in the nineteenth- and twentieth-century Japanese industry, or ending up as ‘informal’ or on the ‘periphery’ as in the textile piecework sector of 1970s Uruguay.17 This begs the question of how gender inherently shaped characteristics of the textile labour force over time and how much power the textile labour force was able to wield throughout history as determined by its gendered make-up.

13 Roy, ‘India’; Blewett, Chapter 21, this volume, ‘USA’. 14 Seidman, Chapter 19, this volume, ‘Turkey’; Camou and. Maubrigades, ‘Uruguay’. 15 Fowler, Chapter 9, this volume, ‘Great Britain’. 16 Bortz, ‘Mexico’; Komlosy, ‘Habsburg Monarchy’. 17 Hunter and Macnaughtan, ‘Japan’; Camou and. Maubrigades, ‘Uruguay’. 706

Gender and the global textile industry

Pre-industrial, rural–artisan distinction It is apparent that textiles have a long history of production in most economies, and that both men and women were closely involved in textile production in the pre-industrial period. However, the gender division of labour in pre-modern textile production showed considerable variation across countries, in conjunction with varying constructions of the gender identity of particular aspects of textile work. In pre-colonial and colonial Argentina, and among the indigenous producers in Mexico, weaving was identified as women’s work.18 In colonial America women were involved in all stages of the production process. More widespread appears to have been the identification of spinning as a mainly female occupation, and weaving as a predominantly male occupation, a pattern that characterized not merely much of pre-modern Europe, but also Egypt, India and the non-indigenous producers in Mexico. A major feature of most of these patterns of pre-industrial production is the dominance of the household as the production unit and the frequent coexistence of both domestic and artisanal production. Both these features had significant gender implications. Firstly, since in most households it was a male who was the family head, that hierarchy tended to be reflected in the family’s organization of production. Secondly, the specialist artisans who often produced the highest valueadded textiles, and the apprentices whom they trained, were invariably male, as demonstrated by the cases of Denmark and the Habsburg Empire in Europe, and further afield in both India and Japan.19 Artisan textile production was in many cases controlled by guilds, which were again invariably restricted to males. Guilds and equivalent bodies therefore became in turn a form of gendering, helping to reinforce male dominance of certain areas of production. In Denmark and the Netherlands, as well as in Egypt and the Ottoman Empire, guilds were highly significant in controlling the male-dominated weaving industry. Whereas in both rural households and urban workshops the pre-industrial male weaver was often referred to as a ‘master’ or ‘artisan’ (and had often undertaken an apprenticeship and/or was a member of a guild), this formal recognition of the male weaver (and his skill) was not something generally accorded to women. Spinning, carding and other processes performed by women were often seen as complementary to her other key tasks of household chores and child-rearing. These two features had particular implications for the gender division of labour within the industry. While female labour still formed a crucial part of householdbased textile production, male labour was often dominant particularly in the case of weaving. Women were frequently excluded from the higher value-added end of textile production (often weaving) and, perhaps more importantly, from the status and training that was associated with it under the guild and apprenticeship system. While there is evidence that females were involved in the administration 18 Lobato, Chapter 2, this volume, ‘Argentina’; Bortz, ‘Mexico’. 19 Christensen, Chapter 6, this volume, ‘Denmark’; Roy, ‘India’; Hunter and Macnaughtan, ‘Japan’. 707

The Ashgate Companion to the History of Textile Workers and marketing of textiles, and in Egypt, India and Japan, for example, a male artisan might work alongside other family members, there is also widespread evidence that female family members were under such circumstances frequently employed in subsidiary tasks (including spinning or winding). A common division of labour in the rural agricultural household appears to have been one where female labour dominated the spinning of yarn, male labour presided over the weaving of cloth, and both female and child labour performed preparatory and auxiliary tasks for male weavers. This ‘model’ of the central figure of male weaver as household head, supported by the labour of his wife and children, was evident in seventeenthcentury Spain, in the seventeenth-century Hapsburg Monarchy, in eighteenthcentury Denmark and in seventeenth- and eighteenth-century India. Similar patterns of work can be identified within some areas of domestic production, where the male household head was the main textile worker, supported by other family members. However, in many, if not more cases, domestic production was mainly carried out by females, while the males of the household focused on agricultural production. In pre-modern Brazil, for example, there was longstanding female involvement in textile production and trade in the home,20 and from Denmark and Spain in Western Europe, through the Ottoman Empire and Egypt as far as India and Japan, textile production was commonly undertaken as female by-employment within agricultural families. In Qing China the capitalist organization of silk weaving employed males, while it was rural females who carried out cotton weaving in the context of the family economy.21 While we know that family economies had to respond flexibly to changing imperatives, and that the internal division of labour did vary over time, the case of the Habsburg Empire suggests the existence of a sharper gender division of labour in artisanal households than in those of domestic producers. The increased specialization and division of labour associated with industrialization was to take this gendering process further.

Family-based production and its persistence into the twentieth century A key feature of the process of commercialization and mechanization of textile production that has characterized the last 250 years is the persistence of existing ways of organizing production, and frequently the coexistence of mechanized factory production with handicraft activity, often conducted within families. It is apparent that there was no simple linear relationship between handicraft and mechanized textile production, in which the former inevitably gave way to the latter. Even in Britain, the rapid growth of textile factories in both wool and cotton 20 Delson, Chapter 4, this volume, ‘Brazil’. 21 Cliver, ‘China’. 708

Gender and the global textile industry only gradually made inroads into the old handicraft practices, and a spectrum of organizational forms persisted, in which the family economy, and the household as a unit of production, retained considerable significance. This was even more the case in many other economies, and has persisted into the start of the twentyfirst century. That persistence was supported by a range of factors, including the flexibility and adaptability of family economy strategies in many countries, an adaptability in which the women who often dominated domestic production invariably played a major role. In nineteenth-century Brazil, for example, factory production of textiles began to emerge, but women’s participation in domestic textile production remained far greater than in any of the new mills.22 A positive interaction between new and existing forms of production was often manifest early in the industrialization process. Early mills in North America put out yarn to be woven by rural women, and rural putting-out systems can be found not only in Qing China, but in nineteenth-century Egypt, and in India and the Middle East through the twentieth century.23 Handicraft weaving by both men and women was often given a new lease of life by the production of more even and higher quality machine spun thread. In addition, however, it became apparent that household and domestic production enjoyed continuing comparative advantages in many areas of textiles, and those families that demonstrated sufficient flexibility to exploit that comparative advantage could continue to play a significant role in the industry. In the Habsburg Empire, and in Egypt, home-based women carried out labour intensive processes or those demanding specialist skills.24 Homeworking often expanded alongside the factory system, as unpaid or underpaid female family members offered seasonal and other flexibility, and a buffer labour force in times of depression or stagnation. As some forms of domestic production became less competitive in the face of factory or import competition, household producers turned to new areas of textile production, or those in which they continued to hold comparative advantage. Homeworking and sweatshops have been widespread in the garment industry in many European countries, and Egypt even developed a formal apprenticeship system for individual seamstresses.25 The advantage of the family economy in the face of the growth of capitalist manufacturing production essentially lay not just in the flexibility of the response it offered, but in its ability to use unpaid labour, and the labour with the lowest opportunity cost at any particular point in time. Household strategies, particularly those of households in rural areas, invariably focused on the collective earnings of the household, and not just the income of an individual member, and this strategy was concerned with diversifying sources of income (thus reducing risk), was willing to combine paid and unpaid labour, and also to contemplate the concept of ‘supplementary earnings’ for females. This priority is manifest not only in the Spain of the eighteenth century and nineteenth-century Japan, but more recently 22 Delson, ‘Brazil’. 23 Cliver, ‘China’; Beinin, Chapter 7, this volume, ‘Egypt’. 24 Beinin, ‘Egypt’. 25 Ibid. 709

The Ashgate Companion to the History of Textile Workers in India, and even the Austria of the 1980s–1990s.26 Even in the context of the strengthening of the male breadwinner model that occurred in many industrializing economies in the nineteenth and twentieth centuries, the cases of the UK, the US and Japan demonstrate that collective family earnings often remained as important a consideration as the adult male wage.27 On balance, it seems that the persistence of family-based production and the family economy served to underline the significance of women in textile production even where mechanized production might be male-dominated. Where the factory sector grew, the family, and hence the gender aspects of the family, exercised a major influence on its operation. In the textile mills of late nineteenth- and early twentieth-century Japan and China, for example, the family was a key player in the decision of an individual to take up employment, and here and elsewhere – for example in Uruguay, India and the Ottoman Empire – family connections were of persistent importance in both recruitment and work patterns.28 Through these means the dominance of the family, and the patriarchal relations that so often characterized it, became transposed to the factory, a process very evident from the experiences of the US, the UK and Japan. Gendered social hierarchies replicated themselves in the new workplaces, helping to establish an environment in which women were more likely to be identified as subordinate members of the workforce and those with domestic responsibilities were expected to perform them in addition to any paid or unpaid work. The burden of domestic responsibilities was particularly great where married women moved into wage work and employment outside the home, a feature not just of Western economies such as the UK and Denmark, but also of Asian economies such as China and Japan. Such tensions could lead to married female textile workers being regarded as less reliable, a situation that might lead to attempts to replace them with males or with short-term, unmarried females, or to employ them in less desirable tasks. A growing tendency in many countries in the nineteenth century to associate women with ‘domesticity’, and to identify work outside the home with maleness served further to undermine many women’s position in paid work, and reinforce their family-based identity. Capitalist development in general tended to idealize the family, associating femininity with love and emotional bonds, and men with earning and with economic and power relationships. This shared ethos helped to shift the position of the family from a cooperative, flexible unit of production to a unit characterized by more rigid and hierarchical individual functions, and with, over time, a lesser economic role.

26 Smith et al., Chapter 17, this volume, ‘Spain’; Komlosy, ‘Habsburg Monarchy’; Roy, ‘India’; Hunter and Macnaughtan, ‘Japan’. 27 Fowler, ‘Great Britain’; Hunter and Macnaughtan, ‘Japan’; Blewett, ‘USA’. 28 Cliver, ‘China’; Roy, ‘India’; Hunter and Macnaughtan, ‘Japan’; Camou and. Maubrigades, ‘Uruguay’; Quataert, Chapter 18, this volume, ‘Ottoman Empire’. 710

Gender and the global textile industry

Production, technology and skill – feminizing and masculinizing the workforce The process of (re)defining the labour force composition in the textile industry by gender can be thought of as a process of either feminization or masculinization. There are several reasons why these processes took hold in many of the countries studied, but technological change was an important part and was often accompanied by changes to the national economy within which the textile industry operated.29 Not only the importance of the household and its gendered operation, but also the pre-industrial gender division of textile labour had a significant impact on the gendering of the new mechanized forms of textile production. In terms of technological innovation, the industry was generally female dominated when the processes were very ‘simple’ and done by hand. With the advent of semi-automation in the proto-industrial and early industrialization eras, masculinization would occur as men took over many of the tasks previously performed by women, using new technology. This would tend to reverse again once industrialization advanced, as automation and mass production allowed for a (re)feminization of the workforce to take place. Similarly, the position of the textile industry within the overall economy of a nation encouraged the same process. With commercialization in the pre-industrial economy, adult males took over weaving from women in handicraft and became masters, artisans or a ‘labour aristocracy’ who were often well paid and operating in exclusive guilds. However, under both mechanization and proletarianization, particularly as mass production in the nineteenth and twentieth centuries designated textiles a labour-intensive, light manufacturing industry, factories moved from employing men to ‘de-skilled’, lower-waged women and children. There was never any simple relationship between technology change and the division of labour in textile production, and technology change was always part and parcel of the ongoing social construction of ‘male’ and ‘female’ jobs, but a number of general points may be made in relation to the gendering of technology and skill, and the consequent feminization or masculinization of jobs and sectors. Firstly, mechanization has in many cases been associated with the feminization of the workforce. In the case of China, for example, as the cotton and silk industries shifted to mechanized production, it was male jobs that were lost, and female jobs that were created.30 Nevertheless, in India it was the work that was done by females that was potentially more susceptible to mechanization, suggesting that in this case it was female jobs that were likely to be lost.31 It is clear, however, that in many cases 29 For some interesting case studies, see G. de Groot and M. Schrover, Women Workers and Technological Change in Europe in the Nineteenth and Twentieth Century (London, 1995) and L. Sommestad, ‘Gendering Work, Interpreting Gender: The Masculinization of Dairy Work in Sweden, 1850–1950’, History Workshop 37 (Spring, 1994): 57–75. 30 Cliver, ‘China’. 31 Roy, ‘India’. 711

The Ashgate Companion to the History of Textile Workers the introduction of new technology permitted cheaper female workers to replace more expensive male ones. The introduction of ring spinning in countries such as Japan, the UK and the US offered an opportunity to employ more unskilled females, while the introduction of power-looms removed some of the requirements for skill and strength that had helped male weavers to retain their jobs.32 However, as the nature of jobs changed, and along with it the identity of those who performed them, those same tasks became recategorized in terms of ‘skill’ designation and pay. Mule spinning in the UK was undertaken by male workers.33 With spinning equipment that no longer required such skill or strength, nor the employment of a number of assistants, females could be employed, but the task came to be constructed as low skill, ‘female’ work. Such changes were, however, often fiercely resisted. Technical change in the US Fall River mills undermined male spinners’ authority as men supervising other adults, posing a challenge to their very identity as male workers in possession of some status.34 The association of female labour with low skill, absence of qualifications and training, and low pay was not a new phenomenon associated exclusively with the development of mechanized production. As noted earlier, pre-modern apprenticeship was invariably associated with males, while women were excluded from many tasks and were often engaged in ‘supplementary’ processes in support of a core male worker. At the same time, the evidence from Denmark suggests that some pre-industrial female textile workers were relatively skilled.35 Moreover, the advent of rapid technological change posed new challenges both to the organization of textile production and to social norms and constructs of gender. Male and female workers in early factories in Japan and Argentina, for example, were often the subject of acute complaints from employers concerning the general lack of discipline, their reluctance to engage in hard work and their inclination for absenteeism.36 What is clear is that the process of technological change, whether imported or endogenous, had the capacity to change the relations of production and promote an increasing division of labour. In that process many scholars accept that a degree of ‘de-skilling’ took place, in which both males and females suffered. We need to ask, however, how meaningful the concept of skill, and hence de-skilling, is. In responding to this challenge, whether real or perceived, it was male workers who were normally able to arrogate to themselves the highest paid, highest status jobs. They were also able to reconstruct the notion of skill to identify it with masculinity. Skilled jobs were by definition those undertaken by males. If a female worker undertook a task, it could not be skilled work, since females were by definition untrained and unskilled. In times of crisis, it is suggested for the Netherlands, there may have been a process of male institutionalization of inferior female jobs,37 but more visible has been the 32 Fowler, ‘Great Britain’; Hunter and Macnaughtan, ‘Japan’; Blewett, ‘USA’. 33 Fowler, ‘Great Britain’. 34 Blewett, ‘USA’. 35 Christensen, ‘Denmark’. 36 Lobato, ‘Argentina’. 37 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 712

Gender and the global textile industry tendency within textile production everywhere for men to command the specialist skilled jobs such as engineer or foreman, and for women to take the burden of shop floor production. The process of technological change in conjunction with an ongoing reconstruction of gender in the workplace therefore established the foundations for a less flexible gender division of labour and justifications for that division of labour that have continued to be articulated through to the present. The emerging availability of alternative manufacturing jobs in an economy also ensured a feminization of the textile industry, as men were more likely to move into higher paid and heavy sectors of manufacturing. Women, on the other hand, often remained longer in the textile industry and were less likely to be employed in heavier manufacturing industries. On the flip side, when there were fewer alternatives to textile production or when jobs were scarce under economic downturn, male workers tended to remain or reclaim textile jobs from women. In such cases, they were often found to occupy a more powerful position in the process itself, ensure that the environment favoured them and carry out a process of assigning better status to their jobs via titles like ‘master’ or through the formation of guilds and later unions. These were processes that were far more likely to happen under masculinization than under feminization. Similarly, as demonstrated by the case of Mexico, contraction and/or decline of the industry meant that textiles moved from being a booming industry comprising well paid, skilled and unionized male workers in the nineteenth and twentieth centuries to an industry of ‘unskilled, non-unionized, low-paid women workers’ under globalization.38 And in the USA, factories employed non-unionized black, Asian and Hispanic females under globalization, displaying not only a feminization but also a ‘race to the bottom’ construction of the workforce under economic pressure.39 A common reason for feminizing the textile labour force was therefore economic and the inherent costeffectiveness of female labour, with female wages classified as lower and ‘cheap’ by virtue of gender. This cost advantage encouraged textile employers to feminize the workforce in their factories. While the nature of the industry was deemed suitable for the employment of women, as Quataert’s chapter on the Ottoman Empire points out: ‘such massive participation of women and girls in expanding sectors of textile production was no accident, nor was it because of nimble fingers … [It] was part of the process of lowering labour costs in the struggle for competitive survival.’40 The notion of gender itself played a large role in shaping the process of either masculinization or feminization, and there were social and cultural reasons for gendering the textile workforce. We know from looking at the US and Brazil that early factories separated their workers by gender, although in other countries, including Japan, women and men worked side by side in the first cotton mills. However, the participation of men and women in a single industry was a sine qua non for the emergence of a gendered division of labour. The gendered division of labour interacted with the process of technological change to produce a division 38 Bortz, ‘Mexico’. 39 Blewett, ‘USA’. 40 Quataert, ‘Ottoman Empire’. 713

The Ashgate Companion to the History of Textile Workers of labour between the sexes that was far from being equal, but which also, as Roy notes for India, had the capacity to lead to an overall higher female workforce participation rate.41 While pre-industrial household production was generally female-dominated or broadly divided between the sexes, entering employment in workshops or factories outside of the home was not deemed suitable work for women in some countries at certain points in history – for example China in the late Ming period and India prior to the 1970s.42 Another reason for feminization of the textile workforce was the seemingly inherent nature of being ‘female’. This can be demonstrated by the Shanghai mills in the 1920s which substituted strikeactive male workers with female workers and the USA mills in the 1860s which turned to Irish immigrant females because they would provide a more ‘docile’ and ‘stable’ labour force.43 In Brazil and India, few women were employed in early factories and, when they were, were segregated on the shop floor in different areas from the male workers.44 In Egypt, males ‘fiercely opposed the entry of females to their workplace. Some even physically attacked them’, while new factories (in this case for the making of hats) were set up with female labour because they were seen as carrying on a ‘tradition’ of female work and therefore did not ‘violate the established gender division of labour’.45 Over time, the social exclusion of women from some textile sectors was often legitimized, which allowed for a change in cultural and social norms and a feminization, or at the very least ‘incorporation’ of females, within the industry’s workforce. In 1860s China, male lineages and local male elites influenced and legitimized the employment of girls in textile factories, while declarations from religious authorities in the Ottoman Empire and foreign female ‘role models’ helped persuade girls to enter textile employment.46 Thus, during the course of textile history there has been a continual process of negotiating and constructing ‘feminine’ and ‘masculine’ roles within the industry. In Japan, as in Denmark and Egypt, for example, prevailing assumptions about what constituted men’s and women’s work were brought to the factory. However, new ways of organizing production could also offer opportunities for changing the gendering of textile work. Beinin suggests that the growth of the new industries in nineteenth-century Egypt had the capacity to change the gender division of labour, as symbolized by the tarbush (fez) industry with its female workforce.47 In the UK, by contrast, some workers sought to establish new trades as being in the artisan tradition, hence preserving them for males. The fact that in most pre-industrial economies women had engaged in some form of textile work did not necessarily mean that factories replicated existing divisions, although they often reflected them. The consequence was that the gendering of textile factory production during 41 Roy, ‘India’. 42 Cliver, ‘China’; Roy, ‘India’. 43 Cliver, ‘China’; Blewett, ‘USA’. 44 Delson, ‘Brazil’; Roy, ‘India’. 45 Beinin, ‘Egypt’. 46 Cliver, ‘China’; Quataert, ‘Ottoman Empire’. 47 Beinin, ‘Egypt’. 714

Gender and the global textile industry the nineteenth and twentieth centuries varied according to country and region, and over time. In nineteenth-century Mexico, for example, textile production was seen as ‘male’, whereas the late nineteenth-century silk filatures that proliferated in Japan, China and the Ottoman Empire were worked almost entirely by young women. Within England, the cotton, woollen and worsted industries embraced very different workforces with very different gender compositions, while in China the gender division of labour varied from factory to factory depending on the nature of native place ties and the extent of the workers’ handicraft background.48 In textiles, however, there were very few sectors of production that did not involve both women and men, although there were some that employed predominantly one sex or the other. In this respect textiles, through its employment of a large number of females, may not necessarily be representative of manufacturing as a whole. Nevertheless, as far as the gendering of the work process within industries and within factories is concerned, the case of textile production is not necessarily so distinct from other industries.

Gender, wages and the workplace environment Gender played a key role in determining the level of skill and the level of payment in the majority of national textile industries. The same task or process when transferred from male to female worker or simply performed by a female was generally considered less skilled and deserving of a lower payment by virtue of the operative being female. Women often earned as little as 45–50 per cent of male wages under fixed wage structures. Even when nominal equality existed, for example when both sexes were paid using the same piecework levels, in reality gender differences within the workforce persisted. Women generally earned less than men because of their lower output, often due to men being given better quality materials and equipment. While male and female cotton weavers in Lancashire in theory received equal rates, males were assigned cloth types that brought higher rates of pay.49 The work was never equal. Overwhelmingly in textile factories women were increasingly assigned to tasks that were inferior in terms of status, pay, training provision and prospects. The division of tasks was never constant, but evolved over time depending on the conjuncture of technology, labour cost and social constructions of gender. The perceived ‘lower output’ of female workers was no doubt also due to women (particularly in household production) having to divide their time between textile production and domestic work to a greater degree than men. In 1925, Danish employers declared that it would be an injustice to give equal pay to women and men given that it was ‘a generally established fact’ that women were less physically productive than men.50 This was a view similarly held 48 Fowler, ‘Great Britain’; Cliver, ‘China’. 49 Fowler, ‘Great Britain’. 50 Dansk Tekstilfabrikantforening, Beretning om Foreningens Virksomhed (1925), pp. 29–30. 715

The Ashgate Companion to the History of Textile Workers by the vast majority of employers in textile history. However, as the same author also points out, keeping female wages low provided employers with a good price to productivity ratio. Certainly, the ‘lesser’ physical abilities of women did not inhibit the textile industry’s great use of female workers, and allowed companies to reduce their costs by utilizing gendered (female) cheaper labour. It should be noted, however, that while female workers could often be regarded as having the potential to undermine the position of their male counterparts, they could also themselves be vulnerable to such competition. Even cheaper or more easily exploited immigrant workers or ethnic minorities, for example, could replace or supplement the supposedly more tractable females who were already less expensive than men. Similarly, minors were a cheap and easily exploitable form of labour prevalent in nineteenth- and twentieth-century factories. In the nineteenth century, girls under the age of 14 were common in Japanese factories, while the Chinese industry employed girls as young as 8 in the silk industry and some Macedonian factories employed girls as young as 6 years old.51 Child labour often came from the poorest households and those without a home. In the nineteenth century, foundling children were employed in Brazilian factories, and orphaned children in government factories of the Ottoman Empire.52 Orphans were also utilized in the seventeenth-century Netherlands, though a gender distinction was made between orphan boys who undertook apprenticeships to train as weavers and orphan girls who remained (un-apprenticed) spinners.53 A common view held by textile employers across the globe was that of the ‘family wage’ or ‘adult male wage’, where the male household head was considered the key breadwinner and any income earned by wives and children was considered ‘supplementary’ to this. This gave employers the excuse to maintain male wages at a higher level than those of their female colleagues (and indeed children) on the shop floor, despite the disadvantages this brought male workers in respect of their labour being replaced by more cost-effective female labour. In Egypt, the idea of women’s earnings being supplementary to household income was considered a reason to pay female operatives in early twentieth-century factories below subsistence wages.54 Alongside gender, race and ethnicity was also used as a basis for determining wage levels in the industry. In late eighteenth-century Brazil, indigenous workers were paid in cloth rather than cash and plantation owners’ wives and daughters worked alongside predominantly female slave labour, while in nineteenth-century India spinning was often un(der)paid when performed by wives or by rural labour castes.55 In eighteenth- and nineteenth-century Mexico, slave and convict labour was utilized, while in the USA prior to the 1960s, black

51 Hunter and Macnaughtan, ‘Japan’; Cliver, ‘China’; Quataert, ‘Ottoman Empire’. 52 Delson, ‘Brazil’; Quataert, ‘Ottoman Empire’. 53 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 54 Beinin, ‘Egypt’. 55 Delson, ‘Brazil’; Roy, ‘India’. 716

Gender and the global textile industry men and women were denied production jobs or were given the dirtiest or most dangerous jobs on the shop floor.56 However, it was not all bad news. Even when female workers were paid less well than their male counterparts, textile work often gave them a greater economic strength and earning ability than they had previously held. A women weaving commercially in China in the mid to late eighteenth century could earn enough to support one to two adults and two children, while in the 1920s and 1930s rural women seeking factory work could earn enough to contribute at least half of the household income and in some cases even support the entire household and initiate their migration for industrial employment.57 Young female operatives in Japan could significantly add to the rural household income, and their ability to keep their earnings for themselves and save for their own future rose over time, particularly after the Second World War.58 Moreover, the undertaking of textile work generally gave women a higher wage than their counterparts in agriculture work. The overall demand for female labour in the textile industry therefore gave large numbers of women the opportunity to become wage-earners in their own right and in some cases to develop a greater sense of independence via economic means. It was better news still in the Lancashire cotton industry, where wages were based not on the ‘traditional’ adult male wage but on the collective earnings of family members as a whole. The author comments that this denoted female weaving operatives as the ‘best paid women workers in Britain’.59 Women in the Lancashire industry could also work after giving birth to children, which challenged the notion of a woman’s primary reproductive role.60 While maternity rights were granted in most countries during the twentieth century, textile work for a woman operative had previously usually ended upon her marriage and, if not then, upon childbirth (though in some cases a woman could continue to do piecework from the home). While some textile employers, like the large Japanese companies of the twentieth century, provided childcare facilities on factory grounds to enable mothers to return to work, the hiding of pregnancy from employers often led to terrible consequences for the unborn child as referenced in Cliver’s chapter on China.61 The transfer of textile production from the rural household to the factory over time also increased the ‘double burden’ for women in terms of their productive and reproductive work efforts, and could often lead to a complex ‘triple burden’ of domestic housework and child-rearing, agricultural labour and factory work. Gender also played a part in influencing the shop floor and factory environment in the textile industry. The heavy utilization of young and female workers allowed 56 Blewett, ‘USA’; Bortz, ‘Mexico’. 57 Cliver, ‘China’. 58 Hunter and Macnaughtan, ‘Japan’; Macnaughtan, Women, Work and the Japanese Economic Miracle, pp. 203–205. 59 Fowler, ‘Great Britain’. 60 Fowler, ‘Great Britain’. 61 Cliver, ‘China’. 717

The Ashgate Companion to the History of Textile Workers employers to employ patriarchal and paternal controls over employees. These included housing workers within factory compounds, regulations and restrictions with regard to work breaks and leaving company grounds, reward and punishments systems, and the use of male supervisors to oversee women and children. Factory shift hours were long, ranging from 10–17 hours in the nineteenth century, until legislation prescribed industrial hours and in many cases placed restrictions on the use of female labour (especially the ban on night shifts in Japan, Britain, Russia and India).62 Female workers were often subject to sexual harassment or abuse, either on the shop floor or on the way to and from work. Paternalistic personnel management could be seen as a form of ‘imprisonment’ in its worst light and a form of ‘benevolence’ in its best light. The ‘selling’ of young girls into the industry, exploitation by recruitment agents, the cramped and overcrowded living/working conditions that bred illness, and the six-day working week were not unusual conditions and reflected paternalism at its worst. Improvements over time to working conditions brought seemingly altruistic paternalistic efforts, as young workers from poor backgrounds were given housing, food, education, recreation and shopping facilities and maternity benefits. In the case of Spanish women weavers in the early twentieth century, this paternalism translated into imposed silences during the working shift in stark contrast to their male colleagues who were allowed to chat and sing.63 In the case of young female operatives in 1960s and 1970s Japan it translated into considerable welfare benefits including free or heavily subsidized housing, health benefits and opportunities for personal educational advancement.64 The existence of large numbers of women in modern textile mills aided the development of various female networks, which while they also helped employers in terms of their ability to introduce and train generations of female operatives, gave individuals a group identity and consciousness as working-class factory women. In some respects their growing strength on the shop floor and in the dormitories of industrialized mills no doubt meant, as referred to in the USA chapter, that employers were forced to ‘reinvent paternalism’ in order to stave off union strength and worker power.65 While it is certain that workplace conditions were miserable for many women (and children) in modern textile factories under industrialization, it is also true that their activism and consciousness as women reflected their dynamism as a workforce during the course of history. Young girls migrated long distances for textile employment, while older women combined paid production work with unpaid domestic work to support the prosperity of their families.

62 Hunter and Macnaughtan, ‘Japan’; Roy, ‘India’; Fowler, ‘Great Britain’; Pretty, Chapter 16, this volume, ‘Russia’. 63 Smith et al., ‘Spain’. 64 Hunter and Macnaughtan, ‘Japan’. 65 Blewett, ‘USA’. 718

Gender and the global textile industry

Gender, organization and activism The utilization of both male and female workers in substantial parts of the textile industry set up the potential for both conflict and cooperation within the workforce, and for shifting employer preferences in a context of changing labour costs and production priorities. In many locations early textile industrialization was characterized by conflict, with male workers, with or without collective representation, protesting against their replacement by cheaper female workers. Such protests are recurrent in the national stories of European countries such as Spain, in the Turkish and Egyptian dominions in the Middle East, in China, and in the New World, both Mexico and the United States. Responses to such resistance and protest were varied, but included strategies such as shifting women into less mechanized, less skilled or auxiliary tasks, as shown in the case of Egypt, Mexico and Spain, re-orienting women’s textile work into a life cycle phenomenon, thereby allocating the work lower status and pay (as found in China, Japan, Spain and the USA), or, most recently, finding an entirely new set up employing non-unionized, female workers, as in the case of the Mexican maquiladoras. Formal organization of textile workers came relatively early, and pre-industrial guilds were formed by textile workers in many countries, particularly in Europe. However, these were very much a ‘fraternity’ in nature. Guilds usually restricted their membership to male textile artisans, though sometimes widows were allowed entry if they continued their husband’s craft. However, guilds required the supportive labour of non-member women and children and actively solicited this labour through their male (household-head) members. As noted earlier, therefore, guild membership was a direct way of gendering the formal and commercialized sectors of the pre-industrial textile industry, and providing the male workers with more status within the labour force. This continued to be the view of early worker organizations in nineteenth-century textile mills, where female workers were referred to as ‘journeymen in skirts’ in Denmark66 and ‘daughters of freemen’ in the USA,67 in some respects relegating their worker status to an inferior or less independent level by virtue of their gender. There was much resistance, often from male workers, to the incorporation of women in textile labour organizations, which meant their membership took time and energy. There was a failed attempt to organize female weavers in Denmark in 1874, and their organization was not realized until a decade later when ‘male hand weavers realized they had to take colleagues in mechanized mills seriously’.68 Male textile workers would often initiate the formation of unions and female workers would follow their lead or be permitted access at a later date. There was of course a definite fear amongst male workers of being replaceable by female workers, particularly as lower-waged female labour was more cost-effective for textile 66 Christensen, ‘Denmark’, original version as presented to the conference, quoted from http://www.iisg.nl/research/textilenational.php (consulted 1 October 2009). 67 Blewett, ‘USA’. 68 Christensen, ‘Denmark’, original conference version. 719

The Ashgate Companion to the History of Textile Workers employers. Strikes in China in the 1920s and 1930s were strikes by male workers to try to prevent such replacements.69 Those women workers who raised their heads from their seemingly ‘docile’ status and participated in early strike activity were something to be feared, referred to as ‘babbling Amazons’ in the USA and ‘hot-blooded’ in Denmark.70 Although textile unions were led by and dominated by male workers, over time they realized the need to rally women workers in order for unions to gain overall momentum and power as organizations. There was sometimes an ulterior motive, however, as in the case of Spain where male workers hoped the unionization of females and their increased wages would lead to their replacement by more male workers.71 Sometimes, arguments for the rights of female workers could be more simply viewed in the context of their female role within the family and within society, such as the willingness of male workers in Spain to increase the pay of women because it would add to the overall family income, or the male workers in Mexico who were ‘defending traditional female virtue with traditional male honour’ against the abuse of female operatives by male supervisors.72 In the case of China, however, female workers took matters into their own hands and formed powerful ‘sisterhoods’ for mutual support and protection.73 In part, this was probably a response to a tradition in China of male workers joining large-scale sworn brotherhoods, but it also demonstrates the ability of women to organize in informal but powerful ways. The chapter on China notes that these sisterhoods and their YWCA links were later mobilized and used to great effect by the Chinese Communist Party after the Second World War.74 Informal textile organizations where female workers could gather together were also found in friendly societies and mutual associations in England and Japan.75 Over time textile organizations transformed themselves from male-only memberships to modern unions making demands for equal pay. Throughout the nineteenth and first half of the twentieth centuries, however, it would be fair to say that women textile workers often participated in strikes and other forms of organized protest in spite of lower rates of unionization compared to male workers and their often weak representation by union organizations. The participation of female workers in labour activism and strikes was witnessed early on in the industrialization process – as early as 1821 in the USA, in Denmark in 1886, in the Ottoman Empire in 1908, in Mexico in 1912 and in Russia in 1917.76 While strikes were often male-led forms of activism, women participated in large 69 Cliver, ‘China’. 70 Christensen, ‘Denmark’; Blewett, ‘USA’. 71 Smith et al., ‘Spain’. 72 Bortz, ‘Mexico’; Smith et al., ‘Spain’. 73 Cliver, ‘China’. 74 Cliver, ‘China’. 75 Fowler, ‘Great Britain’. 76 Quataert, ‘Ottoman Empire’; Christensen, ‘Denmark’; Bortz, ‘Mexico’; Blewett, ‘USA’; Pretty, ‘Russia’. 720

Gender and the global textile industry numbers and with ferocity, as demonstrated in the 1870 Fall River strike in the USA when female weavers ‘confronted the knobsticks with violence: yelling, jeering, kicking and pummelling strike-breakers with their fists and throwing stones and dirt clods’ alongside their male strikers.77 The overview of Spain by Smith et al., commenting on a 1913 strike, noted that ‘women proved every bit as militant as their male folk, contradicting stereotypes which portrayed them as meek and submissive’.78 Women also participated in less direct and more subtle forms of protest alongside their direct strike activity, which has challenged their historically ‘passive’ image. Such action was discernable in the 1912 ‘Bread and Roses’ strike in the USA when women supported and sustained strike activity from behind the scenes on ‘the tenement steps and [in the] kitchens’.79 Christensen’s Chapter on Denmark also suggests that less direct activism by women workers (for example lower union membership) does not indicate that women were more indifferent to their working roles than men, but rather that high expectation of women’s reproductive role in society meant they had a different experience as industrial workers than their male colleagues and thus had to develop different strategies.80 This often led to their acceptance of their ‘female subordination’ and their participation in activism which defended male-centred jobs and wages in the industry. It could also be argued that women were keen to protect their families and their role within the household. When Spanish textile women participated in a 1913 strike to reduce hours it could be suggested that they were protesting the long factory hours which added to their double burden of productive and reproductive work. While any successful worker activism throughout textile history can be attributed to the power of labour unions, it is also fair to say that the participation of female textile operatives in both direct and indirect forms of protest contributed to the strength and success of union action. It is clear from the Chinese case, for example, that sometimes female textile workers had a very unique and powerful role to play, as demonstrated by the case of the ‘Number Ones’. Women who had risen up through the textile ranks, the Number Ones operated as recruiting agents, shop floor bosses and leaders of activism and sisterhoods. Supported often by criminal gangs, they wielded a lot of power in the industry, to the extent they could also be accused of reinforcing a ‘patriarchal’ exploitation of young rural women, the so-called ‘plucking mulberry leaves’ trade in girls. They were a matriarchy in their own right and were often referred to as ‘godmothers’, standing out in a largely paternal industrial history. The Chinese textile sisterhoods were so strong that the Chinese Communist Party utilized their help in their 1949 revolution and liberation of cities. Powerful leadership was also displayed by women workers in 1875 Fall River activism in the USA, when female operatives shamed their male colleagues by 77 Blewett, ‘USA’. 78 Smith et al., ‘Spain’. 79 Blewett, ‘USA’. 80 Christensen, ‘Denmark’. 721

The Ashgate Companion to the History of Textile Workers successfully striking after the men had already reluctantly accepted reductions to their wages.81 These women also spoke on public platforms, were vocal at strike meetings and were active fund-raisers for the labour movement. The strident voices of these Chinese and American women, and the recorded participation of female textile workers in strikes and other protests in many of the countries in this study, demonstrate a degree of consciousness and identity as proletariat workers that was developed by women themselves. This increased consciousness, autonomy and economic strength (as individual wage-earners) often led women to settle in the urban areas rather than return to their rural homes, to resist or delay marriage, to fashion themselves as ‘modern’ women and to save for their own future.

Conclusions The changes that have taken place in the gendering of textile work over the last four centuries have in many respects reflected gendered aspects of the evolution of the manufacturing sector as a whole. One aspect of this is the importance of state policy. It is apparent that in a number of the countries considered here state policy had a significant impact on the development of the textile industry as a whole. In that context states also influenced the gendering of production. Consideration of these indirect effects of state policy lies beyond the scope of this chapter, but two direct ways in which state policy has had an impact on gender aspects of textile work can be briefly noted. The first is through the broad thrust of labour policy as a whole, which was invariably closely tied to the wider ideology of the regime in question, including its adherence or non-adherence to free market principles. In Communist China improved benefits for workers were imposed on textile producers as on others, and the glorification of industrial labour associated with the party ideology had a significant impact on the identity and image of female workers in the industry.82 Other Marxist-influenced countries, such as the Soviet Union, also promoted the image of female workers and sought to guarantee equal pay and status. The political control enjoyed by the ruling party in these cases went some way towards ensuring that the dictates of equality were met. Nationalist imperatives in wartime Japan also pushed towards greater gender equality in the workforce, including in textiles. By contrast, the populist Perón regime in Argentina also implemented policies designed to benefit workers, but found it was often resisted by employers.83 In general, it may be suggested that the gender division of labour, and, indeed, divisions along ethnic lines, for example, have tended towards greater equality where political regimes have shown a tendency to curb the excesses of free market operation.

81 Blewett, ‘USA’. 82 Cliver, ‘China’. 83 Lobato, ‘Argentina’. 722

Gender and the global textile industry The second direct way in which states have been able to influence the gendering of textile work is through the implementation of targeted, specific legislation, usually in the form of protective legislation. There are many potential areas for such legislation, ranging from minimum wage legislation in recent years (which may arguably undermine women’s comparative wage advantage compared with men), through to maternity provision, childcare leave and health and safety legislation. Most conspicuous in the case of the modern textile industry has been protective legislation aimed at curbing labour abuses or safeguarding the interests of particular segments of the workforce. The prohibiting of night work for women in the UK in the nineteenth century was a major factor limiting the 24-hour operation of textile mills, since employers were reluctant to employ an additional shift of male workers to fill the night time gap.84 In Japan, the introduction of a ban on night work for women was resisted by employers for years, but when eventually imposed in 1929 if anything enhanced labour productivity and stimulated further feminization of the workforce. The issue of night time working was also debated in India.85 The impact of such legislation has tended to vary across countries, but the extent to which protective legislation restricting the work conditions of one particular segment of the labour force, normally women, also acts to promote or undermine gender equality in that workforce lies at the heart of modern textile production. The process of the gendering of textile work in each country has also mirrored changing social perceptions of gender roles, and interacted with the process of technological change and capitalist development to produce new patterns of production and new understandings of what it means to be a male or female textile worker. Unlike some other industries, however, textiles has in all countries been more likely to employ both sexes within the same sector, often in similar jobs, laying the foundations for a sharper gender division of labour that reached its apogee in the most industrialized economies in the first half of the twentieth century. The details of that process have varied considerably across the globe, but the prominence of the textile industry in manufacturing production everywhere and its particular importance in the work of females, in both pre-industrial and modern times, means that textiles is a crucial area for understanding the significance of gender in the labour process. In many respects gender is the most profound, and the most important division in the composition of the textile workforce. In some countries the historical experience suggests that under certain circumstances gender differences can be subordinated to the interests of class. More often, though, gender differences and the unequal relations of power between men and women have shaped the production process, the worker experience and the overall process of capitalist development, and have in turn been shaped by them. In some respects it may be problematic to speak of common experiences that define the male or female textile worker, as we must be sensitive to enormous differentiation within the global industry. On the other hand, many of the issues on 84 Fowler, ‘Great Britain’. 85 Roy, ‘India’. 723

The Ashgate Companion to the History of Textile Workers gender discussed in this chapter seemed to emerge as commonalities and themes from many of the national overviews. The male–female differentiations and the voices of young women were not always specifically recorded in industrial history, but they can be uncovered through comparative studies such as these. And we may discover that women have gone from being ‘broken shoes’ (China) and ‘pitiful factory girls’ (Japan) to become the heroines of a new historical and industrial narrative.

724

29 Investigating identities within the global textile workforce Mary H. Blewett

Identity is both multi-faceted and embedded in history and culture. The interaction of personal subjectivity, of perceptions by others and of the lived experiences of skill, class, gender, race/caste, citizenship/legal standing/family status, religion, town/ region/nation and political ideology define the human identity of textile workers. All of these categories overlap and may be challenged, renewed, renegotiated and redefined or shattered during conflicts within the global textile industry. The complicated issues of identity require evidence on a level of experience often lacking to historians. Thus, many of the national overviews in this volume focus on economic analysis, but the explanatory power of social history, however devilish, is often in the details of the context. Likewise, any attempt to decide which elements of the overlapping categories of identity take precedence in a particular situation requires a detailed historic context and sensitive analysis. Evidence on self-imagery for textile workers is scarce and scattered. Oral histories, memoirs, autobiographies, novels and public statements usually during strikes in the press or in state hearings by textile operatives can reveal how they perceive themselves. Attitudes inferred from behaviour or situation cannot be proven but can be suggestive. Views of textile workers by the middle class, the state and employers contribute to identity perceptions, frequently through disparagement, racialization or political demonization. Although sexuality is a primary category of identity expression, sources on working-class experiences from their own perspective are few, especially for women. The concept of identity remains complex, subjective and changeable over time. I will analyse identity as separated into two overlapping spheres of class/occupation, gender, age and race and of self-images and perceptions by others, ending with regional, cultural and political experience.

 For a recent exploration of these perceptions of identity, see Lex Heerma van Voss and Marcel van der Linden (eds), Class and Other Identities: Gender, Religion and Ethnicity in the Writing of European Labour History (New York, 2002).

The Ashgate Companion to the History of Textile Workers

Identity as class/occupation, gender, age and race/caste hierarchies Alice Kessler-Harris has argued that historians should imagine class as both a process and an identity. Identity for textile workers involves primarily what work they do and where they do it, but the processes of production are altered by employers and shift in response to changes in the global market. Both objective and subjective power relations shape the class identity of textile workers. Furthermore various stages of pre-industrial and industrialized work often existed simultaneously as ‘enormous differentiation’ or in overlapping forms between 1650 and 2000. The ‘simultaneity of different types of labour in a given moment in time’ can encompass myriad forms of development and reorganizations in which identities are embedded. Analysing shifts in class identity requires an examination of differing modes of production.

Handicraft production in households Seventeenth and eighteenth-century textile production depended on rural households, such as the ‘clothiers’ and their families in Lancashire and Yorkshire, England. This domestic system appeared wherever people had access to fibres and ingenuity. Family-based production always meant the employment of children in simple tasks. It could evolve into ‘big clothiers’ employing wage labourers in household production as in Yorkshire, into small or large workshops as in Denmark or remain part of a more diverse system of production well into the twentieth century as in Japan, Turkey, China and India. Agricultural seasons, such as the Indian monsoon cycle and the ripening of wine grapes in Spain, interrupted hand-spinning with the lure of other income-earning opportunities in rural areas. Income from agriculture, subsistence or market-oriented, gave handicraft workers a little independence from economic forces, which when challenged by mechanization rapidly formed the grounds in early nineteenth-century Yorkshire for machine-breaking. In the American South, farm families claimed the rights of independent, landowning ‘yeomen’, asserting traditional cultural and political values to maintain their distance from the control of cotton mills. Seventeenth and eighteenth-century Chinese farm households, relatively secure on their land, produced cotton fabrics for household consumption, some 

Alice Kessler-Harris, ‘‘Two Labour Histories or One?’, in Heerma van Voss and Van der Linden (eds), Class and Other Identities, p. 135.  Roy, Chapter 10, this volume, ‘India’.  Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’.  Fowler, Chapter 9, this volume, ‘Great Britain’; Seidman, Chapter 19, this volume, ‘Turkey’; Christensen, Chapter 6, this volume, ‘Denmark’.  Fowler, ‘Great Britain’; Blewett, Chapter 21, this volume, ‘USA’. 726

Investigating identities within the global textile workforce purchasing the raw cotton imported from other regions with profits from growing sugar cane. The export by cloth merchants of these handcrafted cotton goods called ‘nankeens’ made by the ‘majority of farming families’ reached Western Europe and the Americas by the early eighteenth century. Chinese and especially Indian household production by rural men and women dominated world trade until the late eighteenth-century changes in European manufacture. The culture of rural family life shaped this Chinese household economy but included a long period of commercial success in regional and international trading. When faced with competition from Western manufacturing, rural households continued to produce their own cotton cloth well into the twentieth century. In India handicraft production was similarly long-lived, but hand-workers who owned no land tended to migrate into urban centres of production. Textiles were a central component of the material culture of indigenous people in Argentina and Mexico prior to colonization by Spain. Widely used for consumer goods, such as ponchos, and for religious ceremonies, these patterned fabrics woven by indigenous women in Argentina (and in Mexico) sported blue and red warps made from llama wool and other fibres. Before the Portuguese colonized Brazil, exquisitely fine details of cordage twist, as in elegant hammocks designed and made by indigenous craftspeople, delineated a social hierarchy of identity within the larger tribal group.10 European imperialism, the introduction of sheep and the system of encomienda (long-term servitude to the colonizers) nearly crushed this craft production, but it survived in Argentina to offset twentieth-century rural unemployment among people of mixed race and to delight ecotourists. Contemporary retail sales of craft production from Peru and Bolivia can link local economies with the global trading system, as in the arrangements of the late twentieth-century Kansas-based American firm the Peruvian Connection.11 Even in industrially developed nations, such as Britain and the United States, indigenous production of tartans and tweeds in the Scottish Highlands and of Navaho rugs and blankets in the American Southwest create commercially viable textiles. Indigenous textile production confirms regional and native identities, often reflecting resistance to the dominant contemporary culture, and is part of the analysis of global textile workers. In contrast with indigenous textiles, Portuguese settlers in Brazil carried on their traditional household production of cotton and woollen textiles for domestic consumption. African-born and Creole slaves on sugar plantations produced additional handicraft cloth for their own clothing as well as sacking for export materials.12

 Cliver, Chapter 5, this volume, ‘China’.  Roy, ‘India’.  Lobato, Chapter 2, this volume, ‘Argentina’. 10 Marx Delson, Chapter 4, this volume, ‘Brazil’. 11 See newsletter, 9/04 at www.theperuvianconnection.com 12 Marx Delson, ‘Brazil’. 727

The Ashgate Companion to the History of Textile Workers Artisan workshops The classic artisan system of master, journeymen and boy apprentices existed in most societies, often complementing household production. Interference with artisan privilege appears just as widespread. Eighteenth-century Castilian mercaderes (merchants) subordinated woollen guilds and monopolized key processes on which artisan production was dependent. Imported raw cotton from Spain’s distant American colonies led to the development of artisan weaving and cloth printing, dependent on cotton spun by women in their farming households. Spanish artisans who also owned vineyards and other valuable arable land wove scarves and ribbons on handlooms in preference to participating in developing factory work.13 Artisan workshops frequently organized into guilds, while some independent entrepreneurial drapers and middlemen in the key Dutch Republic cities of Leiden and Haarlem came to control dyeing, finishing and trading woollen fabrics. Coordination of textile production by merchants at lower labour costs in Tilburg stimulated an increase of 125 per cent in local hand woollen spinning and weaving between 1650 and 1810. Male artisans in guilds controlled woollen weaving in cities, while women without guild protection but with social rituals of their own – along with some men – dominated hand-spinning in their homes. Dutch men monopolized hand wool combing and shearing or cropping of woven cloth, as in Yorkshire. Women workers spun less valuable textiles, such as linen, associated with low rates of pay and poverty. Separated processes of production shaped by gender and skill formed the class identities of those who worked for wages.14 Prior to 1720, household production and artisan workshops interacted in Habsburg-Austria, with domestic producers using raw materials of local origin and supplying yarns for more highly finished goods of linen and wool made by town artisans organized in guilds. Masters of artisan households with journeymen and apprentices to feed depended on female members for daily sustenance and for ‘the administration and marketing of the trade’. ‘Noble manufacture’ or the establishment of production in centralized workshops relied on the power of eighteenth-century Bohemian feudal lords to coerce labour from their subjects.15 Conflictual class relations preceded the factory system. While the rural handicraft system dominated cotton textiles for centuries in China, silk production centred in artisan workshops, which could extract hard terms from insecure male weavers. Weavers of cotton cloth sometimes entered small workshops in China, but their wages were at the mercy of long apprenticeships and working hours. Intensive household production continued until nationalization of the cotton and silk industries in 1949.16 13 Smith et al., Chapter 17, this volume, ‘Spain’. 14 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’. 15 Komlosy, ‘Habsburg Monarchy’. 16 Cliver, ‘China’. 728

Investigating identities within the global textile workforce The small artisan workshop usually represented strict European traditions, including master, journeymen and young apprentices producing a valued product and associated with the local guild, although this system withered away in the vast wilderness of the British colonies in North America.17 Danish artisans participated in a special culture, involving rituals practised while travelling on the continent. In Copenhagen ‘factory-guilds’, protected by the absolute monarchy, redefined mutual goals of controlling competition and prices by intensifying production and exercising great power over wages and working conditions of early factory labourers.18 The artisan system for textiles when transplanted by Spain to the New World in eighteenth-century Mexico could emerge transformed by European racial categories intensified by imperial conquest. Racial and caste differentiations reshaped the artisan system brought to Mexico by employing both ‘forced and free labour’. Indeed the bottom-end obradores, or workshops, employed an array of forced labour: encomienda or indentured Indians, slaves, debt peons and convict labour, the majority male but a few females. Native Indian women who had dominated indigenous production did not find work in non-indigenous production. Less prison-like workshops employed free blacks and castas or castiza, representing people of mixed parentage, Indians, blacks and whites. The ‘society and culture [of colonialism] rather than technology determined male predominance’ in urban cotton textile production with Spanish-style guilds in Mexico City.19 European technology, imported wholesale to Mexico, including water turbines, English textile machinery, mechanics and skilled workers, transformed the class relations in workshop production into mechanized factories, not by long-term innovation but with pesos. Other Latin American nations, such as Uruguay, escaped this phase of imperial development and industrialized woollen production in the early twentieth century.20 The longevity, productivity and commercial reach of the eighteenth-century Egyptian artisan system, centred in Cairo, remains impressive. Trading cotton and linen cloth throughout the Ottoman Empire and with North Africa and Europe, especially France, this system, which used the family’s female and child members to feed yarn to artisans, journeymen and apprentices, lasted well if unevenly into the twentieth century. Less lucky artisans in the early nineteenth century were forcibly recruited by the state into large workshops equipped with European machines powered by oxen or donkeys. The gender division of labour within the artisan system permitted women to spin in the short-lived state workshops.21 In the face of a flood of cheap European textiles, the seemingly progressive stages of development reversed themselves as a growing putting-out system tapped rural Egyptian female labour for low wages to spin yarn in their homes for local weavers 17 Blewett, ‘USA’. 18 Christensen, ‘Denmark’. 19 Bortz, Chapter 13, this volume, ‘Mexico’. 20 Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’. 21 Benin, Chapter 7, this volume, ‘Egypt’. 729

The Ashgate Companion to the History of Textile Workers who sold locally and also farmed. When the factory system was established in the twentieth century, its locale was urban. Self-sufficiency in textile production in small Brazilian factories was the Portuguese imperial objective after a flood of cheap British textile imports threatened the national economic balance. The scope of these enterprises leapt over combined handicraft production and artisan workshops into centralized, wage-paying industrial organization using imported but hand-powered machinery. Centralized production for trade within Brazil and to the mother country employed resettled indigenous wage-earning workers, male and female, rather than the wives or slaves of settlers. Their employers regarded them (probably in both racial and class terms) as an exploitable proletariat and paid them only in cloth at rates carefully controlled by managers. Still the material cultural heritage of these indigenous workers resisted Europeanization by introducing geometric design patterns once used for hammocks and baskets. Raw materials to feed these small factories were to be cultivated by resettled indigenous people ‘wrested’ away from Jesuit missions.22 Meanwhile, household production continued in Brazil. Slavery in Brazil was not abolished until 1888 after the introduction of modernized textile machinery. Free and slave labour worked side by side in early nineteenth-century factories in Rio de Janeiro, but only free labourers kept their wages. English supervisors managed their work, while expert English mule spinners and weavers, relocated to Brazil as they did to the American Northeast in the second half of the nineteenth-century, presumably taught skills. Most Brazilian textile workers, illiterate, young and poor, experienced very long working days by electric light (14 to 17 hours) in the 1890s.23 The culture of slavery left its imprint on the nature of Brazilian textile manufacturing and the complicated identity of immigrant and native factory workers.

Occupation and organization Spinning cotton, wool and silk presents a dizzying variety of occupational arrangements in different nations under varying circumstances. Hand-spinning by females and children using spindle/distaffs or spinning wheels in a rural family household for minimal wages reflects isolation and marginality. Managers of child labour in water-powered spinning mills in New England and in Yorkshire treated children brutally and paid them poorly. Widespread use of cheap children’s labour until regulated by the state marked every period of textile industrialization on a global basis.24 In contrast, throstle or ring spinning by women in factories and jenny/mule spinning by men, powered by water, steam or electricity, represented paid work away from the home and some varying measure of independence. In Lancashire 22 Marx Delson, ‘Brazil’. 23 Ibid. 24 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 730

Investigating identities within the global textile workforce females did handicraft spinning but were denied access to powered mule spinning in factories. In New England, young rural women were recruited to water-powered throstle spinning but not mule spinning in integrated factories. In India, handspinning by unpaid women and children in the household system or by agricultural ‘labour castes’ gave way almost entirely to powered-spinning frames in centralized production in nineteenth-century urban centres, such as Bombay.25 For migrants, caste oppression was, however, weaker in Indian cities. The comparatively few women, who worked at the traditional female tasks in Indian cotton factories only where their husbands were employed, were organized into teams supervised by other females. Most Indian women continued to work in rural household settings.26 By the late twentieth century, however, spinning firms in India relocated to ‘semirural’ and impoverished areas to employ some women. Cheap machine-spun cotton yarns from Britain and India in the nineteenth century totally undercut commercial hand-spinning in China. In silk production, however, Chinese investors operated mechanized silk spinning mills or filatures for export in the late nineteenth century, employing rural Chinese women and replacing traditional hand-reeled silk. Local rural elites legitimized the shift in women’s traditional work by recruiting the help of lineage organizations along male bloodlines. Silk spinners centred in Shanghai also laboured in farming areas, where these women were the first to work on textiles outside the Chinese household. By the mid-twentieth century, the workforce in cotton and silk textiles, even in the male bastion of skilled silk weaving, was overwhelmingly feminized.27 As in Lancashire, when men and women weavers shared factory work, joint labour activity followed. In Japan, women hand-workers produced reels from silkworm cocoons. When silk spinning became mechanized in central factories, relatively skilled young women dominated spinning operations, turning reels of ‘raw silk’ into yarns that represented 64 per cent of world production at its peak in 1935.28 Specialist male cotton spinners in eighteenth-century Japanese urban centres who marketed their yarn through wholesalers coincided with widespread rural handicraft production for the domestic market. By the late nineteenth century, women ring spinners in modernized cotton mills supplanted both earlier forms, and as a result by 1935 Japan represented the second largest cotton yarn producer based on the labour of factory workers. Like China, the Ottoman Empire adapted to the rise of nineteenth-century mechanized textile production by abandoning most handicraft spinning operations to import English machine-spun cotton yarn. Although both urban and rural household production for domestic consumption in which women dominated both spinning and weaving continued, male textile workers concentrated on weaving and printing cotton goods for export. Child labour, either in families or orphaned 25 Roy, ‘India’. 26 Ibid. 27 Cliver, ‘China’. 28 Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’. 731

The Ashgate Companion to the History of Textile Workers or poor, contributed to the spinning of yarn whether in mechanized factories or in household production, a phenomenon that disappeared after the Second World War. Girls and teenaged females reeled silk. Steam-powered silk reeling factories for use on European-style power-looms replaced handicraft reeling, then faced widespread shortages from silkworm disease. Still the industry recovered for domestic consumption. Women spinners from a wide array of Middle Eastern religious and ethnic backgrounds provided mohair yarn for use by weavers’ guilds and spun other fibres both in their homes and in factories.29 Spinning wool by machine suitable for use in carpet knotting, however, was not technically possible until the turn of the twentieth century, when spinning mills in the carpet-making region of western Anatolia supplied these needs. In 1908, over a thousand female handicraft spinners ‘sacked’ three such mills, which destroyed their traditional livelihood.30 Adult male mule spinners in Lancashire, who supervised and trained assistants/ apprentices called piecers, enjoyed elite status as ‘barefoot aristocrats’. English immigrant mule spinners in American cotton mills tried to claim similar status by way of their skills, control of the trade and earning power. But shifting industrial circumstances undermined the elite status of migrant mule spinners. New England mill owners denied them the supervisory training of ‘big piecers’ and ‘little piecers’ by creating the unskilled job of ‘backboy’ as the spinner’s only helper. Likewise Dutch textile managers in Tilburg used mules to spin woollen yarn operated by adult males assisted by boys.31 Cutting piece-rates and speeding the machines exposed the mule spinner’s bare feet to splinters and broken bones, while the physical exhaustion of speeding undercut manly strength. Class exploitation broke the status of the mule spinners in North American mills. Still both in Lancashire and in New England (but not in Mexico or Russia), mule spinners became the organizers of exclusive craft unions. These mule-spinning unions never accepted the late nineteenth-century technology of ring spinning nor organized the ring spinners. By the twentieth century English and American cotton spinning mules and those used elsewhere had become technologically obsolete.32 In Yorkshire worsted mills, children and young, unmarried females ran throstle frames, while fine woollen production in the West Riding and in Europe required mule spinning. Community opinion condemned married women who worked in Yorkshire mills, while male woollen workers kept young females out of their unions, in contrast with Lancashire practice. In a concerted effort to remain independent of cheap English yarns, Austrian industrialists hired British engineers in the early nineteenth century to build cotton spinning frames, presumably with self-acting mules for male workers. The availability of these yarns stimulated household

29 Seidman, Chapter 19, this volume, ‘Turkey’. 30 Quataert, Chapter 18, this volume, ‘Ottoman Empire’. 31 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 32 Fowler, ‘Great Britain’; Blewett, ‘USA’. 732

Investigating identities within the global textile workforce weaving in rural areas, where prosperous cottage weavers enjoyed their status for a time.33 In Japan, China and Denmark, the industrial phase of mule frames run by skilled men was entirely skipped over; instead by the 1880s Western-style ring spinning by unskilled young women in cotton factories prevailed. Male workers in Japanese spinning mills did work requiring physical strength (opening cotton bales) and expertise (finishing such as dyeing).34 As in Japan, maleness for Chinese textile workers was associated with skilled work and higher wages usually based on the day rather than by the piece, but only until women moved into these traditional masculine jobs under the pressures of depression or war.35 Powerful early twentieth-century labour unions in Mexico’s leading industrial sector controlled hiring, working conditions, wages and management but, as in Lancashire mule spinning, opposed technological innovations, resulting in low relative productivity and declining wages after the Second World War.36

Weaving cloth Weaving fabrics from hand-spun or machine-spun yarns also represented a wide variety of productive organizations over time and in different cultures. Thus a worker’s identity as a weaver shifted and changed. In ancient China as in eighteenthcentury Japan, weaving defined ‘women’s social, economic and moral role’ but more practically fitted a female for marriage with skills essential to her future household. Commercialization of silk weaving operations in Chinese workshops with opportunities for wages, greater skills and trading outside the home discouraged homebound women and attracted men to weaving. This trend masculinized most silk weaving by the eighteenth century and led to the organization of strong allmale guilds associated with martial arts training and quick to protest grievances. Skilled male weavers claimed the wages and privileges of ‘labour aristocrats’, not unlike mule spinners in Lancashire. They stood at the head of a labour hierarchy descending to the young females who spun in filatures and cotton mill hands, whose lives and wages ranged from ‘tolerable to miserable’.37 Male weavers in Mexico spearheaded ‘societies’ or unionization in the mid-nineteenth century for wage increases and fewer hours. Their sense of traditional manhood and honour led them to protest the abuse of the minority of women textile workers by supervisors. Manhood meant working outside the home for wages, and men dominated textile production in Mexico. When organized cotton textile workers won a general strike in 1911, this historic victory for both men and women workers claimed the national revolution as their own, a moment 33 Komlosy, ‘Habsburg Monarchy’. 34 Hunter and Macnaughtan, ‘Japan’; Cliver, ‘China’. 35 Cliver, ‘China’. 36 Bortz, ‘Mexico’. 37 Cliver, ‘China’. 733

The Ashgate Companion to the History of Textile Workers in worker culture not unlike the role of Chinese silk weavers in Shanghai in 1949 to be remembered and celebrated.38 In both societies, these events represented the peak of progressive and inclusive labour reform in Mexico and in China. The recruitment of rural peasants in China, Mexico, Russian, the United States and elsewhere for textile factories did not necessarily mean a delay in the emergence of class conflict and militant organization, although experienced, urbanized migrants from industrialized societies such as England sometimes led the way.39 Indeed in the late nineteenth-century Central Industrial Region of Russia, weavers, both men and women from rural areas, formed a ‘critical mass’ led by male militants necessary to organize work stoppages and initiate organization. As the workforce in weaving became increasingly feminized, women led turbulent strike activities in wartime Petrograd (St Petersburg) in 1917.40 Some small workshops in China used sophisticated looms to produce complex patterns in silk fabric, but in the 1920s silk weaving mills in Shanghai using powerlooms established the city’s reputation as the major centre for silk weaving in the twentieth century. Weavers were hired as individuals, not as contract labour. Generation and education divided silk weavers in the Shanghai mills into two distinct groups: illiterate former artisans with declining opportunities and young literate men and women with company-sponsored apprenticeships eager to assume the lives and status of urban factory workers. These divisions also shaped how they organized interest groups. Artisans joined traditional brotherhoods, secret societies or gangs, while younger workers emulated the leftist students with whom they associated in cities and formed labour unions. Young women silk weavers participated in class conscious strikes in Shanghai in the 1930s, encouraged by radical young male weavers and the Chinese Communist Party.41 Sharing work and equal piece-rates among weavers as in Lancashire enrolled female militants in organizing worker resistance. Dutch male cotton weavers preferred the help of their eligible children.42 Chinese cotton spinning was quickly mechanized in the 1920s and 1930s by ring frames in centralized production, but handicraft weavers, working in households or small workshops, continued to dominate the export of cotton cloth through merchant/traders until the 1970s. Japanese occupation from 1937 to 1945 resulted in post-war shortages of raw cotton and unstable supplies of electricity. Still power-loom weaving in Shanghai cotton mills began to recover, while most cotton looms even in the city were still powered by foot or hand. The great expansion in power-loom weaving in cotton mills, representing the final end of centuries of ancient handicraft production, would be carried out in the 1950s through nationalizing the entire Chinese textile industry. 38 Bortz, ‘Mexico’. 39 See the discussion of peasants as proletarians, in Pretty, Chapter 16, this volume, ‘Russia’. 40 Pretty, ‘Russia’. 41 Cliver, ‘China’. 42 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 734

Investigating identities within the global textile workforce Handloom weavers and cloth printers in India formed guilds, participated in an artisan system of masters and journeymen, and frequently worked together in small urban workshops, forming a group identity. Some guild members in Indian villages became ‘weaver-traders’ or ‘capitalist weavers’, a pathway to factory ownership. Indian girls, unlike Chinese and Japanese girls, seemed uninterested in casually acquiring weaving skills in contrast with their carefully trained brothers.43 Guilds in Istanbul and other Ottoman cities presented a wide spectrum of organization and policy, yet all tried to control access to the artisan trade. Merchant-traders supplanted these guilds in the mid-nineteenth century by organizing large weaving workshops run by masters and journeymen to whom they paid piecework wages.44 Yet powerful guilds of Spanish artisans stood in the way of the factory system, forming Spain’s first organized challenges to industrial capitalism. In the nineteenthcentury labour movement of Barcelona, both male mule spinners, apparently using British models of union strategies, and male weavers claimed artisan status and privileges on the factory floor.45 Hand-weaving in the underdeveloped nineteenthcentury Russian cotton industry became concentrated in buildings called svetelki with multiple looms run by groups of peasants sometimes on a communal basis but more often financed by well-off peasants who ran them as investments.46 The quality of textiles produced also contributed to an elite identity for some textile workers. Skilled and privileged artisans created the coveted silk brocades, rich combinations of gold, silver and silk fabrics, mohair woollens and ‘Oriental carpets’ of the Ottoman Empire, which dominated European definitions of ultra-luxury for a time.47 Imperial silk weavers for the Yuan and Ming dynasties prior to 1645 created a large hereditary caste of bonded artisans to produce luxury goods for consumption and for commercial uses, but they were also permitted to weave silk for the domestic market.48 By the twentieth century, Indian handloom weavers became differentiated by products produced, such as silk saris and goldthread cloth with complicated designs for the elite. This work conferred status, skill and ‘honour’ on one sector of weavers, while rural artisans who wove coarse cotton cloths for village neighbours occupied the lowest status. This hierarchy led to growing inequality and sustained wage differentials.49 Artisan silk weavers in Japan produced cloth for ruling class warriors or damasks and brocades for the politically dominant elite.50

43 Roy, ‘India’. 44 Quataert, ‘Ottoman Empire’. 45 Smith et al., ‘Spain’. 46 Pretty, ‘Russia’. 47 Quataert, ‘Ottoman Empire’. 48 Cliver, ‘China’. 49 Roy, ‘India’. 50 Hunter and Macnaughtan, ‘Japan’. 735

The Ashgate Companion to the History of Textile Workers Race Bruce Nelson argued that ‘class and racial identities are intertwined’, making ‘a single identity with multiple dimensions’ while creating ‘a process of becoming that both crystallises at particular historical moments but also continues to change over time’.51 Context remains crucial. Racial hierarchies became endemic to Asian textile capitalism. Owners of Japanese textile factories, facing labour shortages, employed racialized Korean workers during the years between the two world wars. When the Japanese came to dominate mechanized cotton spinning mills in Chinese treaty ports of the 1930s, especially in Shanghai, Chinese textile workers became part of a terrorized people.52 The Japanese invasion in 1937 undoubtedly intensified racial antagonism with the investors/invaders seen as ‘foreign devils’, while the Chinese became a racially contemptible underclass. Perceptions of workers as utterly expendable indigenous races turned the Spanish encomienda system in Latin America into both class and racial exploitation of subject peoples. The Spanish colonialists had intense racial consciousness, which shaped a hierarchy of categories for degrees of mixed parentage and place of birth among the indigenous, blacks and whites, but the possession of wealth often determined the degree of ‘whiteness’ in Mexico.53 The takeover of the Sudetenland in Czechoslovakia by the Nazi government in 1938 allowed German authorities to replace despised Jewish textile mill owners and presumably Jewish workers with Aryan Germans.54 Textile employers in the Piedmont area of the American South used the threat of introducing African American workers into jobs commonly done by those who prized their privileged status of whiteness, both male and female. When white workers sought better opportunities in a booming post-war economy, blacks, once confined to the worst work by employers suspicious of their abilities, took over the labour force in Southern textiles, just as it faced precipitous decline.55 Nineteenth-century immigrant textile workers also became racialized or imputed with inherent, naturally ingrained characteristics and values. The result was a racialized hierarchy within the textile workforce. Immigrant textile workers from Lancashire and Yorkshire faced the scorn of native-born Yankees as vulgar, illiterate, dialect speaking ‘jickeys’, a peculiarly New England ethnic slur.56 The more experienced textile workers of both Yankee and English backgrounds in New England racialized the assumed docility of French-Canadian, Catholic Quebecois. Ironically by the early twentieth century, French-Canadian men had formed strong 51 Bruce Nelson, ‘Ethnicity, Race and the Logic of Solidarity: Dock Workers in International Perspective’, in Sam Davies et al., Dock Workers: International Explorations in Comparative Labour History, 1790–1970 (Aldershot, 2000), vol. 2, pp. 657–680, at p. 659. 52 Cliver, ‘China’. 53 E-mail, J. Bortz to M. Blewett, 21 March 2004. 54 Komlosy, ‘Habsburg Monarchy’. 55 Blewett, ‘USA’. 56 Ibid. 736

Investigating identities within the global textile workforce loom fixers’ unions as the mule spinners’ unions declined. Immigrant textile workers from Italy also became racialized in the American Northeast as violent, radical and emotional. In general employers regarded laziness or stupidity as characteristic of easily coerced migrant peasants (even those with long traditions of handicraft skills) recruited to factory work at low wages.57 The introduction of second and third shifts into English textile mills in the 1960s, in spite of bonuses, did not attract men while protective laws prohibited women from working nights. Immigrant labour from Pakistan, India and South East Asia dominated the third shifts.58 Racial friction flared in the mill centres of Yorkshire and Lancashire, contributing to the atmosphere of twentieth-century industrial decline.

Self-image of textile workers and perceptions by others, religion, region/nation, family status/legal standing/ citizenship and political ideology Culture and class are the primary building stones of textile worker identity. Cultural variations are reflected in religious life, region/town loyalty, position in family hierarchies of age and gender, and cultural and political ideologies. Stefan Berger and Angel Smith argued that categories of past (and present) concepts of class, national and ethnic identity must allow for ‘heterogeneity, historical contingency, plurality and ambivalence’.59 This is especially relevant to self-perception and perception by others. Conscious as well as circumstantial perceptions of textile workers reflected their changing experiences of culture and heritage as well as their skills and job. Danish weavers were separated in the late nineteenth century into artisan handweavers and female power-loom operators in factories. Male handicraft weavers, organized into the Weavers’ Union of Copenhagen in 1873, regarded such women as inferior, unskilled ‘journeymen in skirts’. But the future lay in mechanization and female organization while craft exclusion represented the past. Late twentiethcentury Danish and Dutch women weavers regardless of their union protections saw themselves as ‘stressed-out’ with nerves strained and unable to sleep from tending too many looms and working double shifts. By the 1960s these jobs came 57 Ibid. 58 Fowler, ‘Great Britain’. Similarly, migrants from Algeria filled third shift jobs in textile mills in northern France, Leslie Page Moch, ‘Dividing Time: An Analytic Framework for Migration History Periodization’, in Jan Lucassen and Leo Lucassen (eds), Migration, Migration History, History: Old Paradigms and New Perspectives (Bern, 1999), pp. 41–56, at p. 54. 59 Stefan Berger and Angel Smith, ‘Between Scylla and Charybdis: Nationalism, Labour and Ethnicity across Five Continents, 1870–1939’, in Stephan Berger and Angel Smith (eds), Nationalism, Labour and Ethnicity, 1870–1939 (Manchester, 1999), pp. 1–30, at p. 3. 737

The Ashgate Companion to the History of Textile Workers to be regarded by women workers as ‘low status’ and undesirable.60 Similarly Russian women weavers faced a low-paid and high stress occupation and sent their daughters into service work.61 In the 1990s former textile workers in the Czech Republic faced massive rural unemployment, resorting to ancestral patterns of survival including growing their own food while taking in tele-homework or participating in periods of migration.62 Capitalist employers (save perhaps the few genuine paternalists) as well as managers of state-owned enterprises and the state regarded textile workers as commodities as a result of ‘marketization’.63 Virtually any slowdown, resistance, complaint, protest, migration, strike, proposed labour law or trade union contradicted desirable progress or legitimate profits. Supervisors in Mexican textile mills singled out the minority of women workers for verbal sexual abuse in a society where females working outside the home were not entirely respectable.64 When in the 1920s male workers hit Shanghai factories with strikes, employers shifted to more docile, easily controlled and lower waged females. They did not, however, wish to employ any regarded as liabilities: ‘old women, women with bound feet or pregnant women’.65 The importance of controlling hiring, with opportunities for kickbacks and bribes or ‘gifts’, took a unique form in China. The position of ‘Number One’ gave older women an unusual chance in Chinese society to develop supervisory ties with non-related female textile workers based on gratitude, respect and protection against predatory gangs interested in selling young women into various forms of enslaved work. Still these supervisors were often involved in criminal gangs and their own interests determined their support of worker strikes. Employers viewed their ‘contract workers’ as ‘weak, tired and apathetic’, but hiring was controlled by local ‘Number Ones’ and criminal gangs, which took over Shanghai’s labour market after murderous government repression of local Communists in 1927. The Guomindang government remained suspicious of all labour unions, even those controlled by its own agents. Nationalization of all textile production by the Chinese Communist Party changed the perception of poor, exploited working women as ‘broken shoes’ into heroines of the Revolution with rights and dignity.66 Owners and bosses, mostly white foreigners, looked down on Mexican textile workers, skilled or unskilled, because they were dark skinned locals of mixed parentage. The revolution of 1910–1927 turned the tables on these employers who, bowing to the political and economic power wielded by the violence of the textile unions backed by armed gangsters, abandoned their prerogative to discipline 60 Christensen, ‘Denmark’; Van Nederveen Meerkerk, Heerma van Voss and HiemstraKuperus, ‘Netherlands’. 61 Pretty, ‘Russia’. 62 Komlosy, ‘Habsburg Monarchy’. 63 Roy, ‘India’. 64 Bortz, ‘Mexico’. 65 Cliver, ‘China’. 66 Ibid. 738

Investigating identities within the global textile workforce workers and returned deference and respect to proud male workers and locally powerful union bosses.67 Twentieth-century trade union leaders in Denmark, protecting their hard-won status, concessions and agreements, particularly detested the wildcat strikers of Silkeborg in 1934 who were dismissed as ‘communists’ and ‘insane’ children, regardless of widespread local support for their demands for respect and dignity in the workplace. The ideal Danish industrial worker, according to the trade unions, did his duty steadily at work and demanded his rights, thus earning the status of respectable citizen. Women workers were not expected to conform to this ideal. And those who rebelled against trade union authority in wildcat strikes and formed separate female unions in 1886 and 1906, some of whom were Polish, were denounced as erotically undisciplined and sexually dangerous: ‘hot-blooded young women who can be brought swiftly into ecstasy’.68 The association of female militancy with uncontrolled sexuality also reflected male attitudes in the Yorkshire labour movement, while in the Piedmont region of the American South young ‘disorderly’ women teased and challenged owners, police and state militia.69 The miserable poverty of homework weavers in Austria and Bohemia trying to compete with power-looms inspired journalistic sympathy in the Social Democratic press for their notorious exploitation. Many of these weavers, however, preferred to eke out a traditional living at various kinds of work as servants or migrants to avoid factory life.70 Dutch textile workers refused to leave their plots of land to maintain their ability through the status of landowner to provide home-grown food or simply to breathe ‘the free air’.71 In the late nineteenth-century Russian textile factories, many male workers kept allotments of sown land in their home villages, maintaining ties to an agricultural ‘social welfare safety net’ during industrial depressions and strikes. These ‘peasant workers’ straddled industrial and agrarian cultures.72 Managers/owners in declining regions refused to invest in deteriorating textile mills and offered no future for workers, undercutting pride in skills and experience. In mid-twentieth-century Mexico, textile work came to be viewed as ‘shameful’, while in New England mill towns, workers tried to escape the label of ‘mill rats’, insisting on better jobs for their children.73 Southern mill workers, commonly regarded as ‘lint heads’, just as eagerly sought other work. Similarly, de-skilled

67 Bortz, ‘Mexico’. 68 Christensen, ‘Denmark’. 69 M. Blewett, ‘Yorkshire Lasses and their Lads: Sexuality, Sexual Customs, and Gender Antagonisms in Anglo-American Working-Class Culture’, Journal of Social History 40 (2) (2006): 317–336 and Jacquelyn Dowd Hall, ‘Disorderly Women: Gender and Labor Militancy in the Appalachian South, Journal of American History 73 (1986): 354–382. 70 Komlosy, ‘Habsburg Monarchy’. 71 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 72 Pretty, ‘Russia’. 73 Bortz, ‘Mexico’; M. Blewett, The Last Generation: Work and Life Among the Textile Workers of Lowell, Massachusetts, 1910–1960 (Amherst, 1990). 739

The Ashgate Companion to the History of Textile Workers textile workers in the Netherlands, caught in a rapidly declining industry after 1950, also directed their children towards different jobs.74

Religion Religious values could inspire moral critiques of textile capitalism or bind together co-religionists in migration or in organizations. Protestants fleeing from the southern to northern Netherlands to escape the ongoing wars against Spain brought capital and new skills as textile merchants and artisans that impelled development in the woollen trade. The Dutch Republic welcomed these refugees to Leiden, a centre of woollen production, which became the centre of the world trade until 1664. Amsterdam attracted persecuted religious groups, including Catholics and East European Jews. Trade unions in the early twentieth-century Netherlands organized around religious sects: Catholic, Protestant and secularists.75 Widespread religious non-conformity to the established Anglican Church in England supported dissent of all kinds against textile capitalists in Lancashire and Yorkshire. Catholicism became central to worker culture in some paternalist mill villages in Mexico. The religious communities of Old Believers in Russia, widely dispersed but holding wealth in common, moved to Moscow during a seventeenth-century period of tolerance and invested in the development of the cotton industry, the first important Russian industry. Before emancipation in 1861, Old Believer entrepreneurs attracted serfs to factories by paying their peasant communes to free them or exempt them from military service.76 Religious toleration also enabled textile production in the Ottoman Empire. In Cairo, Muslims, Christian Copts and Jews all engaged in silk weaving, presumably as members of the same guild.77 Guilds in Istanbul likewise included Muslims, Christians and Jews, although co-religionists sometimes acted together to their mutual advantage.78 Religious dignitaries – imams, priests or rabbis – in the Ottoman Empire approved of and helped in the recruitment of young girls for work in nineteenth-century silk reeling factories.79 However, in 1947 the Turkish government’s punitive tax on non-Muslim citizens sent a message to Jewish owners of textile factories; Muslims would start up any new firms.80 An institution important to Chinese female textile workers was the American Young Women’s Christian Association (YWCA), whose activities helped young women become independent of their rural families and live together in hostels. YWCA-educated female radicals (somehow trading Christianity for dialectical 74 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 75 Ibid. 76 Pretty, ‘Russia’. 77 Benin, ‘Egypt’; Seidman, ‘Turkey’. 78 Seidman, ‘Turkey’. 79 Quataert, ‘Ottoman Empire’. 80 Seidman, ‘Turkey’. 740

Investigating identities within the global textile workforce materialism) formed a strong contingent supporting the Chinese Communist Revolution.81 On the whole, religious faith and values are underestimated in studies of textile worker identity.

Town/region/nation Spatial loyalties to centres of production, regional distinctiveness or a revered sense of the nation can shape human identity. Towns associated with progressive political movements or with experience and skill, such as Barcelona, Shanghai, Philadelphia and Bombay, or towns in Lancashire boasting high lists of wages, can confer pride and status. Middle-class observers in Spain equated violent anarchists with peasant migrants from the south, not the respected workers of Catalan.82 Household producers in Austria and Bohemia preferred their familiar villages to relocation to an unknown urban area. Nineteenth-century St Petersburg specialized in huge spinning operations, while Ivanovo became the textile-printing centre. The putting-out system in the surrounding regions took the yarn and fed cloth to the printers. By the twentieth century, the Central Industrial Region in the provinces of Moscow and Vladimir employed 80 per cent of all cotton textile workers.83 Perhaps the most pronounced differentiation of regions occurred in the cotton mills of Lancashire and the woollen industry of Yorkshire, each with contrasting cultures of labour protest.84 Both regional cultures sent working-class migrants in the nineteenth century to the American textile industries. Transnational ties based on regional loyalties to culture and labour activity connected the English labour movement with protest activities in the American Northeast, shaping the American Federation of Labor’s exclusive trade unionism.85 The imposing regional model of English textile industrialization spread to Western Europe. In Denmark the goal of negotiated wage schedules – including the basing of piece calculations on Lancashire models using the English inch – and strong male-led, trade union leadership for textile workers among Danish and Spanish labour activists suggests some transnational borrowing of ideas from the successful Lancashire experience. But the organization of jobs by gender in Danish cloth (woollen) mills did not copy Yorkshire practices nor were self-acting mules introduced into cotton spinning.86 Spanish textile mill owners also tried to copy English practices called treball a l’anglesa to cut costs.87 Traditional concepts of textile work done by men and women faced uprooting and rearrangement as 81 Cliver, ‘China’. 82 Smith et al., ‘Spain’. 83 Pretty, ‘Russia’. 84 Fowler, ‘Great Britain’. 85 Blewett, ‘USA’. For regional development in Germany, see Ebeling et al., Chapter 8, this volume, ‘Germany’. 86 Christensen, ‘Denmark’. 87 Smith et al., ‘Spain’. 741

The Ashgate Companion to the History of Textile Workers industrialization proceeded. But the imperatives of industrial capitalism and trade unionist backing for the West European male breadwinner norm – that included unmarried men but not single women or widows – were the major forces in the ‘social reconstruction’ of gendered work.88 Regional identity could divide rather than unite textile workers. The American textile industry guided by the trade association the Arkwright Club of Boston operated textile factories in the twentieth century both in the Northeast and in the South, pitting regional cost advantages against each other to undercut unions and lower wages.89 Merchant traders in the Ottoman Empire in the early nineteenth century organized the most widespread regional implementation of the putting-out system. Purchasing raw cotton in one region (Adana), they hired handicraft spinners in another cluster of towns to produce yarn destined for weavers either in the same locality or in urban centres.90 Other regions throve on traditional skills, such as the carpet-making region of western Anatolia. Privately capitalized factories in Spanish Catalonia centralized the work of weaving, printing and dyeing, while spinning, by peasants, continued in rural areas. Peasants in the region of Muntaya with access to scarce water power could be recruited to factory work or remain somewhere inbetween as half-time textile workers and half-time farmers/vintners. Those linked to the land had less acquaintance with factory skills and the city life of Barcelona and presumably different self-perceptions. Certainly, their employers paid them less and worked them harder. Peasants with different ethnic backgrounds became migrants to mid-twentieth-century Catalonia.91 Migration from regions with abundant labour or special skills helped shaped textile industrialization. The regions of the southern Netherlands and LombardyVenetia located in ‘the transitory margins of the Monarchy’ had played a crucial role in the sixteenth century in introducing advanced textile skills, technology and capital, especially in Viennese silk production, to the adjacent peoples of HabsburgAustria.92 Working-class immigrants from Italy and Spain contributed to the initial phases of textile industrialization in Argentina but were replaced by young women (probably denied access to skilled weaving) and children of native birth.93 The same sources of experienced immigrants to Uruguay in the early twentieth century led to their domination of the more skilled jobs, but unlike English immigrants to the USA, they seemed not to have influenced labour politics.94 Post-Second World War Istanbul became flooded by ‘undocumented workers’ and migrants from the countryside, some finding employment in textile production and thereby undercutting wage standards and contributing to weak unionization, 88 Ibid.; Christensen, ‘Denmark’. 89 Blewett, ‘USA’. 90 Quataert, ‘Ottoman Empire’. 91 Smith et al., ‘Spain’. 92 Komlosy, ‘Habsburg Monarchy’. 93 Lobato, ‘Argentina’. 94 Camou and Maubrigades, ‘Uruguay’. 742

Investigating identities within the global textile workforce made illegal by the state in 1980. Turkish ‘guest workers’ represent the more unusual migrants, especially those who remitted crucial foreign exchange to the homeland and those who set up with their savings clothing and textile firms in Western Europe.95 Dutch textile owners invited similar guest workers from Italy, Spain and Morocco to ease post-Second World War labour shortages.96 Integrated textile factories in twentieth-century China recruited women from rural regions more and more distant from established centres of production. Subei women from farm families with a strange dialect and less experience in handicraft production seemed ‘rough, coarse and dirty’ to skilled factory workers from the Jiangnan region. They accepted the least desirable jobs, usually assigned to unskilled men or even worked for Japanese-owned mills, which intensified the contempt in which they were held. Despised outsiders – whether rural migrants or of different regions, religions, languages and cultures – frequently divided textile workers and prevented the emergence of class solidarity.97 Still, dramatic incidences of multicultural labour protest, such as in the Lawrence, Massachusetts Bread and Roses strike in 1912, suggested that class cohesion could supersede cultural divisions among immigrants.98

Position in family/legal standing/citizenship Wherever patriarchal family relations helped to order textile production, either in pre-factory or factory settings, identity became associated with position and age in the family system, which often meant limited legal identity and citizenship. Family heads and skilled young men might talk and sing during factory hours in Barcelona, while the young and female remained silent. Family remained primary for Mexican and Spanish textile workers who both favoured and protected their relatives in the mills. However, for elderly, experienced family members with stiffening joints and failing energy, textile work became difficult if not impossible. Adaptation to speeded machinery and additional tasks favoured the young. As the Mexican textile industry declined during the 1980s global competition, family traditions and the culture of male worker pride among mill hands ended. De-skilled jobs shifted into the hands of unskilled women, earning low wages.99 Only full citizenship and demonstrated artisan skills in the Dutch Republic in the seventeenth and eighteenth centuries won a man his coveted place in the woollen weavers’ guilds. The exclusion of women weavers, except for a few widows, reflected their lack of full citizenship.100 When serfdom in the Bohemian Lands was abolished in 1781, rural labourers became freer to marry, travel and choose their own 95 Seidman, ‘Turkey’. 96 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 97 Cliver, ‘China’. 98 Blewett, ‘USA’. 99 Bortz, ‘Mexico’; Smith et. al., ‘Spain’. 100 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 743

The Ashgate Companion to the History of Textile Workers work as citizens.101 As the textile industry declined in the post-1950 Netherlands, jobs reserved for members of mill owners’ families were protected, while rank and file textile workers faced unemployment.102 Citizenship remained strictly limited for women and children. Limited citizenship shaped the legal status of Japanese (and many other) female textile workers. Legal identity and citizenship in the Japanese Civil Code was based on family status: young women, ‘older female workers’ and undoubtedly wives were unable to sign contracts or collect wages.103 Married female mill workers in nineteenth-century New England faced similar disabilities under the common law concept of ‘femme couverte’: the civil death of women in marriage. In the Netherlands, women who married or became pregnant were fired.104 One particular Chinese institution that helped women textile workers avoid the legal, economic and cultural disabilities of marriage – the worker sisterhoods – developed out of recruitment practices involving the female supervisor, ‘Number One’. These organizations represented marriage avoidance societies with retirement and burial benefits. Labour solidarity was particularly intense among the worker sisterhoods. Or some experienced and well-paid silk spinners bought a concubine or second wife for their husbands.105 Male operatives with full citizenship that sometimes included suffrage often used their status as ‘free-born’ men to argue for additional political rights as workers in nineteenth-century England and the United States. Even full citizenship for textile workers had its costs. Wars and nationalist revolutions could shatter even the most established textile industries, as in Japan and the Netherlands in the Second World War, and during the early nineteenth-century and early twentiethcentury revolutions in Mexico. The ‘chaos’ during the first Soviet Five Year Plan, which favoured heavy industry, hit textile workers hard with soaring prices and shortages of food and raw materials.106 The loss of former colonies after the Second World War, such as Indonesia for the Dutch, meant the collapse of a previously dominated textile market.107 Civil wars in Uruguayan rural areas disrupted intensive sheep breeding ventures.108 The requirements of twentieth-century warfare shifted English textile pre-eminence from Lancashire cottons to Yorkshire khaki, while Egyptian and Brazilian textile production, spurred by demands from the allied war effort, mechanized and expanded.109 War in the 1940s disrupted international trade creating demands for increased domestic production even in the absence of modern machinery, as in Argentina, while stimulating trade unionism and labour 101 Komlosy, ‘Habsburg Monarchy’. 102 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 103 Hunter and Macnaughtan, ‘Japan’. 104 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 105 Cliver, ‘China’. 106 Pretty, ‘Russia’. 107 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 108 Camou and Maubrigades, ‘Uruguay’. 109 Beinin, ‘Egypt’. 744

Investigating identities within the global textile workforce laws in Uruguay in the 1940s.110 For nearly a century (1760–1840), the exactions by the taxmen serving the military needs of the warring Ottoman Empire discouraged small producers.111 The continental trade blockades during the Napoleonic Wars at first stimulated then undercut Danish textile manufacturing, which came to rely on cheap, imported English cotton yarns for weaving until the early twentieth century, as did the Chinese.112 The Taiping Rebellion in mid-nineteenth-century China killed millions of household producers, wrecked centres of silk production and destroyed acres of mulberry trees. Treaty concessions that hurt Chinese customers followed, allowing the massive importation of Western machine-made cloth and yarn of lower quality and higher price in the 1870s from Britain, the United States, Japan and India.113 Japanese invasion and occupation in 1937 nearly extinguished Chinese silk production and severely disrupted cotton cultivation.114 So promising was the post-war expansion of Shanghai’s textile position under the Chinese Communist government that Nationalist Chinese aircraft from Formosa bombed the main electric power plant in 1950 causing widespread shutdowns and worker dislocations.115

Political ideologies As nineteenth-century textile industrialization and global competition undermined crafts and working conditions, some workers became radicalized and joined nationalist groups. In the German woollen industry, Social Democratic voters elected parliamentary representatives.116 Failures in nineteenth-century strikes and collective bargaining resulted in labour strife and the growth of anarcho-syndicalism in twentieth-century Catalonia among Spanish textile workers. The worse the unrest, the more military the response by the state controlled by industrial elites and backed by conservative Catholics. As in Mexico and China, armed gangs in Spain fought out labour relations in the streets of Barcelona. The brutalities of the fascist Franco regime in Spain eradicated unions in pursuit of national harmony and favoured developing heavy industry.117 In pre-Revolutionary Russia, Marxist activists were literate, skilled textile workers, usually men in the St Petersburg printing trades.118 Popular liberal nationalism demanding social and political reform could be adopted by both working-class activists and colonials fighting imperialist 110 Lobato, ‘Argentina’; Camou and Maubrigades, ‘Uruguay’. 111 Quataert, ‘Ottoman Empire’. 112 Christensen, ‘Denmark’. 113 Cliver, ‘China’. 114 Ibid. 115 Ibid. 116 Ebeling, et al., ‘Germany’. 117 Smith et al., ‘Spain’. 118 Pretty, ‘Russia’. 745

The Ashgate Companion to the History of Textile Workers domination.119 Anti-colonial movements in India led by M.K. Gandhi used the plight of hand-spinners as a symbol of nationalist protest and ‘Swadeshi’ in Bengal in the 1920s associated its efforts with the handloom weavers.120 During the same period, organized Mexican and Egyptian textile workers in factory production forged strong ties with radical politics and anti-colonial nationalism. Rejecting both colonial rule and the dominance of foreign investors, Egyptian textile workers by the 1950s could work for nationalized mills under the Nasserist government while pressing unsuccessfully for radical labour unions independent of the state.121 Mexican operiarios influenced by European socialist ideas, like Italian and many other immigrants in the American Northeast, organized early. Mexican operiarios cooperated with communists and anarcho-syndicalists, who encouraged militancy, women’s activism and direct action including armed violence directed at supervisors to destroy their authority. During the 1920s this revolutionary movement hammered out some of the most progressive labour laws and protections in modern textile production.122 The Chinese Communist Party (CCP) was active among both former artisans and radical younger silk workers in Shanghai in 1927 and 1934. This spectre of radicalism badly scared mill owners and the Guomindang government who turned to policies of mass murder of militant leaders and intensifying class exploitation. As in Mexico, textile workers often turned to criminal gangs for support against employer/state terror. Still the Shanghai silk weavers, including ‘red’ worker sisterhoods, formed the core of CCP underground activities during the Second World War and facilitated the Communist takeover of the city in 1949.123 Nationalization of the cotton and silk industries in the 1950s coincided with enormous improvements in Chinese textile workers’ benefits, protections, education and wages. Still divisions developed between permanent, privileged workers and temporary workers. Chinese women workers received special benefits: pregnancy leave, childcare and new political respect as heroines of the Revolution.124 Textile workers in Communist China in the 1960s and 1970s had a well-earned and enviable position in an industry that by 2000 would command global superiority in production. Yet, the substitution of rural, less skilled, non-union labour for privileged urban workers and the rise of privatization in the 1980s and 1990s made cutting labour costs and the living standards of unionized textile workers the key to increasing worker productivity and insuring global supremacy in Chinese textile production.

119 Berger and Smith, ‘Between Scylla and Charybdis’. 120 Roy, ‘India’. 121 Beinin, ‘Egypt’. 122 Bortz, Mexico’. 123 Cliver, ‘China’. 124 Ibid. 746

Investigating identities within the global textile workforce

Conclusions A global perspective on the identity of textile workers de-centres the Western impression that the human image of textile industrialization can be captured by the Rhode Island child spinner, the Lowell factory girl, the ‘barefoot aristocrat’ mule spinner of Lancashire and the angry, de-skilled wool-comber of Yorkshire. The complexities of global industry add a vast array of Central European, Asian, South Asian, Middle Eastern and Latin American peoples with differing cultural, racial and political identities to the well-known portraits of American and West European operatives. The longevity and market reach of systems of production in India, China, Japan and the Ottoman Empire diminish the significance of Anglo-American production, which in global terms (1650–2000) entered the market place briefly with mechanized factories and after a century of dominance became eclipsed by non-Western textile industrialization. In comparison with the radical labour politics of Mexican, Spanish, Egyptian and Chinese textile workers, the craft unionism and goals of British and American operatives were deeply conservative. The major experience, however, that unites the identities of textile workers in a global sense is class exploitation, whether by mechanization and de-skilling, by repressive regimes or by national policies that undercut hard-won worker rights.

747

This page has been left blank intentionally

30 Institutions in textile production: guilds and trade unions Lars K. Christensen

Textile production is a social activity. As such, it is embedded in an institutional framework. This chapter will discuss those institutions formed by and reproduced by the direct producers: guilds and trade unions. Commenting on industrial relations research, Richard Hyman has defined comparison as ‘the systematic cross-analysis of phenomena displaying both similarities and differences’. However, the empirical basis for this chapter is mainly the national overviews in the present volume. Naturally, the focal points of these overviews were chosen by their respective authors – and not necessarily with the primary focus on crossanalysis of guilds and trade unions. For this reason alone, if not for others, this chapter makes no claim to be comparative in the meaning above. The ambition is more limited: to present questions of guilds and trade unions across the global textile industry in a more or less systematic way, highlighting both common issues and important differences.

Guilds Organizing artisan production in the cities in an institutional framework, known as guilds in the English language, is a phenomenon that dates back to at least the early Middle Ages. In the Islamic Empire, institutions known as naqabat rose in the larger cities from the ninth century, organizing artisans according to their trade, controlling prices and the quality of production. These guilds also developed their own rules and rituals for promoting apprentices to journeymen and journeymen  In this context, institutions are understood in a broad sense: as types of social practice with structural properties and a certain extension in time and/or space. Anthony Giddens, The Constitution of Society (Cambridge, 1984), p. 17.  Richard Hyman, ‘Trade Union Research and Cross-national Comparison’, European Journal of Industrial Relations 7 (2001): 203–232, at p. 205.

The Ashgate Companion to the History of Textile Workers to masters. The guilds had their own statutes and aldermen, either elected by the artisans themselves or appointed by the city council. Institutions similar in structure and functions are known in several parts of the world. In Europe, guilds seem to have come into existence, along with the rise of an urban civilization, from around the eleventh century. In Asia, guild-type institutions existed in India, China and Japan. While in India they were of very ancient origin, in Japan they were not established before the fourteenth century. From this evidence, Bo Gustafsson has concluded that craft guilds were a ‘generally occurring city-based industrial form of production in all pre-industrial/pre-capitalist economies’. Research on the history of textile production seems to confirm this. Apart from Europe, guilds were a feature of pre-industrial textile production in the Ottoman Empire, shortly after its formation, and were established in Egypt as part of that empire. In China and Japan, guild-like institutions in textile production evolved as well. In other non-European countries, guilds were set up as part of European colonial rule, as in Mexico. One example stands out because of an explicit lack of significant guilds: the USA. The reason for this might be that pre-industrial textile production in North America was primarily rural and family based, and that by the time of colonization, guilds were already losing their importance in the colonial power, England. In Europe, guilds are often associated with the Middle Ages. Fernand Braudel, in his monumental work on early modern Europe, claims that the ‘great days’ of the guilds were over already in the fifteenth century, at least in Western Europe, although they ‘stubbornly’ survived in some places, primarily in Germany. Several examples of this volume suggest this claim to be exaggerated. Guilds played a role in eighteenth- and even nineteenth-century textile production in a number of countries. In Europe, male artisans in urban areas were organized in guilds in Austria and Bohemia, Denmark, Germany, the Netherlands and Spain.10 Of the European countries treated in this volume only in England did guilds seem  Bo Gustafsson, ‘The Rise and Economic Behaviour of Medieval Craft Guilds’, in Bo Gustafsson (ed.), Power and Economic Institutions: Reinterpretations in Economic History (Aldershot, 1991), pp. 69–106, at pp. 69–70.  Ibid., p. 69.  Quataert, Chapter 18, this volume, ‘Ottoman Empire’; Beinin, Chapter 7, this volume, ‘Egypt’.  Cliver, Chapter 5, this volume, ‘China’; Hunter and Macnaughtan, Chapter 12, this volume, ‘Japan’.  Bortz, Chapter 13, this volume, ‘Mexico’.  Blewett, Chapter 21, this volume, ‘USA’.  Fernand Braudel, Civilization and Capitalism, 15th–18th Century, vol. II: The Wheels of Commerce (New York, 1982), p. 315. 10 Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’; Christensen, Chapter 6, this volume, ‘Denmark’; Richard Biernacki, The Fabrication of Labor: Germany and Britain, 1640–1914 (Berkeley, 1995), pp. 259ff; Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’; Smith et al., Chapter 17, this volume, ‘Spain’. 750

Institutions in textile production to have been without significance in eighteenth- and nineteenth-century textile production. The functions of guilds normally included controlling the distribution of raw materials and labour power as well as customers between the master artisans. They also upheld internal discipline, preventing masters from competing among themselves, and enforced social norms and values, such as that of delivering ‘honourable’ work. Guilds also had an excluding function, defining which persons could be allowed into the trade and which not. It seems to have been a common feature, from Western Europe over the Ottoman Empire to China, that women were excluded, even though widows in some instances were allowed to carry on the business of their deceased husband.11 Persons considered ‘unworthy’, such as those born outside of marriage, would in some countries not be allowed apprenticeship. To become a master, at least in Europe, citizenship was usually required, which often implied affiliation to a certain religious confession. In contrast, the guilds of the Ottoman Empire normally contained members of more than one religious denomination.12 The causes behind the creation of guilds, as well as their primary functions, have been a matter of debate. Traditionally, their role as up-keepers of a monopoly has been emphasized. In the eyes of Karl Marx, this was part of a class struggle between artisans and merchants. By defending the occupational monopoly of trained artisans and enforcing limitations on the number of journeymen that could be employed by each master, guilds blocked the development of a capitalist employer–labour relationship in the trades.13 Braudel later repeated this viewpoint. In a more recent contribution, Gustafsson has asked the obvious question of how guilds came into a position which allowed them to enforce this monopoly. Guilds were normally dependent on having their statutes approved by city councils and other authorities, which might be dominated by merchants or other classes, with interests dissimilar to the artisans. Gustafsson points to the problem of upholding a certain level of quality of the products. In a pre-industrial economy, he claims, markets have functional problems, because of consumers’ lack of information about product quality. This is a structural problem, which threatens to undermine both consumers’ willingness to buy and sellers’ possibilities of obtaining a price corresponding to the value of their product. For this reason, artisans, consumers and authorities had a shared interest in quality control. Thus, as guilds were able to guarantee a certain minimum level of quality, they were able to obtain and maintain a monopoly position.14 Gustafsson’s argument is supported by the fact that quality control plays a prominent role both in normative evidence such as guild statutes and in evidence of the practice of the guilds. This was not the least the case in textile production. 11 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’; Quataert, ‘Ottoman Empire’; Cliver, ‘China’. 12 Quataert, ‘Ottoman Empire’. 13 Karl Marx, Das Kapital, vol. 1 (Berlin, 1977 [1867]), p. 380. 14 Gustafsson, ‘Rise and Economic Behaviour’, pp. 81f. 751

The Ashgate Companion to the History of Textile Workers It could be hard for customers to judge, for instance, if inferior materials had been mixed into the wool, sufficient numbers of thread had been used or the cloth had been dyed sufficiently. Furthermore, cloth was sometimes exported and sold far away from the point of production. For this reason, some guilds employed trademarks, which were applied only after vigorous quality control. Quality control might be an important function of the guilds, and an explanation for their success in upholding monopoly, but it does not exclude the class struggle, which Marx and Braudel pointed to. It is even plausible that as capitalism and an embryonic market economy evolved, this function of the guilds became more important. Some of the national overviews show that emergent capitalism and increased market competition made guilds strengthen their inner organization and/or their political appearance, at least for a limited period. In some parts of the Ottoman Empire, guild masters were able to strengthen their monopolistic practices during the second half of the eighteenth century. In other parts of the empire, economic developments swept away the guilds.15 There is also evidence that guilds adapted to changes in the production process, caused by economic development, and thereby broke away from the traditional model of a small workshop with a master and a few journeymen. When faced with competition from manufactories favoured by state privileges, guilds in the Habsburg Monarchy established their own putting-out networks, as well as joining those set up by manufactures.16 On the other hand, master artisans were sometimes beaten on their own battleground. The Spanish example is illuminating: during the seventeenth and eighteenth centuries the Spanish guilds were transformed. Apart from regional differences, the process mainly comprised the establishment of the merchant as the central figure. This was partly a result of specialization among the artisan masters, which led to an atomized production process and limited investment power. The merchants were able to acquire the role not only of investors, but also of coordinators of the production. In some places the guilds were reorganized, and masters and journeymen were subordinated under one single merchant guild.17 In this case, the function of the guilds in fact changed from representing the artisans’ interests against the merchants to the opposite. When the capitalist mode of production became dominant, and free markets were established, the guilds were often fierce – but usually futile – in their resistance to liberal capitalism. In Germany they petitioned massively against free trade during the revolution of 1848–1849. As late as 1860 some artisans decried free trade as an artificial invention, dreamed up by intellectuals with no insight into the trades.18 Guilds and states seem to have had an ambivalent relationship to each other. From the point of view of the early modern state – especially in its absolutist form – the guilds could be seen as competitors for power in the cities. Nevertheless, absolute monarchies could also benefit from the guilds. As part of its mercantilist 15 Quataert, ‘Ottoman Empire’. 16 Komlosy, ‘Habsburg Monarchy’. 17 Smith et al., ‘Spain’. 18 Biernacki, Fabrication of Labor, pp. 288–289. 752

Institutions in textile production aspirations, the Danish monarchy established so-called factory guilds in the late eighteenth century, in order to secure a supply of labour and fixed wages to textile manufactories.19 In the same period, the French guilds were restored and consolidated by the monarchy after attacks by reform-minded philosophers and politicians, supposedly because they were seen as a valuable aid to the upholding of a stable, hierarchical society.20 It was also the desire for political stability, combined with tax policies, which made the Ottoman sultan support the consolidation of guilds in the early nineteenth century.21 But generally, the overthrow of absolutism and creation of liberal regimes, as it took place in parts of Europe in the late eighteenth and early nineteenth centuries, seems to have been a major deathblow to the guilds. In England, the guilds had already been weakened by the revolution in the late seventeenth century and were virtually extinct by the mid-eighteenth century. In France, the revolutionary government abolished any obstruction to free trade, including guilds and any other regulations of labour, in 1791. In 1810–1811 the Prussian government lifted the requirements that entrepreneurs should belong to a guild, and abolished wage regulations. Other German states soon followed this example. But guilds as such were not abolished, and they continued to be influential throughout the rest of the century. German guilds were able to exert influence on technical training, and guild membership remained a sign of social status. In Japan, the Meiji Restoration government of 1868 formally abolished guild privileges.22 As capitalism progressed, the guilds proved unable to handle the new type of social relations between labour and capital. The new antagonism between labour and capital superseded the traditional internal community of the trade. Some journeymen’s guilds took on new forms. As an example, journeymen in Damascus regularly performed collective, disciplined actions against their masters from at least the 1870s, in order to achieve higher wages and better living conditions.23 In the decades around 1900, silk weavers’ guilds protested actively against declining wages and layoffs in several parts of China.24 Nevertheless, guilds were hampered in taking on the challenges of capitalism by their origin and fundamental interconnectedness with earlier modes of production. In their attempts to defend the interests of journeymen, they were generally looking backward, promoting ideas of restoring earlier, and now historically obsolete, types of social relations. As institutions they were becoming dysfunctional.

19 20 21 22

Christensen, ‘Denmark’. Biernacki, Fabrication of Labor, p. 324. Quataert, ‘Ottoman Empire’. Biernacki, Fabrication of Labor, pp. 235–236, 324–325, 266, 285f; Hunter and Macnaughtan, ‘Japan’. 23 Quataert, ‘Ottoman Empire’. 24 Cliver, ‘China’. 753

The Ashgate Companion to the History of Textile Workers

Trade unions With new types of social relations in the capitalist mode of production, a new type of institution emerged: trade unions. Ross M. Martin points out that trade unions are not only perhaps the largest group of non-governmental organizations in the world today, but furthermore ‘from a bird’s-eye view, they exhibit a bewildering diversity of structure and, particularly, of behaviour’.25 In some countries textile trade unions emerged early, and grew into mass organizations of national scope in only a few decades. The history of the textile labour movement in these countries was certainly not without sacrifice, conflict and dramatic moments. Nevertheless, the overall narrative is one of evolution. Yet from a global point of view, these cases – which we could label ‘the North European model’ – are in fact an exception. The majority of national overviews in this volume tell us about discontinuities and abrupt changes in the organization of textile workers. Not only in their scope, but also in their strategies, ideology and relationship to employers and state, textile unions display a great variety over time and place. There is a close link – empirically as well as logically – between the historical transformation of labour power into a commodity and the emergence of trade unions. According to classical liberal economy, sellers and buyers of labour power should face each other as individuals on the labour market of the capitalist economy. But almost anywhere at some time, workers grouped together in associations in order to strengthen their position in the market versus the employers. A famous definition of trade unions is the one offered by Sidney and Beatrice Webb in 1894: ‘A continuous association of wage earners for the purpose of maintaining or improving the conditions of their employment.’26 This definition, however, has been subject to much criticism. As demonstrated by Martin, examples can be found of a very broad spectrum of positions concerning the purpose of trade unions. This leads him to conclude that any definition of trade unions which includes purpose is basically an expression of ideological preferences, and trade unions as institutions should be defined only by their membership.27 However, this would in fact lead to further confusion. Without taking purpose into consideration, we would be unable to distinguish between a trade union and a sports club or a choir organized by workers. Consequently, a trade union should be understood as an association of wage labourers for the purpose of advancing their common economic and social interests, at least in relation to their employers but sometimes beyond. This definition implies that a trade union must strive to advance the interests of its members versus the employers, but it does not stipulate whether these interests are limited to higher wages and lower working hours, or whether the ultimate goal should be much more 25 Ross M. Martin, Trade Unionism: Purpose and Forms (Oxford, 1989), p. 106. 26 Quoted from Richard Hyman, Understanding European Trade Unionism: Between Market, Class & Society (London, 2001), p. 6. 27 Martin, Trade Unionism, esp. pp. 95–97. 754

Institutions in textile production encompassing. Sometimes the concept of trade union has been used specifically to designate an association comprising skilled (male) workers, affiliated to a certain trade and often local or regional in scope. In this respect a trade union is a certain type of association, different from, for example, an amalgamation of unskilled industrial workers. However, in the following discussion trade union will be used as a general concept, including all types of institutions fitting the above definition.

Labour power as a commodity and the rise of unionism Historically, the collective effort of improving working conditions has often been initiated by a certain group of workers, with their own particular interests in mind. The attempt would be to control the supply of labour power, and to keep others away from the same jobs. Logically, some differences inside the workforce, which could be used to set up distinctions between ‘us’ and ‘them’, had to be accentuated. These differences could be related to the work process, such as the skill required to carry out certain jobs, but they did not have to be. Any difference experienced as culturally meaningful would do. A prototypical example is that of the spinners in the Lancashire cotton industry. There was no formal apprenticeship for mule spinners, but nonetheless it was a job requiring technical skills. These skills were obtained by working as a spinner’s assistant. Since it was the privilege of Lancashire mule spinners to hire their own assistants, they were in fact able to exercise a high degree of control over the supply of labour. On this basis, mass trade unionism was established among the spinners as early as the 1870s.28 In this example, the establishing and upholding of a monopoly was based on clever exploitation of certain aspects of the organization of production. In other cases, formal training and artisan traditions could be a repository for attempts to monopolize certain jobs. However, as industrialization progressed, textile work seems to have become generally unskilled, or semi-skilled based on a short period of training in the workplace, while only jobs as overseers and foremen required more formal training. Distinctions based on formal training thus became less relevant among blue-collar workers. Another difference used for creating distinctions was that of gender. An argument widely used for reserving certain jobs for a certain sex was that of physiology: jobs requiring physical strength necessarily had to be carried out by men. This was the alleged reason why mule spinners had to be men, not only in Lancashire, but also in other places where the same technology was adopted, such as in the Catalan textile industry. When self-acting mules were introduced in the 1850s, there was no longer the same demand for physical strength. However, attempts to replace male spinners with female were successfully resisted in both Lancashire and Catalonia, and the male workers were largely successful in retaining relatively high wages and a powerful position in the production process.29 28 Fowler, Chapter 9, this volume, ‘Great Britain’. 29 Smith et al., ‘Spain’. 755

The Ashgate Companion to the History of Textile Workers Jobs performed by one sex in one context – geographically or historically – could very well be performed by the other in another context. Furthermore, the gender division of jobs sometimes displayed a striking lack of consequence – while female workers were banned from some jobs, such as mule spinning, on the pretext of the job being too physically demanding, they could very well carry out other physically demanding jobs, such as wool cleaning.30 Gender divisions of jobs were basically a social construction, influenced by a number of variables such as technological changes in the work process or traditional concepts of male and female – but in the end determined by the organized strength of the actors. It seems to be a more or less general experience that male workers were the first organizers of trade unions. Where national narratives give examples of early attempts by women to form trade unions, it is mostly as rare exceptions to this overall rule.31 It seems plausible that the early labour movement, at least to some extent, reflected and reproduced the concepts of gender dominant in the society it was part of. Other differences that could be used for claiming certain rights or jobs for certain groups were those of culture in general, whether based on ethnicity, race, religion, place of origin or something else. But the attempt to defend a certain job territory, based on skill, gender, birthplace or other premises was a vulnerable one. It might create internal conflicts among workers, such as when a strike initiated by Catholic immigrants was broken by native-born Protestants in mid-nineteenthcentury New England, USA.32 One way to overcome this was to expand the scope of organization, so that groups of workers who might otherwise become competitors in the labour market would become members of the same organization. Typically, unions exclusively organizing men with skilled or semi-skilled status experienced the need to open up their union to unskilled female workers.33 When facing the challenges of the labour market, workers were often inspired by cultural practices already known to them, and tried to apply them to their present needs. In Britain, where the industrialization of textile production began, nascent trade unions in the mid-eighteenth-century cotton industry drew upon existing traditions among handloom weavers, such as friendly societies.34 In Denmark, male handloom weavers, with a tradition of apprenticeship and zünft, were also the first textile workers to form a trade union.35 Drawing upon traditional artisan forms of bonding is a well-known feature, especially in the early European labour movement. The fact that the cultural forms and practices of some trade unions borrowed a lot from guild traditions should not lead us to obscure the principal differences 30 Christensen, ‘Denmark’; Pretty, Chapter 16, this volume, ‘Russia’. 31 One notable example, as mentioned later, is the Lowell Female Labor Reform Association. 32 Blewett, ‘USA’. 33 Christensen, ‘Denmark’. 34 Fowler, ‘Great Britain’. 35 Christensen, ‘Denmark’. 756

Institutions in textile production between the two types of organization. While guilds defended the interests of a certain trade, unions are basically defending the interests of a certain class. Guilds were traditional institutions, often backing their claims with reference to history. Unions are modern institutions, aiming to defend labour interests by adapting to the challenges of modernity. If not before, this becomes evident as unions grew into mass organizations.36

Union growth Late nineteenth-century Europe experienced a general wave of worker activism and mass unionization, especially in the economic growth period from around 1889. The textile industry followed this general pattern. In Britain, where mass trade unionism had already developed earlier in Lancashire, a national federation, the United Textile Factory Workers Association, was established in the 1880s.37 Even though other European countries were industrialized later than Britain, the development of trade unions took place not much later. In Germany, an important wave of unionizing took place in the 1890s, and a national federation, the Deutscher Textilarbeiterverband (DTAV) was created.38 In Denmark a national federation, established in 1885, grew into a mass organization during the 1890s.39 The first weavers’ and spinners’ union in the Netherlands was founded in 1889.40 Despite unions being illegal, there was a mayor strike wave in the Russian textile industry in 1895–1900, which was also a period of accelerating industrialization.41 The textile industry came early to the United States, in the beginning concentrated in New England, and with it arrived trade unions. The Lowell Female Labor Reform Association, active from the 1820s and into the 1850s, is not just an example of an early trade union, but also exceptional because it organized female workers and was led by one.42 Among the Latin American nations Mexico stands out because both the textile industry and textile workers’ unions were established quite early. Trade union activity dates back to the early 1870s, and from being heavily repressed unions gained a strong position following the Revolution in 1910.43

36 For a further discussion on the concept of modernity in relation to labour and unions, see Lars Christensen, Det moderne arbejde. Kulturelle og institutionelle forandringer af arbejdet i den danske tekstilindustri 1895–1940 (Ph.D. Dissertation, University of Copenhagen, 1998), esp. pp. 283f. 37 Hyman, Understanding European Trade Unionism, p. 79. 38 J. Daumas, R. Cazals, M. Ellerkamp and A. Fowler, ‘Trade Unionism in Textile Towns and Areas’, in J. Robert, A. Prost and C. Wrigley (eds), The Emergence of European Trade Unionism (Aldershot, 2004), pp. 54–90, at p. 57. 39 Christensen, ‘Denmark’. 40 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 41 Pretty, ‘Russia’. 42 Blewett, ‘USA’. 43 Bortz, ‘Mexico’. 757

The Ashgate Companion to the History of Textile Workers But in most other parts of the world, trade unions developed later. Possible reasons are later industrialization of textile production, a workforce recruited from a rural background with no tradition for organization, repression from employers and state – or a combination of these. In Asia, unions seem to have existed in Japan from the 1890s, but without much involvement from textile workers. At the end of the First World War there was a rise in activity among textile workers, including a number of major strikes. However, textile unions remained weak, and the labour movement in general was politically repressed in the 1930s.44 In the premier textile mill centre of India, Bombay, sporadic strikes and labour protest also took place in the early twentieth century. The first successful attempt at organizing was the Bombay Textile Labour Union in 1926. Its membership fluctuated widely, but in 1938 one-third of the factory workforce in this city was organized.45 Textile workers played a central role in the forming of the Turkish labour movement. Unions, however, were not made legal before 1946. It was the Textile and Knitting Workers Federation (TEKSIF) that took the initiative to establish the first national trade union federation, TÜRK-IS, in 1952.46 The global history of textile work confirms the close relationship between industrialization and trade unionism. Trade unions can, and did, arise in pre-industrialized, market-based types of textile production. But it was industrialized production that created the basis for mass unionism.

Politics and ideology Trade unions have been categorized in a number of ways in relation to ideology.47 Any ideological categorization, however, should be deduced from the subject itself, with the primary purpose of trade unions as its starting point. Starting from the labour–employer relationship, two principal positions can be distinguished. The first regards labour and employers as principal antagonists and class struggle as inevitable. The second may recognize the existence of class conflicts, but does not view class struggle as inevitable. Any empirical example of trade union ideology is a sub-category of one of these. The first major position is mainly made up of anarchists, syndicalists and different kinds of socialists. The earliest example of a relatively widespread socialist labour movement among textile workers is probably Germany. Already in the 1860s, strikes and actions among German weavers had acquired a political character. The Lasallean Allgemeine Deutsche Arbeiterverein in particular, but also the competing Sozialdemokratische Arbeiterpartei were influential among textile 44 Hunter and Macnaughtan, ‘Japan’. 45 Roy, Chapter 10, this volume, ‘India’. 46 Seidman, Chapter 19, this volume, ‘Turkey’. 47 As an example, Ross M. Martin has divided trade unions into five principal categories: Pluralists, Syndicalists, Marxist-Leninist, Organicists and Authoritarians. Martin, Trade Unionism, pp. 12f. 758

Institutions in textile production workers.48 In the same period, the International Working Men’s Association, led by Karl Marx in London, was an inspiration to organize socialist labour movements, including weavers’ unions, in such different corners of the world as Denmark and Mexico. Karl Marx had emphasized the role of trade unions, not only in the day-today struggles, but also as ‘schools’ for the proletariat, teaching them the necessity of fighting against capitalism as such. But Marx and Engels also believed that in order to raise the political awareness of trade unions, an external force was needed, in the form of a socialist political party. This was the model followed by the socialist labour movement in countries like Germany, Denmark and – with some modifications – Britain, where mass trade unions became interlinked with Social Democratic parties. Lenin later accentuated this division of labour between trade unions and a political party into a more radical position, strongly emphasizing the need for a party comprised of professional revolutionaries. Trade unions should be subordinated to the Communist Party.49 An almost opposite viewpoint was that of syndicalism, which held that the road to socialism went through militant unionism, culminating in a general strike. Involvement in party politics would only divert the revolutionary energy of the working class into parliamentary intrigues. Syndicalism was an important tendency in French and Spanish trade unions, including those in the textile industry, as early as the beginning of the twentieth century.50 In the turbulent years around the First World War, and especially following the Russian Revolution, there was a general swing to the left in many labour movements. Revolutionary socialist parties were formed in most European countries. An example of radical labour movements outside Europe can be found in Mexico, where a national revolution in 1910 led to the radicalization of trade unions. In China, another country with a growing textile industry, the Russian Revolution inspired the creation of the Chinese Communist Party (CCP). However, even though younger workers in the textile centre of Shanghai were becoming politically radicalized, the CCP was not able to gain any significant support among textile workers before the Second World War.51 After the defeat of fascism, socialists either reaffirmed or regained their role in the trade unions of North and western Europe. In countries such as England, Germany and Denmark, social democracy was the de facto leading ideology in the trade union movement. But at the same time, this ideology had been redefined. During the twentieth century, many Social Democratic parties became convinced that economic growth through improved productivity would make class struggle obsolete. In this respect, social democracy made a bold move out of its original 48 Ebeling et al., Chapter 8, this volume, ‘Germany’. 49 V.I. Lenin, ‘What is to be done?’, in V.I. Lenin, Collected Works, vol. 5 (Moscow, 1961), pp. 347–530, esp. pp. 397f. 50 Hyman, Understanding European Trade Unionism, pp. 22–23. Daumas et al., Trade Unionism in Textile Towns, p. 83. 51 Cliver, ‘China’. 759

The Ashgate Companion to the History of Textile Workers position of class antagonism towards class cooperation. More radical versions of socialism were still influential, though, in eastern and southern Europe as well as in countries outside Europe. All in all, socialism – in a great number of variations – has been an important ideological and political current in the global textile labour movement. However, it has always been contested – and in some countries clearly outnumbered – by other ideological currents which belong to the other principal position: that class conflicts are either reconcilable or must be subordinated to a greater principle, such as nationality, the state or religion. In several countries in Europe, unions based on a Christian ideology have had relatively strong support. In Germany, the Catholic Church had a strong position in textile-producing regions. Christian unions were formed, which raised much the same types of demands as the secular unions. The growth of Christian trade unions among German textile workers resulted in the formation of the Christlich-soziale Textilarbeiterverband in 1896 and the Zentralverband christlicher Textilarbeiter in 1900/1901.52 Shortly after the establishment of the socialist trade union in the Netherlands, the Catholic Church founded an alternative union. This was again followed by initiatives from the Protestant Church.53 Christian unions could be very critical of the effects of capitalism, but at the same time denounced the concept of class struggle. In Europe, Christian trade unions became gradually ‘de-confessionalized’ after the Second World War. In Germany, the Catholic workers became integrated in the new trade unions of the Federal Republic of Germany, which, even if mostly lead by social democrats, were politically unified and committed to ‘social partnership’.54 Trade unions based on religion were not limited to Christianity. As late as 1976 a new national trade unions council, HAK-IS, was created in Turkey, with an Islamic focus, competing with two already existing organizations. HAK-IS also represented textile trade unions.55 Other ideological orientations that have played an important role in textile labour movements are those of nationalism, predominantly in countries in a process of national liberation such as India or Egypt, and populist or authoritarian variants of social reformism, such as Peronism in Argentina. Finally, some unions have been built on the basis of being specifically apolitical, concentrating on securing immediate, mostly economic benefits. This is also known as ‘Business Unionism’. A prominent example of a union of this kind is the United States national union federation, the American Federation of Labor,56 which also included the United Textile Workers. Another example is the first national trade union federation of

52 Ebeling et al., ‘Germany’. 53 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 54 Hyman, Understanding European Trade Unionism, pp. 52–53, 116–117. 55 Seidman, ‘Turkey’. 56 Roy J. Adams, Industrial Relations Under Liberal Democracy: North America in Comparative Perspective (Columbia, 1995), pp. 36f. 760

Institutions in textile production Turkey, the TÜRK-IS, conceived in 1952 at the congress of the Textile and Knitting Workers Federation.57 While attitudes towards class and class struggle can be used to characterize the ideological and political radicalism of trade unions, they should not be confused with the question of militancy. Christian trade unions in late nineteenth- and early twentieth-century Europe did not necessarily refrain from bringing forward considerable demands for the improvement of workers’ conditions. And even though these unions often are characterized as ‘integrative’ in sociological terms, they did occasionally organize strikes and other militant actions.58

Collective bargaining As unions grew in size and strength, a new type of relationship between unions and employers evolved in some countries, generally known as collective bargaining. Collective bargaining can be defined as a situation where terms of work and pay are made subject to negotiations between employers and workers, with the latter represented as a group by their union(s). However, this basic principle can be realized in a very wide variety of ways. Most importantly the differences can be in the level of bargaining (from factory level to nationwide), the extent (the relative proportion of workers covered) and the scope (the range of working conditions covered).59 Collective bargaining was established early in British Lancashire, parallel to the creation of mass unionism. Collective bargaining and agreements in Britain, however, for a long time mostly took place at local and county level, and were not consolidated at the national level before the interwar period.60 Denmark, with its very high rate of organization – reaching more than 80 per cent of the textile workforce in one single union in the interwar period – is also one of the countries with the earliest and most far-reaching example of industrial relations built on bargaining and agreements. Here, a national agreement for the textile industry had already been reached by 1898, and centralized collective bargaining at the national level has been in effect ever since.61 As a non-European example, the Argentinean textile workers union, the UOT, succeeded in establishing a bargained agreement in the wool sector in 1936.62 However, there are in fact rather few reports of extensive use of collective 57 Seidman, ‘Turkey’. 58 Hyman, Understanding European Trade Unionism, p. 41. 59 Hugh A. Clegg, Trade Unionism under Collective Bargaining: A Theory Based on Comparisons of Six Countries (Oxford, 1976), pp. 5–9; Keith Sissons, The Management of Collective Bargaining: An International Comparison (Oxford, 1987), pp. 2–3. 60 Fowler, ‘Great Britain’; Daumas et al., Trade Unionism in Textile Towns, pp. 56–57, Hyman, Understanding European Trade Unionism, pp. 78, 91. 61 Christensen, ‘Denmark’. 62 Lobato, Chapter 2, this volume, ‘Argentina’. 761

The Ashgate Companion to the History of Textile Workers bargaining in the textile industry, at least until the second part of the twentieth century. In Germany, collective agreements did not follow mass unionism in the textile industry, even though they became widespread in other sectors.63 In other countries, attempts by trade unions to establish collective bargaining were met with fierce resistance from employers or even repression from the state. The model of collective bargaining was effectively challenged by another model in the mid-twentieth century, that of corporatism. According to corporatism, labour and employers should not be seen as classes, but as parts of the common body of the state, and preferably subsumed under its leadership. Corporatist structures for regulating industrial relations by the state were established by fascist or conservative authoritarian regimes in a number of European countries in the 1920s and 1930s. However, corporatism has been part of the ideology of a number of other regimes during the twentieth century. Examples relevant to this volume are the populist government of Perón in Argentina and the nationalist government of Egypt.64 After the Second World War the model of collective bargaining gained new impetus in some countries. In Japan, where trade unions had been suppressed since the 1930s, a national textile union was reorganized. In 1955 a wage strike in the textile industry started what was to become a permanent feature of post-war Japan: a yearly ‘spring labour offensive’ of collective bargaining. From the 1960s onwards, disputes in the textile industry were largely confined to wages, and mostly ended with settlements. Another feature of the Japanese textile industry during this period of economic expansion was the establishing of close cooperation between labour, employers and the state, aimed at modernizing and furthering the national industry. The textile union federation became one of the most powerful unions in Japan, and was politically linked to the all-dominating political party. The union, the employers’ organization and the Ministry of International Trade and Industry cooperated closely on issues such as labour politics, technology development and market information.65 In the Federal Republic of Germany, autonomous trade unions were re-established after the defeat of fascism. The original programme of these unions included demands such as nationalization of core industries and workers codetermination in management and economy. However, in the framework of the Wirtschaftswunder – the growing strength of the German economy – such demands were downplayed. Instead, trade unions were integrated as partners in a ‘social market economy’, based on collective bargaining.66 Collective bargaining had existed continuously in Denmark since the end of the nineteenth century. Cooperation between labour, employers and the state was strengthened during the economic growth of the 1960s. The agreements reached between unions and employers became elements

63 Daumas et al., Trade Unionism in Textile Towns, pp. 59–60. 64 I shall return to these examples in more detail below. 65 Hunter and Macnaughtan, ‘Japan’. 66 Hyman, Understanding European Trade Unionism, pp. 115f. 762

Institutions in textile production of an income policy, negotiated at tripartite meetings between unions, employers’ associations and the government.67 These countries are examples of an effort to rebuild or modernize production and integrate it into a ‘Western’ market economy during the period of the Cold War. Based on economic growth, industry or state welfare plans were introduced, and standards of living improved considerably. Labour was politically and ideologically integrated into society, and more radical elements of the labour movement were marginalized. How do we explain the differences in the development of collective bargaining in different countries? In his comparative study of collective bargaining in six countries, Hugh Clegg emphasizes the role of the employers. This includes not only the attitudes towards recognizing unions as bargaining partners, but also the level and structure of organization among employers. Strong and centralized employers’ organization seems to favour a uniform policy towards labour among employers, including acceptance of collective bargaining.68 Examples from the textile industry do not counter this. But it cannot be the sole explanation of the differences between national developments.

Unions and the state In the classical Marxist view, the state is an instrument for the ruling class, used to dominate the subordinated classes and upholding an existing social order. One example of state operation that comes close to this is the Spanish textile labour movement, from its beginning in the mid-nineteenth century and well into the post-war period. The first viable union of textile workers, Federación de las Tres Clases de Vapor (TCV), was established in Catalonia in 1869. The TCV was striving to establish collective bargaining rights in a spirit of class collaborationism. However, mill owners resisted negotiation, and a combination of a conservative government and hard-line Catalan industrialists succeeded in breaking the union. A number of attempts to revive the Catalan textile labour movement in the early twentieth century were likewise met with fierce resistance from employers. In the beginning, a socialist leadership renewed the pressure for collective bargaining. The employers continued to reject this, and in an environment characterized by social conflict, the anarcho-syndicalist labour confederation, CNT, grew rapidly in 67 Some scholars have introduced the concept of neo- or social corporatism to describe especially the northern European types of labour market and economic regulations of the late twentieth century. For an introduction, see Göran Therborn, ‘Lessons from ‘Corporatist’ Theorizations’, in J. Pekkarinen, M. Pohjola and B. Rowthorn (eds), Social Corporatism (Oxford, 1992), pp. 24–43. 68 Clegg, Trade Unionism under Collective Bargaining, pp. 106–108. The countries examined by Clegg are Australia, France, Sweden, the United Kingdom, the USA and West Germany. 763

The Ashgate Companion to the History of Textile Workers the years 1918–1919. Militancy among the workers and repression from employers and state were growing. In 1923, a military dictatorship was established. In the Second Republic, established in 1931, the labour movement was able, for the first time, to influence the state through government. But, as is well known, parts of the Spanish military initiated the civil war of 1936–1939, and the victorious military dictatorship eradicated independent, class-based unions and established an authoritarian, corporatist regime.69 The strategy of the initial Catalan textile trade union, the TCV, in many ways resembled what led to stable, union-based industrial relations in other countries. Later on, Socialists and other Marxists argued for collective bargaining. However, neither of these opportunities was developed, because for most of the period the Spanish employers had at their disposal a state either passively accepting employers’ repression of trade unions or actively engaged in the same repression. However, in the same period a union–state relationship almost opposite to the Spanish experience evolved in countries such as Britain and Denmark. In both countries, a labour movement arose during the process of industrialization. In both countries, the initial response from the state was that of repression. But nevertheless trade unions had established themselves by the second half of the nineteenth century and were de facto, if somewhat grudgingly accepted by both employers and the state. At the heart of this process lay the formation of industrial relations which distributed and balanced power and responsibilities between unions, employers and the state. As unions became a mass phenomenon in some countries, both leading industrialists and representatives of the state began to see the advantages of being able to deal with labour in an organized and institutionalized manner. A Royal Commission on Labour, established in Britain in 1891–1894, stated the very clear conclusion that ‘peaceable relations [on the labour market] are, upon the whole, the result of strong and firmly established trade unionism’.70 Trade unions were, in the words of Richard Hyman, ‘accepted as part of the solution rather than part of the problem’.71 What was established in Britain and Denmark was a class compromise, with the state acting as an intermediary. Trade union activities – including, under certain specific conditions, strikes – were recognized as legitimate. Trade unions, on the other hand, accepted employers’ right to exercise their private ownership of the means of production, and the legitimacy of a state, committed to upholding that principle. Trade unions also committed themselves to uphold internal discipline in the working class. The Social Democratic leadership of the Danish Textile Workers’ Federation was relentless in combating militant tendencies among the union’s regular members, including expulsion of members involved in non-sanctioned strikes.72 While the Spanish experience can be seen as an example of a contesting 69 Smith et al., ‘Spain’. 70 Hyman, Understanding European Trade Unionism, p. 77. 71 Ibid. 72 Christensen, ‘Denmark’. 764

Institutions in textile production relationship between unions and state, the British and Danish experiences can be seen as an example of mutual recognition. This is clearly one dimension of the union–state relationship.

Autonomy or dependence? In the above examples, regardless if unions were repressed or recognized as partners, they were institutionally and politically independent from the state. But there are also a number of textile labour movements that managed to extort substantial concessions from employers and/or the state, but lost their autonomy as part of the process. In periods of power struggle between different factions of the dominant classes, trade unions could enter into a sort of ‘clientilist’ relationship with the forces in control of the state. In Mexico, a dictatorship was overthrown in 1910 by a national revolution, and in the following years different factions fought for control of the state. During this period, militancy grew among textile labourers, and radical military commanders adopted the demands of workers. As a result, a new federal constitution incorporated labour rights, which were far-reaching for their time. Parallel to this, a bureaucracy of government and allied unions evolved, symbolized by the president of the national trade union confederation being appointed Minister of Industry, Labour and Commerce. In a combination of dissent against this development, as well as an effort to gain power by supporting rival political leaders, the labour movement became fragmented, and in 1940 the 425 local textile workers’ unions were affiliated to no less than eight different federations. But close alliances continued between the dominant political forces and the dominant trade unions. From the late 1940s, both employment and real wages started to drop in the Mexican textile industry, and government abandoned its pro-labour policies. However, trade unions were still loyal to their political patrons, preventing a rank and file opposition to the declining wages.73 Clientilism was even more prevalent in Argentina. Under the government of Juan Perón from 1946, the Argentinean state pursued a policy of social peace, by giving benefits to the workers, controlling labour conflicts and supporting union organizations. In the same period a new, parallel trade union structure, closely related to the Peronist movement, was created. These favoured unions grew quickly. During the first years of the regime, strikes were common in the textile industry. In the late 1950s, one of the issues of conflict was rationalization. Some militant actions took place, but unions generally accepted rationalization plans as a premise for higher wages.74 Another type of dependence can be observed in nations where textile trade unions emerged when the country was still under colonial rule and an independent nation state did not yet exist. Such was the case in India, where the labour 73 Bortz, ‘Mexico’. 74 Lobato, ‘Argentina’. 765

The Ashgate Companion to the History of Textile Workers movement inevitably became connected with the struggle against colonialism. The dominating union in the textile centre of Bombay, RMMS, was affiliated to the nationalist Congress Party. When India became independent, the nationalist government pursued a policy of state intervention in industry, with the goal of job creation. Labour laws were set up, which established the state as a powerful third party, intervening in disputes on the labour market. The RMMS was, by law, given a monopoly as the sole representative of textile workers.75 In Egypt, the fate of trade unions was even more closely connected to the nationalist regime. Trade unions were allied to the struggle against British colonialism almost from the start. However, only very few textile workers were union members. Between 1919 and 1923 there was a wave of strikes, and a textile workers’ union was established. But after the country’s nominal independence, unions were suppressed, and attempts to organize workers were mostly part of the internal power struggle between different nationalist movements. From 1942 trade unions were made legal and until the establishment of the nationalist military regime in 1952, the textile workers, especially of Cairo, formed the core of a radical labour movement. The nationalist government was widely supported by Egyptian workers. A corporatist structure was created, enacting several laws improving workers’ conditions, and trade unions effectively became part of the state apparatus. While most union leaders accepted the role as clients of the regime, the textile unions, with a strong Marxist influence, sought to maintain an independent status. The regime counteracted this with a combination of repression and social benefits for workers. Consequently, the Textile Workers’ Federation was gradually incorporated into the state-controlled labour movement.76 In both India and Egypt, nationalist governments regarded the textile industry as important, not only economically but also symbolically. In India, textile production was associated with a golden age of self-reliance, before colonialism. In Egypt, textiles was the most industrialized sector of production, thus symbolizing the modernization that the nationalist regime was aiming at.

Unions in non-capitalist states In the examples discussed so far, the mode of production has been that of capitalism – even if in some cases, like Egypt, substantial parts of the textile industry had been nationalized. But in other countries, such as Russia after 1917 and China after 1949, the dominant type of ownership was by the state, which was officially an instrument of the ruling working class. As noted earlier, Russian Bolsheviks prior to 1917 believed that trade unions should be subordinated to the party, acting as the political vanguard of the working class. This position can, at least partially, be explained by the specific circumstances 75 Roy, ‘India’. 76 Beinin, ‘Egypt’. 766

Institutions in textile production of the tsarist regime. Political repression was extremely harsh, and – except for a period following the 1905 Revolution – any labour movement had to be clandestine. These conditions did not favour mass organizations like trade unions. Instead the Russian labour movement became predominantly political. However, trade unions did in fact exist. And it was a strike by female textile workers in Petrograd that sparked the revolution in February 1917, overthrowing the tsar.77 Following these events, trade union membership grew enormously, and continued to do so after the socialist October Revolution and during the civil war. Consequently, trade unions came to be far the biggest mass organizations in the new Soviet state. Especially during the years of wartime communism, trade unions gradually adopted new functions such as the working out of wage tariffs, the distribution of labour power and even recruitment of voluntary soldiers to the Red Army. During these turbulent times, some factories were taken over by workers’ committees, if only to keep them open. One of the unions revived was the Textile Workers’, which under Bolshevik leadership grew from a small local union to a national federation in just a few months during 1917, claiming to represent 178,000 members.78 In the years 1920–1921, when the civil war ended and war communism was about to be abandoned in favour of the more liberal New Economic Policy, discussions evolved inside the Russian Communist Party (RCP)79 on the proper role of trade unions in a socialist state. A ‘workers’ opposition’ wanted to strengthen the role of trade unions and give them more control over the economy, both at the national level and in the factories. On the other hand, a group led by Trotsky, among others, wanted to transform the trade unions into primarily vehicles for the state in promoting productivity. Between these positions was a group that included Lenin. The state, the party and the trade unions, Lenin argued, should be seen as a system of cogwheels, interacting with each other. Unions were important for the party to reach the masses, and as such they had the function of ‘schools of communism’. On the other hand, since the state was not yet a proper workers’ state, but a ‘workers’ state with a bureaucratic twist to it’, the proletariat still needed trade unions. While Lenin was clearly against subordinating trade unions completely to the state, he was not very clear on how their independent role should be secured. He was, however, clear about giving priority to production in the last instance: ‘Industry is indispensable, democracy is not.’80 Lenin’s position finally became the official position of the party. However, as the dictatorship of the proletariat gradually turned into a dictatorship of party officials, trade unions were in reality subordinated to the interests of the party and state bureaucracy, and transformed into vehicles for promoting productivity.

77 Pretty, ‘Russia’. 78 Ibid. 79 Renamed the Communist Party of the Soviet Union in 1925. 80 V.I. Lenin, ‘The Trade Unions, the Present Situation and Trotsky’s Mistakes’, in V.I. Lenin, Collected Works, vol. 32 (Moscow, 1965), pp. 19–42. 767

The Ashgate Companion to the History of Textile Workers During the 1920s, strikes were still not an unusual phenomenon. The forced industrialization during the initial five-year plan led to increased consumer prices, and labour unrest in the textile-producing region of Ivanovo is reported as late as 1932. But from that point strikes seems to have been effectively repressed by the regime.81 The five-year plans and forced industrialization laid emphasis on heavy industry. This was seen as the backbone of socialist development, and consequently it was the male metal worker who became the role model of communist ideology. While textile production was increased in absolute numbers during the planned economy, its relative importance diminished. Textile work became a women’s job with low pay and low status, and consequently moved away from political attention. The second example of a non-capitalist state in this volume is China. After seizing power in 1949, the CCP introduced a number of welfare benefits to workers, both through the state and through the unions, now under CCP leadership. The unions, on the other hand, cooperated with the state in its aim of controlling private employers and weakening the capitalist class. The government established labour– capital consultative conferences, used by workers to achieve a degree of collective control over their factories. However, when industry was finally socialized in 1956, there was a turn in development: when a factory came under state management, workers’ influence generally diminished. Frustration with the weakness of the CCP-led unions contributed to a large wave of strikes in 1957. These were repressed, and under the subsequent political campaigns, such as the Great Leap Forward, it became increasingly difficult for workers to express their interests. The urban textile workers benefited greatly from the Chinese Revolution measured in terms of job security and welfare benefits. And the status of industrial workers – not least the female textile workers – in Chinese society was improved immensely. But parallel to developments in the Soviet Union, unions were transferred into state instruments of economic development and political stability.82 And in both cases, workers’ direct control over production was confined to the experiments of the turbulent period immediately after the revolutions.

Unions and the state – concluding remarks There is an immense wealth of theories on the nature and role of states. There are also several theories that link a typology of trade unions to different types of states or political regimes. Ross M. Martin, for example, suggests that there is a causal relationship, in which certain types of political systems invariably lead to certain types of trade unions.83 I shall confine myself to more general observations. 81 Pretty, ‘Russia’. 82 Cliver, ‘China’. 83 Non-competitive party systems lead to state-ancillary unions, competitive party systems to either autonomous or party-ancillary unions and so on. Martin, Trade Unionism, pp. 235–237. 768

Institutions in textile production While the state might principally be thought of as an instrument for the ruling class in subordinating the other classes, historical reality shows a somewhat more complicated picture. In order to uphold the existing social order, the state must also ensure the basic functions of society. Consequently, in an industrialized society, the state must see to it that industrial production can take place, and preferably increase. The global history of trade unions in textile industries shows a number of ways to achieve this, depending on the relative strength of the various classes and on other historical circumstances – from repression to consent. Furthermore, the ruling class is seldom a monolithic entity. Sometimes it is divided into factions with different interests, which will also leave room for labour to acquire certain concessions. While a typology of trade unions might be a valuable heuristic aid in trying to get a global overview, it might just as easily be limiting our viewpoint. Most notably, typologies of the kind offered by Ross M. Martin and others have a tendency to obscure the very important dimension of historical change. Certainly, the history of industrial relations of specific countries seems to reveal their ‘decisive moment’, which may be formative for union structure and behaviour for decades to follow. An example is the system of collective bargaining, established in a number of European countries in the late nineteenth and early twentieth centuries, the ‘clientilist’ relationship with the state or dominant political party entered into by some unions during periods of national reconstruction or the subordinate relation of unions to the state established in some non-capitalist countries. However, changes can and will happen – and sometimes abruptly. As with any institution, unions should be considered historical entities. Referring to the initial discussion above on the nature of trade unions, these examples from the global textile industry confirm that we should be very careful in pointing to certain models of trade union developments as being prototypical. Textile unions might have been contesting or recognizing the state, they might have been autonomous or subordinated to party or state interests. However, all of them were representatives of parts of the working class, either by active mobilization or at least by passive consent. And they should be analysed in a historical context, in which the possibilities and constraints offered by the different and changing political regimes should be given due consideration. This, of course, does not exclude a critical appraisal of the path chosen by – or forced upon – individual unions.

Recent developments For some decades now, textile production has been declining in many countries, especially in Europe and North America. One of the consequences is that textile trade unions have substantially lost members. In the Netherlands, the textile industry’s relative share of industry has dropped from more than 20 per cent to little more than 2 per cent over the last 50 years, measured in terms of both

769

The Ashgate Companion to the History of Textile Workers economy and employment.84 Another symbolic example is that of the Danish textile workers’ federation, which – after 113 years of existence – was dissolved as an independent federation in 1998 to become just a small part of the larger General Workers’ Union.85 It is not only the production of textiles, but of a great many industrial goods, which has declined in what were formerly known as ‘the industrialized countries’ of the world. This has led to the popular belief that we are entering into so-called ‘post-industrial societies’. However, a glance at the streets of any city in the world will tell us that people do not walk naked, but usually wear industrially fabricated clothes. What has happened on a global scale is not a general transition to post-industrialism, but a new phase in the ongoing restructuring and relocation of investments, production and labour markets in an industrialized world. In a number of countries, the textile industry has benefited from protectionist policies in the post-Second World War period. However, the last two to three decades have also been characterized by a global swing towards liberalization and deregulation. In some cases, this has revived the textile industry, in others it has led to further decrease – but generally it has been a challenge to textile workers. The neo-liberal economic policy pursued by the Argentinean government in the 1990s, made the textile industry, already in decay, almost extinct. During the last few years, new textile companies have been established. But the trade unions, which had already suffered from the repressions during the military regime in the late 1970s, almost disappeared during the crisis of the industry.86 In Egypt, de-nationalization has created a growing private sector at the expense of the publicly owned industries. Neo-liberalism has led to decreasing real wages, provoking outbursts of independent labour activities, especially in the public sector. In the private sector social benefits are few, and workers are often hired on temporary contracts. There are few unions and few strikes. In other words: a significant disparity has evolved between public sector and private sector workers, concerning work experience as well as trade union and political activities.87 The policy followed by the Indian government of protecting the national textile industry has become less successful since around 1980. The industry lost its ability to compete, and the workers’ expectations of job security could no longer be fulfilled. RMMS, the union closely associated with the leading political force, lost its credibility among both workers and employers. As the established structures for negotiation collapsed, the workers became more militant. A major strike in the Bombay textile industry in 1982–1983 was lost by the workers, but gave the mill owners a chance to rationalize and promote an alliance between textile capitalists and the state, eventually leading to a deregulation of the industry. A number of new major textile companies were created, often in semi-rural areas, recruiting

84 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 85 Christensen, ‘Denmark’. 86 Lobato, ‘Argentina’. 87 Beinin, ‘Egypt’. 770

Institutions in textile production labour from an impoverished rural population. National trade unions seldom have a presence there.88 The most striking and far-reaching example, however, is perhaps that of the Chinese textile workers. During the last two decades of the twentieth century, China not only drastically increased its production and export of textiles, but the Chinese economic system also underwent radical transformations. Economic growth was achieved by market reforms and privatization. Many of the benefits won by industrial workers during the Revolution were abandoned. In the rapidly expanding private sector, labour was often recruited among rural migrants. They face long working hours and low pay, and are usually denied even their legal rights. In principle, China may still have progressive labour laws. However, as long as it is denied the right to form independent unions, this new proletariat, emerging in the world’s largest textile-producing nation, has no means of claiming its right from the state or defending itself against intensified exploitation by employers.89 These recent developments are evidence that despite what some historians have claimed, history has not yet come to an end. As long as textile production continues, there will be a need for workers to organize themselves, whether in existing unions or by creating new ones. Thus we can expect new chapters to be added to the global history of textile trade unionism in the future.

88 Roy, ‘India’. 89 Cliver, ‘China’. 771

This page has been left blank intentionally

31 Covering the world: some conclusions to the project Elise van Nederveen Meerkerk, Lex Heerma van Voss and Els Hiemstra-Kuperus

The ultimate goal of a collaborative project like this is of course to sketch some broad lines and draw more wide-ranging conclusions. What, in the end, do the national overviews and thematic essays collected in this volume tell us about the global history of textile workers from 1650 to the present? We do not pretend to be able to construct an all-embracing, exhaustive history of textile production in all its diversity and covering all the textile workers concerned. Comparing just the spinning and weaving of wool and cotton between some twenty important textileproducing countries means eliminating many aspects of textile history beforehand, as we have explained in our introduction. A large-scale international comparison is difficult enough within the limits to which we have confined ourselves. The historical experiences of textile workers taken into account here already show a great deal of variation and diversity across time and space. All too general conclusions would run the danger of becoming a-historical. However, to state that generalization in textile history is utterly impossible, or even ‘dangerous’, would be to discredit any attempt to analyse historical development. The central argument of this conclusion relies on three determinants of the global history of textile workers that we believe to be of crucial importance. First 

 

We thank the participants in the Amsterdam conference that was part of this project and the participants at sessions at the 1st European Global History Conference (Leipzig, 22–24 September 2005), the 30th Social Science History Conference (Portland, 2–6 November 2005) and the 6th European Social Science History Conference (Amsterdam, 22–25 March 2006), and Jan Lucassen and Marcel van der Linden for comments on earlier versions of this concluding chapter. As many of the participants asserted, for example Hunter and Macnaughtan, Chapter 28, this volume, ‘Gender’; McIvor, Chapter 24, this volume, ‘Textile Firm’; Scholliers, Chapter 27, this volume, ‘Working Conditions’. As put forward by David Jenkins, Cambridge History of Western Textiles (Cambridge, 2003), p. 4.

The Ashgate Companion to the History of Textile Workers of all, our global comparison offers many examples clearly confirming earlier research that a teleological view on the history of textile work should be avoided as much as possible. Secondly, in the ongoing up-scaling of markets, we observe a trend in which textile entrepreneurs are constantly searching for ways to lower production costs, most notably by cutting down on the wage bill, here defined as a ‘race to the bottom’. Thirdly, we want to emphasize the observation that, in general, throughout history the bargaining power of textile workers has been weak. We do not claim to be the first to come up with any of these insights. However, we do think that the outcomes of our global comparison more systematically and clearly point to the interrelations between the three mechanisms mentioned here, and the way in which ‘globalization’ influenced them all. In the following, we will further elucidate our arguments.

Synchronicity of the non-synchronous Since all good things come in threes, it is tempting to distinguish between three different organizational configurations of spinning and weaving of cotton and wool that have existed around the globe in the last 350 years. Based on existing literature and the national overviews and comparative essays in this volume, a typology in three crude categories can be constructed, representing an ‘artisan’, a ‘proto-industrial’ and an ‘industrial’ configuration. However, such a model seems to imply that there are a set number of developmental stages through which textile industries are determined to move chronologically, with the phase a country or region is in inherently representing its degree of progress or ‘modernity’. It is therefore questionable whether such a typology is useful, especially since the various essays in this volume show that the global history of textile workers is actually characterized by far more dynamic mechanisms, leading, for instance, to combinations of modes of production and, in any case, definitely not to unidirectional developments. These dynamics can also be typified as the ‘synchronicity of the non-synchronous’, as one of the participants in this project has proposed. One marked example of these dynamics is the historical significance and the continued importance of non-factory production. The first theorists of protoindustry often considered it as a transitional phase between pre-industrial and industrial production, either because of its specific demographic characteristics or because it laid the foundation for a capitalist system. In areas where ‘protoindustrialization’ occurred, mechanization and the transition to a factory system would eventually follow. More recent research has already indicated that this logic  See for an elaborate critique of this conception of industrialization, Quataert, Chapter 23, this volume, ‘Proto-industrialization and industrialization’.  Komlosy, Chapter 25, this volume, ‘Spatial division’.  For example, Franklin Mendels, ‘Proto-industrialization: The First Phase of the 774

Covering the world did not necessarily apply to the whole of Europe. Not in every ‘proto-industrial’ region did factory production follow, and de-industrialization could also take place. In many cases, home manufacturing continued to exist alongside factory production, sometimes independently, and sometimes as supplementary labour in commission of the factories. Moreover, several of the national overviews included here show that factory production did not inevitably prevail over manufacturing. As Prasannan Parthasarathi has already shown elsewhere, non-western hand-weaving held its own against mechanized weaving in the West for a longer time than previously thought. This was not necessarily because wages were lower in the East, but because productivity was high. The national overviews on India, the Ottoman Empire, Egypt and Uruguay stress the continued viability of (hand) production at home.10 In the Ottoman Empire, by 1900, hand-spinning still produced one quarter of all yarn consumed, and spinning and weaving took place in practically every home. In North China a putting-out system was developed as recently as the early twentieth century, profiting from small, inexpensive and fast looms. Smaller workshops and workers at home could adapt in a flexible way to market changes, or to put it differently, merchants could avert the risk of the markets to their workforce.11 In India, the relative importance of hand-weaving compared to mechanical looms was greater in 1995 than it had been in 1940.12 It is altogether plausible that many textile workers working at home as well as their production escaped the attention of the statisticians, especially if the production was intended for the domestic market. In the ‘developed’ Western world too, until the present day, hand-manufacturing, albeit usually hidden and under abysmal (but hence competitive) conditions, continues to be important. In these cases, it mostly concerns illegal production in Industrialization Process’, Journal of Economic History 32 (1972): 241–261; David Levine, ‘The Demographic Implications of Rural Industrialization: A Family Reconstitution Study of Shepshed, Leicestershire, 1600–1851’, Social History 2 (1976): 177–196; Peter Kriedte, Hans Medick and Jürgen Schlumbohm, Industrialization before Industrialization: Rural Industry in the Genesis of Capitalism (Cambridge, 1981).  This last issue was noted by Kriedte, Medick and Schlumbohm, Industrialization before Industrialization, p. 145. See also Sheilagh C. Ogilvie and Markus Cerman, ‘The Theories of Proto-industrialization’, in Sheilagh C. Ogilvie and Markus Cerman, European Protoindustrialization (Cambridge, 1996), pp. 1–11, at pp. 10–11.  See, for example, Komlosy, Chapter 3, this volume, ‘Habsburg Monarchy’; Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, Chapter 14, this volume, ‘Netherlands’.  Prasannan Parthasarathi, ‘Rethinking Wages and Competitiveness in the Eighteenth Century: Britain and South India’, Past and Present 158 (1998): 79–109. 10 Roy, Chapter 10, this volume, ‘India’; Quataert, Chapter 18, this volume, ‘Ottoman Empire’; Beinin, Chapter 7, this volume, ‘Egypt’; Camou and Maubrigades, Chapter 20, this volume, ‘Uruguay’. 11 Cliver, Chapter 5, this volume, ‘China’. 12 Tirthankar Roy, ‘Indian Handlooms in the 20th Century’, Jahrbuch für Wirtschaftsgeschichte 2 (1998): 129–150, at p. 149. 775

The Ashgate Companion to the History of Textile Workers the clothing industry, for instance the sweatshops that are present in many large cities, like Paris, Manchester and Amsterdam. The work here is usually done by illegal immigrants, working for minimal wages and under very poor conditions. In the Netherlands in the late 1990s, for instance, there were an estimated 15,000 workers in clandestine jobs in the clothing sector; a number that actually exceeded the officially registered workers in the sector. These malpractices are by no means restricted to Europe, as the many examples of unauthorized workshops in the United States show. Here too, the work is done by thousands of illegal immigrants, who pay enormous sums of money to traffickers for bringing them to California, where they then perform forced labour in horrifying circumstances.13 Keeping all this in mind, the use of the concept ‘proto-industrialization’ is probably too teleological and will therefore have to be reconsidered, as was put forward by Don Quataert, who suggested replacing it with ‘putting-out’.14 However, this term may be too specific and narrow to describe the hand-manufacturing of textiles for (supra-)regional markets. Instead, it is proposed here to introduce the concept of ‘industrious’ production, since the productivity of labour in this system was pushed up by a different way of organizing textile production (for example through the putting-out system) rather than by mechanization. This industriousness often implied a more efficient use of all members of the household – notably women and children – within textile production, a process that has also been described by proponents of proto-industrialism.15 The term ‘industrious production’ is quite intentionally analogous with Jan de Vries’ concept of the ‘industrious revolution’, with which he referred to changing consumer patterns in pre-industrial societies. De Vries claimed that a higher purchasing power of workers’ families and, consequently, the rising demand in consumption commodities was made possible by the intensification of labour. Workers reached this goal mainly by re-allocating their labour time at the cost of leisure, and by employing more members of the family.16 The same dynamics that make us question (proto-)industrialization occur in other areas of textile history, which makes categorization even more difficult. For instance, in many of the textile histories a widespread presence and enduring importance of guilds and guild traditions is visible, not necessarily restricted to artisan production. Guilds are not mentioned in the national overviews of the American countries, except Mexico, and they also did not exist in Australia. But they were present in all countries of Eurasia that we have overviews for.17 Their late 13 Examples taken from ILO, Labour Practices in the Footwear, Leather, Textiles and Clothing Industries (Geneva, 2000), pp. 76–82. 14 Quataert, ‘Proto-industrialization and industrialization’. 15 Kriedte, Medick and Schlumbohm, Industrialization before Industrialization, pp. 61–62. 16 Jan de Vries, ‘Between Purchasing Power and the World of Goods: Understanding the Household Economy in Early Modern Europe’, in John Brewer and Roy Porter, Consumption and the World of Goods (London, 1993), pp. 107–121, at p. 110; Jan de Vries, The Industrious Revolution, Consumer Behavior and the Household Economy, 1650 to the Present (Cambridge, 2008). 17 Bortz, Chapter 13, this volume, ‘Mexico’; Cliver, ‘China’. Hunter and Macnaughtan, 776

Covering the world survival dates (for instance until 1862 in Denmark and into the twentieth century in China) alone suggest a possible continuity between the old guilds and newer forms of workers’ organizations, such as trade unions. Another example of the synchronicity of the non-synchronous is the ever-shifting constructions of gendered work patterns through space and time.18 Although the textile industry was (and still is) remarkable in the sense that it simultaneously employed women as well as men in most of history and throughout the globe, gender specific work roles were nowhere ultimately fixed. By no means was there ever a direct link between the gendering of work and technology, nor did ‘culture’ or ‘institutions’ determine stable gendered work relations.19 If synchronicity of the non-synchronous is so important, does this mean that everything happens everywhere all the time and that there are no continuous or long-term trends in textile history? On the contrary, we think that the global perspective changes our view of cotton workers’ history and that there are trends to be found that transcend the universal historians’ conclusion that we have reached a deeper insight in the differences in historical experiences from place to place and from period to period.

A ‘race to the bottom’ Although a comparison of labour costs might have its shortcomings, it nevertheless provides useful information on the reasons for the concentration of textile … production and employment in certain regions of the world, and for the ongoing shifts in the localization of factories in the framework of global production strategies.20 A central long-term trend that is discernable in the history of textile work from the seventeenth century onwards is a continuous attempt to lower the wage bill. Of course, production costs can be cut in numerous ways, for example by lowering raw material costs or price-fixing. But in textiles, the strategy was mainly reducing the share of labour costs in total production costs. Why would lowering the wage bill be a particular trend in the production of textile goods? There is a clear connection with certain characteristics of textile production and textile products. Compared to other essential goods, such as housing and food, clothing and its raw materials proved to be very movable Chapter 12, this volume, ‘Japan’; Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’; Quataert, ‘Ottoman Empire’. 18 For example, Hunter and Macnaughtan, ‘Gender’; Christensen, Chapter 6, this volume, ‘Denmark’; Quataert, ‘Ottoman Empire’. 19 See for a good overview of these arguments, Sheilagh C. Ogilvie, A Bitter Living: Women, Markets, and Social Capital in Early Modern Germany (Cambridge, 2003) pp. 7–15. 20 ILO, Labour Practices, p. 37. 777

The Ashgate Companion to the History of Textile Workers goods early on. Most basic foodstuffs are perishable, once processed, and at least require special transport conditions. Houses and other products of construction in most societies have been quite immobile goods, which made transport of these goods highly inefficient. As a consequence, perishable foodstuffs until the coming of refrigerated trains and ships in the mid-nineteenth century used to be mainly sold locally and housing still is so.21 Textiles were relatively easy to transport and became more influenced by changing consumer demands than food or houses.22 Over time, in supra-regional markets, competition becomes increasingly important. In the absence of trade restrictions, prices on the world market for tradable goods such as textiles will converge, whereas prices of non-tradable goods such as construction and bread will depend on domestic markets.23 While it is hard to compete with non-tradable goods, competition with tradable goods such as textiles will be all the more fierce. In countries with a high national income per capita, prices of non-tradables will be relatively high, driving up nominal wages. Since wages have the tendency to converge within countries, wages in the industries producing tradable goods will rise as well. Therefore, in countries where wages are high, productivity will have to be higher in order to compensate for the loss of income caused by the wage level.24 In order to raise profits, production costs constantly need to be lowered, and since labour costs in textile production formed a substantial part of total costs, and a part that the employers could often influence better than those of raw materials or transport, there were constant attempts to lower the wage bill. There are of course many ways to achieve a lower wage bill. One is the rationalization or mechanization of production, by which employers try to increase the productivity of labour. Rationalization does not necessarily equal mechanization: 21 Kevin H. O’Rourke and Jeffrey G. Williamson, Globalization and History: The Evolution of a Nineteenth-Century Atlantic Economy (Cambridge, Mass. and London, 1999) pp. 33–36. 22 Of course, changing tastes in foodstuffs and building styles also influenced demand in food and construction, but they were and are relatively less important than changing style in clothing. It is no coincidence that the phenomenon of fashion is first and foremost associated with apparel. For the importance of fashion, see De Vries, Industrious Revolution, pp. 140–145; Werner Sombart, Luxus und Kapitalismus (Munich and Leipzig, 1913); Chandra Mukerji, From Graven Images: Patters of Modern Materialism (New York, 1983); John Brewer and Roy Porter (eds), Consumption and the World of Goods (London and New York, 1993); Daniel Miller (ed.), Acknowledging Consumption: A Review of New Studies (London and New York, 1995); Diana Crane, Fashion and its Social Agendas: Class, Gender, and Identity in Clothing (Chicago and London, 2000); Peter N. Stearns, Consumerism in World History: The Global Transformation of Desire (London and New York, 2001). 23 Bread was a non-tradable in most of history, and at least until well into the nineteenth century. Jan Luiten van Zanden, Arbeid tijdens het handelskapitalisme. Opkomst en neergang van moderne economische groei (Bergen, 1991), p. 146. Of course, bread grains like wheat and rye and other grains like rice or corn are tradable. 24 Bela Balassa, ‘The Purchasing-power Parity Doctrine: A Reappraisal’, Journal of Political History 72 (1964): 584–596, at pp. 586, 593; Van Zanden, Arbeid, p. 145. 778

Covering the world it can also mean raising an industry’s efficiency by more efficient organization, for instance by introducing a putting-out system or concentrating formerly scattered workers under one roof in workshops or manufactories.25 However, historically mechanization certainly has been an effective means of rationalizing production. In cotton spinning, for instance, labour productivity (the amount of fibres spun per hour) increased spectacularly after the introduction of Hargreaves’ spinning jenny in 1767: by the 1960s it had multiplied by 500 times with the then common type of automatic ring spinning machines. In the same period, productivity in cotton weaving had increased 220–390 times.26 In combination with import restrictions, the technological advances in the eighteenth century finally enabled Britain to compete with high-quality, but cheap, Indian cottons. Given stiff international competition in textiles, other regions soon reacted, either by import substitutions or by copying British technology.27 It is remarkable how many of the overviews mention that their national industry turned towards Britain for technological guidance, at least until the interwar period.28 It was not only countries in ‘the West’ that responded by mechanizing their textile industries in the nineteenth century: India, Japan and China also industrialized from the 1860s onwards.29 In most cases, workers in branches that were mechanized experienced a considerable loss of income and employment, although sometimes new employment opportunities appeared in other, as yet non-mechanized textile branches.30 Another means of lowering the wage bill is actually reducing nominal wages in the home industry. This was possible when supply of labour exceeded demand, as was the case in the Yorkshire worsted industry in the 1820s, and in nineteenthcentury China, where household units produced cotton for wages below

25 Maxine Berg, The Age of Manufactures 1700–1820: Industry, Innovation and Work in Britain (London, 1985), pp. 220–233. 26 Almut Bohnsack, Spinnen und Weben (Bramsche, 2002), p. 152. 27 Beverly J. Silver, Forces of Labor: Worker’s Movements and Globalization since 1870 (Cambridge, 2003) pp. 85–86; Komlosy, ‘Habsburg Monarchy’. 28 Komlosy, ‘Habsburg Monarchy’, for instance, mentions the adoption of spinning machines ‘nach britischer Art’. British spinners and weavers migrated to Brazil and the USA, carrying British technology with them. For a broader perspective on this, see David J. Jeremy, ‘The International Diffusion of Cotton Manufacturing Technology, 1750–1990s’, in Douglas A. Farnie and David J. Jeremy (eds), The Fibre that Changed the World: The Cotton Industry in International Perspective, 1600–1990s (Oxford, 2004), pp. 85–127. 29 Roy, ‘India’; Hunter and Macnaughtan, ‘Japan’; Cliver, ‘China’; Silver, Forces of Labor, pp. 88–89. 30 This was, for instance, the case in the late eighteenth, early nineteenth century, when spinning was mechanized, while weaving was not. This initially created a larger demand for handloom weavers. See, for example, Fowler, Chapter 9, this volume, ‘Great Britain’; G. Gullickson, Spinners and Weavers of Auffay: Rural Industry and the Sexual Division of Labour in a French Village, 1750–1850 (Cambridge, 1986), p. 105. 779

The Ashgate Companion to the History of Textile Workers subsistence level.31 Another mechanism underlying the reduction of wages was the rationalization of production. Hand-weavers in Austria, for instance, experienced a fall in wages when they faced competition from mechanized weaving factories.32 Also within the factory system, the introduction of new machinery could lead to wage losses. The multiple loom system in the Danish textile industry would not have been profitable if employers had not decided to cut piece-rates for every extra loom that was attended by one worker.33 As a matter of fact, the very use of piece-rates (as opposed to time wages) provided employers with a tool to control production and raise productivity, eventually leading to cutbacks in the piece-rates themselves.34 Often, this was accentuated by gender differences, as in the Chinese textile industry in the 1930s, where men were working on time wages, but women usually received piece-rates for their work.35 The use of female and child labour could also be an adequate way for employers to lower the wage bill. There are a number of reasons why women in the textile industry were historically paid less than men. An argument often mentioned is that the physical strength of women, and therefore their productivity, was lower than that of men.36 Furthermore, their reproductive tasks in most of history bound them to the domestic sphere and to chores that were easy to combine with household work and child- rearing. However, empirical research has refuted these arguments.37 More likely, gender was used as an instrument to keep wages low in certain labourintensive textile occupations and to protect (white) male employment in better paid jobs.38 Already in the Middle Ages most women were excluded from vocational training and from guild-regulated jobs, which made it difficult for them to acquire the necessary skills for well-paid jobs.39 This institutionally and culturally defined division of labour implied that the number of women available for low-skilled jobs, notably hand-spinning, was high. The large supply of female labourers kept spinning wage rates low.40 Sometimes institutions like guilds or the state even set 31 Fowler, ‘Great Britain’; Cliver, ‘China’. 32 Komlosy, ‘Habsburg Monarchy’. 33 Christensen, ‘Denmark’. 34 Harry Braverman, Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century, 2nd edn (New York, 1998), pp. 68–69. 35 Cliver, ‘China’. 36 J. Burnette, ‘An Investigation of the Female–Male Wage Gap during the Industrial Revolution in Britain’, Economic History Review 50 (1997): 257–281. 37 See, for example, Deborah Simonton, A History of European Women’s Work: 1700 to the Present (London, 1998), esp. pp. 70–83; Ogilvie, Bitter Living, pp. 7–8. 38 Hunter and Macnaughtan, ‘Gender’, p. 10. 39 Judith C. Brown, ‘A Woman’s Place was in the Home: Women’s Work in Renaissance Tuscany’, in M.W. Ferguson, M. Quilligan and N.J. Vickers (eds), Rewriting the Renaissance: The Discourses of Sexual Difference in Early Modern Europe (Chicago and London, 1986), pp. 206–224, at p. 214; Sheilagh C. Ogilvie, ‘How Does Social Capital Affect Women? Guilds and Communities in Early Modern Germany’, American Historical Review 109 (2) (2004): 334–336. 40 Brown, ‘A Woman’s Place’, p. 218. 780

Covering the world maximum wages for female spinners specifically to protect the wage level of (male) weavers.41 Skill is at least partly a social construction, and it is obvious that throughout history skill was closely related to the construction of gendered work and low payment for women. Female workers were paid less because they were supposed to be low-skilled, but at the same time, work could be considered low-skilled and thus low-paid simply because it was performed by women. Or, to put it differently, status attaches to the worker, not to the work.42 In the national overviews, there is abundant evidence of the ‘gendering’ of skill.43 This could also imply a devaluation of work, when women started to perform tasks in the textile industry previously carried out by men.44 The other way around, status or wages might increase when men took up formerly ‘female’ jobs.45 Gender, skill, status and wage levels were to a large extent intertwined. The work of women and children could also be paid less because it was perceived as supplementary to the family income.46 Nevertheless, historical reality was often different, for instance when there was no male head of household. Employers of course did not take this into account when they set their wage levels. The low payment of textile work performed within the household could lead to wages below subsistence level. In many cases, this was only possible because families continued to farm a plot of land and managed to stay alive by the returns from their agrarian activities, as in eighteenth-century China or early twentieth-century Egypt.47 This leads us to yet another strategy in the employers’ quest for lower wages: the relocation of the industry to areas where nominal wages were lower. Such a ‘migration of the industry’ (rather than of the workers) was not necessarily disadvantageous to the new groups of workers drawn into the textile industry, which might actually experience a real wage increase.48 But competition in the labour market or the lack of job opportunities outside the textile industry could make even these new workers take on work at below subsistence wages, especially when their wages from textile production were supplemented by income or food acquired from agricultural activities. A historical relocation of the textile industry is visible on three different levels. The first level was a regional shift, in which parts of the textile industry, usually 41 Ogilvie, ‘How Does Social Capital Affect Women?’, p. 342. See also Blewett, Chapter 21, this volume, ‘USA’. 42 Simonton, A History, p. 83, Evelyne Sullerot, Histoire et sociologie du travail féminin (Paris, 1968). 43 Cliver, ‘China’; Hunter and Macnaughtan, ‘Japan’; Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’; Pretty, Chapter 16, this volume, ‘Russia’; Roy, ‘India’. 44 For example, Beinin, ‘Egypt’; Christensen, ‘Denmark’. 45 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’; Smith et al., Chapter 17, this volume, ‘Spain’. 46 Hunter and Macnaughtan, ‘Gender’. 47 Cliver, ‘China’; Beinin, ‘Egypt’. 48 See, for example, Parthasarathi, Chapter 22, this volume, ‘Global Trade’. 781

The Ashgate Companion to the History of Textile Workers spinning and sometimes also weaving, were moved from an urban location to relatively nearby rural areas. This development was already apparent in textile centres in north-western Europe before 1650, when this book begins to cover developments.49 An early example treated here is the shift of the woollen industry from the cities of Holland to rural provinces in the Dutch Republic in the seventeenth and eighteenth centuries.50 Similar developments took place in other parts of the world, in areas as far apart as Japan and the Ottoman Empire.51 In Egypt, the ruralization of textile production in the nineteenth and twentieth centuries also put a downward pressure on urban wages.52 The second level on which a shifting geographical division of labour can be discerned was the interregional level. One early example is the contracting out of linen and cotton spinning from Holland to German territories and Silesia in the seventeenth and eighteenth centuries.53 In the USA, a similar development took place in the post-Civil War period, when the Northeast textile industry slowly but surely lost ground to the southern textile mills.54 More recently, shifts on this level occurred when the markets in East European countries, such as the Czech Republic, started to open up and textile firms proved to be competitive with West European industries.55 The third level of displacement of the textile industry is the global level. Globalization of textile production actually has a long history. As we have seen, India was the most important cotton producer on the world market before British industrialization took off in the late eighteenth century.56 In the late nineteenth century, it was recognized by both foreign and domestic investors that countries like India, China and Japan could compete with Western textile industries precisely because of their cheap labour.57 Clark has shown how productivity and low wages interplay in this relocation process. He calculated manufacturing costs in cotton weaving for three high-wage areas, three intermediate countries and three low-wage countries in 1910. His results are summarized in Table 31.1.

49 Herman van der Wee, ‘The Western European Woollen Industries, 1500–1750’, in David Jenkins (ed.), The Cambridge History of Western Textiles (Cambridge, 2003), pp. 397–472, at pp. 398–399. 50 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 51 Hunter and Macnaughtan, ‘Japan’; Quataert, ‘Ottoman Empire’. 52 Beinin, ‘Egypt’. 53 J.A. Boot, ‘Silezische en andere linnengarens’, Textielhistorische Bijdragen 20 (1979): 3–36 and 22 (1981): 41–79. 54 Blewett, ‘USA’. 55 Komlosy, ‘Habsburg Empire’. 56 Parthasarathi, ‘Global Trade’; Silver, Forces of Labor, p. 86. 57 Silver, Forces of Labor, pp. 86–89; Parthasarathi, ‘Global Trade’. 782

Covering the world Table 31.1

Comparative costs in cotton weaving, nine countries, c.1910 Weekly wage rate (US  $)

Machinery per worker (loomequivalent)

Labour costs, corrected for labour productivity (US  $)

Manufacturing costs (England = 100)

New England

8.80

2.97

6.04

125

USA South

6.50

2.65

5.00

115

England

5.00

2.04

5.00

100

Spain

2.70

0.91

6.05

132

Mexico

2.60

1.15

4.61

119

Italy

2.40

0.88

5.56

120

Japan

0.80

0.53

3.08

101

India

0.78

0.50

3.18

91

China

0.54

0.48

2.30

75

Source: G. Clark, ‘Why isn’t the Whole World Developed? Lessons from the Cotton Mills’, Journal of Economic History 47 (1987): 141–173, here cited after Singleton, World Textile Industry, p. 27.

As is clear from the table, the low-wage countries could compete with English production costs. Their lower labour productivity was more than offset by low wages. The countries in the intermediate group could not. Both regions of the USA only partly offset their high wages through higher productivity. The relocation process accelerated after the Second World War, when de-colonization changed the international market situation and stimulated national industrialization programmes in many non-western countries. Simultaneously, the development of the welfare state drove up wages in many Western countries and, as a consequence, the textile industry was one of the first to be relocated to low-wage parts of the world. In the 1950s through 1970s, many Western countries such as Great Britain, Germany, Denmark and the Netherlands experienced a steep decline in their textile industries.58 From the 1980s onwards, the last bulwarks of past industrial greatness, the USA and Japan, collapsed under the competition of low-wage areas, particularly China and the Indian subcontinent.59 But non-western countries such as Turkey also lost their share of the world market. This is no wonder, since average nominal wage levels show remarkable differences. In 1998, hourly wages in the textile industry of India, China and Indonesia (averages 58 Fowler, ‘Great Britain’; Christensen, ‘Denmark’; Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 59 Parthasarathi, ‘Global Trade’. 783

The Ashgate Companion to the History of Textile Workers of US  $0.25–0.50 per hour) were only a fraction of wages elsewhere in the world. In Turkey, this average was already remarkably higher (US  $2.50), not to mention hourly wage levels in the UK or US textile industries (US  $13–15) and especially countries like Japan, the Netherlands, Denmark and Austria (over US  $20).60 No matter how it was achieved, there is ample evidence for a historical trend in – usually successful – attempts to reduce labour costs as a percentage of total production costs. A rough calculation of the developments in the Dutch–Belgian textile industry may serve as an empirical example for this. In the Dutch city of Leiden in the seventeenth century, labour costs constituted almost 40 per cent of the total production costs of a piece of woollen cloth.61 Not without reason Leiden textile merchants decided to contract out the spinning and weaving of wool to the countryside of Tilburg, where wages formed only 27 per cent of the total costs of one piece of cloth in the second half of the seventeenth century.62 Around 1800, wage costs in the Belgian cotton industry formed 40–45 per cent of the total production costs. Developments at the Voortman firm, one of the largest cotton-producing companies in the city of Ghent until the 1870s, show a continuous decline of the share of wage costs in the nineteenth century, to less than 20 per cent in 1900.63 In the 1920s, the percentages were even lower. At that time, in the Dutch cotton industry, labour costs formed about 10 per cent of the total production costs in spinning factories, 13 per cent in combined (spinning and weaving) factories and 18 per cent in weaving factories. In the wool industry, labour costs were somewhat higher: 18 per cent and 24 per cent for spinning and weaving respectively.64 With the development of labour legislation and the social welfare state, however, wage costs in most ‘Western’ countries started to rise in the decades following the Second World War. Around 2000, the average labour costs in EU textile production had risen again to 40 per cent of total manufacturing costs.65 These percentages were much lower in Turkey (17 per cent in 1998)66 and China (c.10 per cent in 1997). India 60 ILO, Labour Practices, p. 39. 61 N.W. Posthumus, Bronnen tot de geschiedenis van de Leidsche textielnijverheid. Vijfde deel, 1651–1702 (’s-Gravenhage, 1918), no. 434, 13 March 1663. 62 The total price of one piece of cloth was 90 guilders, and total weavers’ wages comprised 24 guilders (27 per cent). In these calculations, it was assumed that weavers themselves paid the spinners out of their own wages. Gerard van Gurp, Brabantse stoffenop de wereldmarkt. Proto-industrialisering in de Meierij van ’s-Hertogenbosch 1620–1820 (Tilburg, 2004), p. 150. See also Van Nederveen Meerkerk, Heerma van Voss and HiemstraKuperus, ‘Netherlands’. 63 Peter Scholliers, Wages, Manufacturers and Workers in the Nineteenth-Century Factory: The Voortman Cotton Mill in Ghent (Oxford, WA, 1996), pp. 63, 220–223. 64 Lex Heerma van Voss, De doodsklok voor den goeden ouden tijd. De achturendag in de jaren twintig (Amsterdam, 1996), p. 196. Note that the figures for the wool industry are only based on single observations and therefore are not necessarily representative of the entire industry. 65 ILO, Labour Practices, p. 38. It has to be mentioned, though, that the total costs are calculated excluding raw material costs. 66 Baris Tan, ‘Overview of the Turkish Textile and Apparel Industry’, Internet publication: 784

Covering the world and Indonesia represented the lowest percentage labour costs in textile production: under 5 per cent in 1997.67 Where does this race to the bottom end? In sub-Saharan countries, where average nominal wage levels are even consistently lower than in Asia?68 New shifts are to be expected, certainly since the highest relative increases in hourly labour costs over the 1980–1998 period were registered in Asian countries. In 1998, labour costs in Asia were almost four times as high as in 1980. On the other hand, in spite of the relatively high labour costs in many Western countries, some of their specialized products continue to be competitive in world markets. This is a result of permanent efforts to rationalize and modernize production processes, including the application of the newest technologies and more efficient marketing techniques. As the production of cheap bulk goods moves to lower wage areas, textile centres which have become uncompetitive in this field still have the option of specializing in or developing higher quality products, for which there is an increasing demand.69 Historically, luxury textile trades have at least been slower to move to lower wage areas.

The importance of power relations The preceding argumentation has mainly offered economic explanations for trends in the global history of textile workers. Nevertheless it is important to realize that at the basis of the ‘search for cheap labour’, as proposed above, lay indisputable differences in balances of power. Unequal power relations in the textile industry were constituted in many ways: master versus journeyman, merchant versus putting-out worker, capitalist versus factory worker, male versus female, adult versus child, colonizer versus colonized, rich countries versus poor countries. Surely, many of the patterns analysed here were also direct or indirect attempts by the powerful to expand their control over the workforce. Changes in the organization of production were not always just motivated by increases in efficiency or to achieve lower wage costs, but also aimed to enlarge control over the workers.70 Likewise, the preference of many employers for piece-rates over time wages can be seen as a means to increase control over the workers.71 The same interpretation could be applied to an increased mobilization of female workers. It is widely alleged that women formed a more docile and easily http://www.hctar.org/pdfs/GS06.pdf (2005), p. 41. 67 International Textile Manufacturers Federation, International Production Cost Comparison (Geneva, 1997), p. 19. 68 Average wages in textiles: Europe: US  $9.25; Americas: $5.86; Asia: $3.66; Africa: $1.05. ILO, Labour Practices, p. 45. 69 ILO, Labour Practices, p. 40. 70 For example, Berg, Age of Manufactures, pp. 189–190. 71 Braverman, Labor and Monopoly Capital. 785

The Ashgate Companion to the History of Textile Workers manageable workforce than men. Employing women would thus not only be a way of attracting cheap labour, but also serve as a means to increase control over the workforce. Abundant empirical evidence shows that this was often the rhetoric of employers, although it is questionable whether their assumptions were correct. Moreover, it is not always clear whether they believed in their own logic or if they just used it as an excuse to employ cheap labour.72 Inequality on a global scale has been very important throughout our period of research. For instance, colonial relations were essential for the take-off of industrialization in the Western world. Parthasarathi shows that without Indian consumption, British textile manufacturing would have collapsed in the nineteenth century. In the process, of course, India’s once flourishing domestic textile industry was temporarily wiped out by British competition.73 But in post-colonial relations as well, the gains for individual firms and workers in non-western countries are not always obvious.74 Although the relocation of the industry after the Second World War favoured employment in these areas, wages were low, working conditions often poor and profits not seldom drained away to foreign (mostly Western) investors.75 Furthermore, there was always the risk of relocation of the industry to regions where wages were even lower. All these imbalances of power seem to lie at the basis of the processes described above. For a great many, they also explain why the structural bargaining power of textile workers has traditionally been weak. In most of history, the textile workers’ trade unions were fighting an uphill battle. The social distance between workers and factory owners was large. The importance of cotton prices and the business cycle in cotton manufacturing brought the employers periodic spells in which they were well stocked but markets were slow. In these circumstances they were quite willing to fight for lower wages, as a strike or lockout only saved them wages. Typically, lockouts were more pervasive in textile mills than in most other industries. The successes that the trade unions could claim often took the form of preventing, delaying or softening turns for the worse.76 In the countries this project has gathered information on, textile workers’ trade unions were clearly successful for a longer period of time only in England, at least in the Lancashire cotton industry, and perhaps also in Denmark.77 The Lancashire trade unions of cotton operatives knew a period of successes and strength from the 1870s, which also translated into high wages for their members. Following Silver’s analysis of global strike patterns, we can explain this by the dominant position of

72 See, for example, Blewett, ‘USA’; Christensen, ‘Denmark’; Cliver, ‘China’; Hunter and Macnaughtan, ‘Japan’; Pretty, ‘Russia’. 73 Parthasarathi, ‘Global Trade’. 74 Roy, ‘India’. 75 As was, for instance, the case in Mexico (Bortz, ‘Mexico’). 76 Clete Daniel, Culture of Misfortune: An Interpretive History of Textile Unionism in the United States (Ithaca, 2001). 77 Fowler, ‘Great Britain’; Christensen, ‘Denmark’. 786

Covering the world Britain in the development of the global textile industry, especially in cotton.78 This enabled the Amalgamated Association of Operative Cotton Spinners to become one of Britain’s strongest unions. British cotton spinners, who long continued working mules instead of ring spinning machines, were more skilled than, for instance, American spinners working on ring machines.79 The stable rate list established in England by employers and the union gave both parties an incentive to increase production. This was, however, the exception. Usually, when the labour movement gathered force, employers effectively threatened to give jobs to other categories of workers (females, children, immigrants)80 or to relocate the industry.81 This last option was probably easier to effectuate in the textile industry, because barriers to entering the industry were relatively low, and there was a widespread tradition of textile production all over the world.82 The threat of relocation of the textile industry explains why recently, even in countries with a trade union tradition, workers who remain in the industry are reluctant to join unions, preferring direct negotiations with their employers which, they believe, are less risky. This trend weakens the unions even more.83 However, even if workers’ organizations in the textile industry have been weak in the long run, they have of course been very important to large numbers of textile workers and their families as a brake on the race to the bottom. They thus created periods of respite, when wages and working conditions improved or at least remained at a constant level for some time.

Labour militancy and globalization In her study of globalization and workers’ movements Beverly Silver presents an overview of global labour unrest in the textile industry between the 1870s and the 1990s, which is reproduced here as Table 31.2. The table is based on data gathered from the New York Times and The Times of London about labour unrest worldwide. The countries mentioned are those that were responsible for at least 1 per cent of labour unrest in the world textile industry, as registered by these two newspapers. The decade or decades in which each country experienced high points of labour unrest are marked.

78 Silver, Forces of Labor, pp. 81–97. 79 Gary Saxonhouse and Gavin Wright, ‘Technological Evolution in Cotton Spinning, 1878–1933’, in Farnie and Jeremy (eds), The Fibre that Changed the World, pp. 129–152. 80 Van Nederveen Meerkerk, Heerma van Voss and Hiemstra-Kuperus, ‘Netherlands’. 81 These practices still continue, much to the disadvantage of working conditions of textile workers, who are harassed into working under deteriorating conditions. ILO, Labour Practices, p. 72. 82 Silver, Forces of Labor, p. 85. 83 ILO, Labour Practices, p. 114. 787

The Ashgate Companion to the History of Textile Workers Table 31.2 Relative high points of labour unrest in the world textile industry, 1870–1996 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s UK Russia

X

X X

US

X

X

Spain

X

X

Poland

X

China

X

Germany

X

Australia

X

India

X

France

X

Belgium

X

Canada

X

Mexico

X

X

X

Egypt

X

Pakistan

X

Note: Data collected in the World Labor Group Database. High points defined as the peak year for that country or as years in which the number of registrations of labour unrest is above 20 per cent of the total mentions for that country. For technical reasons, for the UK the two highest years are marked. Source: Silver, Forces of Labor, p. 82

Broadly speaking, we can see the race to the bottom at work: over time high points of labour unrest tend to proceed from countries with relatively high wages, to countries with a relatively low-wage standard. On the whole it is of course to be expected that textile labour unrest is only to be found where substantial numbers of textile workers are actually present. Furthermore, the table marks high points of textile workers’ militancy in the 1900s in Russia and 1950s in Egypt. In Russia, the strike movement of 1905, interwoven with the Revolution of that year, was indeed at its height. The national overview on Egypt confirms militancy in the 1950s, but traces its origins back to the end of the 1940s. This points to a trend also discussed in the national overviews

788

Covering the world for other countries.84 In emerging industrial countries, textile workers often formed an important section of the urban proletariat. In Egypt in the 1950s textile workers formed more than 40 per cent of the industrial workforce.85 Under these circumstances textile workers were understandably in the vanguard of Nationalist and Socialist/ Communist revolutions when these took place: the Russian Revolutions of 1905 and 1917, the Young Turk movement of 1908, the Mexican Revolution of 1910, the Egyptian Revolutions of 1919 and 1952, the Indian anti-colonial movement which led to Independence in 1947, the Chinese Revolution of 1949. These revolutionary movements were all intertwined with textile workers’ militancy and strikes.86 Some of these revolutions ended in dictatorships, and other dictatorships were established by military coups. Whether taking place in the name of the proletariat, or directed against growing workers’ influence, these coups always curtailed the real power of trade unions and workers themselves to represent workers. The increased risk involved in labour militancy then led to a change in repertoire and an overall decrease in labour unrest. Whether in Russia in the Soviet period, in Germany in 1933–1945, in Poland and Czechoslovakia after the Iron Curtain descended or in China after 1949, in Uruguay in 1973–1984, in Argentina in 1976–1983 or in Spain in 1924–1931 and 1939–1975, the establishment of dictatorships sharply diminished the opportunities for trade unions to organize strikes and other forms of labour militancy. As there is a strong overlap between the countries with an important textile industry and with important textile workers’ unrest, there is also an overlap between the countries mentioned in the table and the ones discussed in this book. On the whole the decades of labour unrest in the table can also be identified in the national overviews presented here. The 1870s saw conflicts in Britain with the establishment of a permanent national federation of cotton spinners, and the defeat of the cotton weavers after a long and at times violent conflict in 1878, which would eventually lead to an amalgamated organization of weavers and a uniform wage list.87 The nineteenth-century wave of globalization and free trade, of which the dominant position of the Lancashire cotton industry may be considered a central fact, ended in 1914.88 The period 1914–1945 was characterized by increased tariff barriers and sharp international competition. Employers sought wage decreases; textile workers defended their gains, which resulted in sharp conflicts. That this period was a high point of textile workers’ militancy is on the whole also borne out in the national overviews, with examples such as the 1934 General Textile strike in 84 Beinin, ‘Egypt’. 85 Ibid. In St Petersburg the metal worker became the proverbial industrial worker, but actually textile workers were more numerous and very present in the revolution (Pretty, ‘Russia’). 86 Pretty, ‘Russia’; Quataert, ‘Ottoman Empire’; Bortz, ‘Mexico’; Beinin, ‘Egypt’; Roy, ‘India’; Cliver, ‘China’. 87 Alan Fowler, Lancashire Cotton Operatives and Work, 1900–1950: A Social History of Lancashire Cotton Operatives in the Twentieth Century (Aldershot, 2003), p. 23. 88 Kevin H. O’Rourke and Jeffrey G. Williamson, Globalization and History: The Evolution of a Nineteenth-Century Atlantic Economy (Cambridge and London, 1999). 789

The Ashgate Companion to the History of Textile Workers the USA, in Spain with the militant years in Barcelona between 1919 and 1923, and again in the 1930s and during the Civil War, in China with spontaneous protests of silk reelers and cotton mill workers in the 1920s and 1930s and the Communistled militancy of 1926/7, and in India with the first successful large-scale union mobilization in Bombay in 1926.89 Among countries not mentioned in the table, the national overviews also report high points of militancy in the interwar period in Argentina and the Netherlands.90 On a global level, we may conclude that textile workers’ militancy was strongly concentrated in the de-globalization period of 1914– 1945. Neither the nineteenth-century period of globalization nor the globalization period of the second half of the twentieth century led to a comparable wave of militancy. The race to the bottom certainly saw struggles of workers resisting the disappearance of their jobs in textile mills and struggles to establish better working conditions in the countries where they were relocated. Roy remarks that the 1982/83 Bombay textile strike was by some indices the largest industrial conflict the world has ever seen.91 Important exceptions like this one notwithstanding, textile militancy on a world scale has dwindled more than the industry has.92

Establishing connections In her 1994 presidential address to the American Historical Association, Louise Tilly set out to describe how the connections between the developments in the industrialized cotton industries of India, England and France influenced each 89 Blewett, ‘USA’; Fowler, ‘Great Britain’; Cliver, ‘China’; Roy, ‘India’. The national overview for Poland does not discuss worker militancy, but marks the decline of the industry in 1923–1925 (and the 1930s) which will have created protest. The national overview for Germany also does not discuss labour militancy in this period, but in Germany the first half of the 1920s was in general a high point of trade union organization and militancy, in which the textile sector was no exception (Hans-Ulrich Wehler, Deutsche Gesellschaftsgeschichte. Vierter Band. Vom Beginn des Ersten Weltkriegs bis zur Gründung der beiden deutschen Staaten 1914–1949 (Munich, 2003), pp. 314–315). The national overview for Mexico does not single out the 1930s as a particularly militant decade for Mexican textile workers. 90 Lobato, ‘Argentina’; Van Nederveen Meerkerk, Heerma van Voss and HiemstraKuperus, ‘Netherlands’. A possible exception is Japan, where unions in the interwar period were still too weak to organize substantial protest (Hunter and Macnaughtan, ‘Japan’). Still, within this period the years 1930–1931 formed a high point in militancy (Janet Hunter, Women and the Labour Market in Japan’s Industrialising Economy: The Textile Industry before the Pacific War (London and New York, 2003), p. 257). As textile workers’ militancy was also not very prominent after 1945, the interwar period may still be the relative high point of Japanese textile workers’ militancy. 91 Roy, ‘India’. 92 That the relocation of industries to areas where employers face a lower wage bill does not necessarily lead to the disappearance of peaks of labour unrest is clear from the very different results for the automobile industry (Silver, Forces of Labor, pp. 77–82). 790

Covering the world other.93 Since then, the awareness of transnational, even global connections and comparisons have become increasingly important in historical research. One way of coming to a global comparison is the strategy of the collective model, as practised in this project.94 By bringing together scholars on the history of textile workers from different countries, expertise and work effort were combined in order to reconstruct a long-term, global history of textile workers and at the same time formulate some all-encompassing conclusions. In doing so, the specificity of each case, the distinctions and varieties in the different regional, national and supranational histories should not be overlooked. It was noted that in actual historical circumstances ‘artisan’, ‘industrious’ and ‘industrial’ production relations existed simultaneously. However important differentiation is, the global and long-term approach has also enabled us to make general statements. The trends suggested here indicate that patterns and movements in the history of textile work and textile workers are rooted in both universal economic logic and unequal power relations on different levels. A comprehensive analysis of the history of textile workers of this type can only be made within a global comparative framework. The overarching trend distilled here is that of a drive towards the lowering of labour costs, achieved in many different ways (mechanization, employing women and children, wage systems, relocation to low-wage areas). The trend to lower wage costs as a percentage of total product costs could be countered. Although this volume has focused on the production of relatively cheap, basic textiles, there is a long tradition in the production of higher prized, higher quality fabrics. When the production of cheap bulk textiles had moved elsewhere, the remaining textile production usually consisted of high-value fibres and fabrics. In the best of cases, productive forces had been freed to concentrate on high-value products. Government measures could protect local and national markets from external competition, thus shielding entrepreneurs and workers alike. Guilds and trade unions demanded this kind of protection from local and national governments to shore up their members’ living standards. Internationally, the 1920s and 1930s formed a peak in workers’ resistance in the textile sector. This was a period not only of increased international competition due to downturns in the business cycle, but also of intensified government protection of national markets. The activities of trade unions also in other ways at least temporarily shielded their members from the drive to lower the wage bill. The global approach to the history of textile workers enables us to connect several trends pointed out in this chapter. Once more, it has become clear that a unidirectional perception of industrial development is flawed and should best be avoided. Textile production in both ‘Western’ and ‘non-western’ regions not necessarily followed the staged development from artisan via ‘proto-industrial’ 93 Louise A. Tilly, ‘Connections’, American Historical Review 99 (1994): 1–20. 94 See also Marcel van der Linden and Jan Lucassen, Prolegomena for a Global Labour History (Amsterdam, 1999), p. 21; Marcel van der Linden, ‘Hacer historia comparativa del trabajo: algunos preliminares esenciales’, Historia Social 33 (1999): 111–131. 791

The Ashgate Companion to the History of Textile Workers towards ‘modern’ industrialization. Increased globalization also led to relocation of textile industry on a global scale, often instigated by Western entrepreneurs in their search for cheaper labour power. And, finally, through the threat of relocation globalization further weakened the bargaining power of textile workers all over the world. Therefore, the relative militant behaviour of textile workers in most parts of the world between 1914 and 1945 might not be surprising. This was a period of increased national protection, and international tensions, or, one might argue, de-globalization. Textile workers’ militancy was thus most forceful in between two waves of globalization.

792

Index

Note: Page numbers in italic type refer to information in the Tables; those in bold type refer to Figures and Maps. Countries which are the subject of case studies are indicated in bold type; the case study is indicated by the sub-heading ‘history of textile production’ with further sub-headings arranged in chronological rather than alphabetical order. Aachen (Germany) 204, 212, 212, 215, 218, 223, 225, 227, 229, 571 Άbbas Halim 186, 187 absentee ownership 669 added value: Germany 209, 210, 213 India 263 Netherlands 389 Turkey 499 Uruguay 511, 513, 514, 515, 518, 529, 530 Administration of Spinning and Weaving (Egypt) 178 Aegean region (Ottoman Empire) 483–4 AFL (American Federation of Labour, USA) 541, 545, 760 Africa, textile imports from India 563, 564 Alamán, Lucas 338, 339 Aleppo (Ottoman Empire) 478, 484, 489, 490, 655–6 Algemene Bond van Textielarbeiders De Eendracht (Netherlands) 387 Άli, Mehmed 177, 179, 197, 639, 640 Allgemeine Deutscher Arbeiterverein (Germany) 228 Alpargatas (Argentina) 29, 35, 40, 41 Alsace-Lorraine (Germany) 200, 203, 205, 214, 217, 217, 221 Alto-Perú (Argentina) 21–2 Álvara (Law, Brazil) 78, 91–2, 94, 100

Amalgamated Union of Operative Cotton Spinners (Great Britain) 240–1, 787 American Civil War: cotton famine 23, 179, 207, 216, 260, 428, 429, 514, 569, 633 impact on USA textile industry 538–41 American Federation of Labour (AFL, USA) 541, 545, 760 American Woollen Company (USA) 549, 550 Americas: indigenous textile production 658 migrant workers 658–62, 676 proto-industrialization 658–9 slave labour 658–9 workshop production of textiles 657–8, 675 Anatolia (Ottoman Empire) 478, 482, 483, 490 Anderson Clayton 28 Anglo-Ottoman Trade Convention (1838) 179, 180, 640 Antuñano, Esteban de 338–9 AOT (Argentina) 36, 37, 40–1 apparel see clothing industry apprentice system 599, 687, 707, 712, 728, 729, 749, 751 China 123, 125, 728, 734 Denmark 154, 756 Egypt 709

The Ashgate Companion to the History of Textile Workers knitting industry 23, 27–8, 30, 32, 34 labour relations 24, 31–2, 36–7, 40–1, 611–12 see also collective bargaining migrant workers 24, 665, 687 network migration 672 night work, restrictions on 37 number of workers 30, 37 obrajes (workshops) 21, 661, 684 price control 34–5 production processes 29–30 production statistics 12, 27, 33, 34, 38, 39, 573 protectionism 17, 24 silk industry 28, 30 spinning: cotton 26–7, 27, 38 number of spindles 12, 26, 27, 38 wool 26–7 yarn production 12, 26, 27, 573 strikes 31–2, 32, 36–7, 41 see also labour relations textile machinery industry 28 tourism 20 trade unions 24, 31–2, 36–7, 40–1, 611–12, 765 see also collective bargaining value added of textiles 26 value of production 34 weaving: handloom weaving 19 number of looms 12, 38 number of weavers 30 women, role in textile production 30 wool industry 12, 22–4, 30 spinning 26–7 workers, number of 21, 25–6, 30, 30, 33, 34, 37, 39 working conditions 24, 31, 40–1 working hours 36, 37, 39 see also Latin America Argentinean Federation of the Textile Industry 37 see also CAIT Argentinean Textile Association (ATA) 36–7 Argentinean Textile Industry Confederation (CAIT) 28 Arkwright Club of Boston 546, 742 Arkwright, Richard 235, 236, 238, 244, 533 see also water frames artificial fibres 561–2, 574, 613–14, 614, 619 Argentina 25, 39

Great Britain 231, 235, 236–7, 732, 755 Habsburg Monarchy 49, 52, 654, 728 India 256, 265, 593 Japan 309, 310 Mexico 337, 338, 353 Netherlands 375, 376, 716 Ottoman Empire 491, 494 Spain 451, 456, 458, 459, 591 Uruguay 519, 521, 521 Arab-Israeli War (1967) 193 arbitration: Denmark 159–60, 162 Egypt 191 Spain 469, 470 Argentina 18 artificial fibres 25, 39 bureaucratization of work 611–12 capital investment 25, 34 child labour 30 civil war 22 collective bargaining 761 see also labour relations; strikes; trade unions cotton cultivation 25 cotton industry 12, 23, 25, 26–7, 27, 30, 573 spinning 26–7, 27, 38 customs policy 35–6 de-industrialization 39–40 decline of textile industry 37–41, 574, 770 employers’ associations 611–12 encomienda system 21, 661, 685, 727 ethnicity of textile workers 672 exports 22–3, 29, 33, 36 factory production of textiles 22, 23–4, 26, 28, 29, 30, 31, 38, 39 gender division of labour 30 growth rate of textile industry 37 health of textile workers 31 history of textile production 17–42, 574 indigenous textile production 17, 19–20, 21, 661, 727 colonial period 17, 20–2 modern textile industry 17, 19, 22–41 impact of First World War 25 import substitution 25, 28, 33, 36, 40 imports 28, 33, 36, 40 industrialization of textile production 23–4, 34 794

Index wool industry 12, 571 Austria 43, 44 cotton industry 12 decline of textile industry 72–3 development of spinning machines 53 exports 65, 66, 70 factory production of textiles 67–8 gender division of labour 68, 72 history of textile production 64–70, 71, 72 industrialization of textile production 68 labour relations 70 looms, number of 12 migrant workers 68 production statistics 12 protectionism 65 ‘Social Partnership’ 70 trade unions 70 women, role in textile production 68, 72 wool industry 12 workers, number of 12, 54, 71–2 yarn production 12 see also Czechoslovakia; Habsburg Monarchy Austrian Lands (Habsburg Monarchy) 48, 49 Austrian Netherlands (Habsburg Monarchy) 43, 45–6, 52 automatic looms 246, 641, 691 Denmark 151, 166 Great Britain 246, 247, 250 Japan 314, 327 Mexico 343 Netherlands 382 Russia/USSR 428 Uruguay 519–20 USA 553

China 137, 574 Europe 574 France 614 Germany 614 Great Britain 614 India 574 Indonesia 574 Italy 299–300 Japan 311, 324, 325, 574, 614 Mexico 357–8, 729 Netherlands 614 South Korea 574 Spain 472 Thailand 574 Turkey 574 USA 552–3, 574, 614 artisan production of textiles 599, 600, 707–8, 728–30 Brazil 730 China 118, 728 Denmark 154, 155, 729 Egypt 174, 176, 729 Habsburg Monarchy 47–8, 51, 52, 728 India 258–65, 266, 274 Italy 281 Japan 308, 309 Mexico 333, 335, 338 Netherlands 370–1, 728 Spain 603, 728, 752 see also workshop production of textiles Ashton, Thomas 241 Asia: de-industrialization 623, 634, 636 ethnicity of textile workers 656–7 factory production of textiles 622–3 migrant workers 656–7 spinning 634 textile imports from Great Britain 568 weaving 634 see also China; India; Japan Asociación Textil Argentina (ATA) 36–7 assimilation of indigenous populations, Brazil 82–3, 89–90 ATA (Argentinean Textile Association) 36–7 Atlantic Islands 90n60, 652, 657 see also Cape Verde Islands; Portugal Atlantic Triangle 564, 627, 637 Augsburg (Germany) 201, 213, 221, 222 Australia: labour relations 788

Baden (Germany) 202, 203, 225 Bangladesh, maternity rights of textile workers 615 baomaizhuzhi (putting-out system, China) 108, 117–18, 685 barafula fabrics (Cape Verde Islands) 657–8 Barcelona (Spain) 450, 455, 456, 458–9, 462, 463, 467, 469, 470–1, 566, 586, 603, 607, 652, 692 ‘barefoot aristocrats’ 237, 537, 699, 732, 747 Barnes, Ruth 561, 562 795

The Ashgate Companion to the History of Textile Workers Bavaria (Germany) 200, 202, 203, 217, 217, 224 Bedaux system 610, 611–12, 696 Denmark 152, 162–4 see also multiple loom system Beida Dyers (Egypt) 185, 190 Belém do Pará (Brazil) 82 Belgium 12, 13, 204, 784, 788 Benetton 616 Bengal (India) 254, 256, 257, 260 Białystok (Poland) 398, 401, 403, 407, 412 Biella (Italy) 276, 279, 290, 299 Bielsko-Biała (Poland) 398, 399, 407 Big Ten companies (Japan) 326 Blewitt, Mary 666 Bohemian Lands (Habsburg Monarchy) 43, 44, 47–8, 49, 50, 51, 54, 59, 566 Bolshevik revolution, Russia 440–1, 766–7 Bombay (India) 254, 269, 271 Bombay Textile Labour Union 758 bonded labour see slave labour Boott Mills (USA) 5, 547, 612–13 Bosnia and Herzegovina (Habsburg Monarchy) 44, 45 Boston Associates 533, 535 Bourgeois Revolution 57 Brandenburg (Germany) 200, 205, 205, 206, 206, 220 Brazil 76 artisan production of textiles 730 assimilation of indigenous populations 82–3, 89–90 ban on production of luxury products 78, 91–2, 94, 100 capital investment 96 child labour 96–7 cloth as a form of currency 83, 88, 661, 695, 714 cloth production 87, 97 clothing industry 98 cotton industry 12 cotton gins 85, 86 raw cotton consumption 12, 78, 92, 100 spinning 84, 85 cultural traditions in textiles 90, 91 decline of textile industry 99 domestic production of textiles 79, 84, 590–1 ethnicity of textile workers 667

exports 83–4, 92, 94, 98–9, 629 factory production of textiles 75, 78, 80, 82, 83, 84–5, 87, 88–90, 92, 93, 94–6, 100–1, 667 gender division of labour 88, 96 history of textile production 75–101, 566 pre-colonial period 78–9 sixteenth and seventeenth centuries 79–80 eighteenth century 80–93 nineteenth century 93–7 twentieth century to present 97–9 housing of textile workers 97, 668, 689 import substitution 8, 81, 98 imports 94, 95, 568 indigenous textile production 78–9, 86, 87, 88–90, 93, 661–2 industrialization of textile production 84, 95–7 labour 87–90 living conditions 97 manufactory production of textiles 95, 693, 696 men, role in textile production 88, 96 migrant workers 96, 665, 667, 687 night work 96, 692 paternalism 97, 668 pre-industrial textile production 77, 84 production statistics 12, 87, 94–5, 97 protectionism 77, 94, 638 proto-industrialization 78, 81–4, 100 putting-out system 600 slave labour 79–80, 87–8, 96, 600, 604, 658, 659, 667, 685, 730 spinning, cotton 12, 84, 85 textile exports 83–4, 92–3 wages 83, 88, 661, 695, 716 weaving, number of looms 12 women, role in textile production 88, 96 wool industry 12 workers, number of 12, 98, 99 working hours 89, 96, 98, 100, 692, 693 workshop production of textiles 79, 730 see also Latin America ‘Bread and Roses’ strike (USA) 549–50, 721 Bremen (Germany) 200, 206–7 Britain see Great Britain Brno (Habsburg Monarchy) 44, 55, 57, 655 796

Index Brooklands Agreement 1893 (Great Britain) 242–3 brotherhoods: China 129 Italy 284 Brügelmann, Johann Gottfried 216, 219 Buenos Aires (Argentina) 18, 23–4, 29, 30, 30, 31, 36, 38 Bulgaria 478, 483 bureaucratization of work 609–13 Argentina 611–12 Japan 611 Netherlands 611 USSR 612 Uruguay 612 USA 611, 612–13 Burlington Industries (USA) 359, 553, 554, 556–7 Bursa (Ottoman Empire) 478, 485 ‘Business Unionism’ 760–1

Spain 451, 452, 454 Uruguay 516, 518, 519 USA 531 Cardroom Amalgamation (Great Britain) 241 carpet production: Denmark 153 Ottoman Empire 482–3, 486, 489, 491, 493, 495, 583, 589, 623, 626, 665, 732, 735, 742 Turkey 497, 501, 503–4, 507, 508 USA 553, 557 Cartwright, Edmund 54 CAS (Continuous Automated Spinning) system (Japan) 327 Castile (Spain) 450, 452, 652 Catalonia (Spain) 450, 453–7, 454, 457, 459, 462–71, 473–4, 474, 586, 603, 607, 652, 672, 742, 763–4 Catholicism: France 653 Germany 229, 760 Ireland 654 Italy 284 Mexico 341, 351, 689, 740 Netherlands 385, 387, 692, 740, 760 Ottoman Empire 590 trade unions 229, 351, 387, 760 USA 535, 666, 756 CC.OO (Comisiones Obreras, Spain) 476 CCP (Chinese Communist Party) 135, 746, 756, 759, 768 Çelebi, Evliya 174, 175 Central Industrial Region (Russia) 422, 423, 425, 431, 433, 438, 440, 734, 741 see also Moscow (Russia); Vladimir (Russia) CETU (Congress of Egyptian Trade Unions) 188 CGOCM (Mexico) 355 CGT (Confederación General de Trabajadores, Mexico) 350, 351, 355 CGT (General Labour Confederation, Argentina) 31, 37, 41 chain migration see network migration Chambers of Commerce, Argentina 32 Chao, Kang 109 chemical fibres see artificial fibres Chemnitz (Germany) 214–15, 223–4, 228

CAD/CAM technologies 99 Cairo (Egypt) 172, 174–5, 177–8, 180, 181, 183–4, 190, 667–8, 684, 691, 692–3 Shubra al-Khayma 185, 188–9, 190, 192, 193 CAIT (Argentinean Textile Industry Confederation) 28 see also Argentinean Federation of the Textile Industry Calwer Moderation 204, 221 camlet production, Germany 213, 221 Campomar & Soulas (Uruguay) 516, 518–22, 520, 521, 526–7 Campomar (Argentina) 29 Canada, labour relations 788 Cape Verde Islands 657–8 see also Atlantic Islands capital investment: Argentina 25, 34 Brazil 96 carding 591, 599, 679, 707 Denmark 146 Germany 202, 205, 206, 206, 207, 208, 212, 218, 220 Great Britain 236, 240, 241 Italy 279, 280–1, 291 Japan 309, 327 Mexico 335, 336 Russia/USSR 445 797

The Ashgate Companion to the History of Textile Workers child labour 698, 716, 780, 781 Argentina 30 Brazil 96–7 China 121, 123 Egypt 179 Germany 223, 224–5 Great Britain 236, 241, 242, 730 Italy 281, 284, 288–9, 295 Japan 319 Mexico 340, 349 Netherlands 376, 377, 380, 382, 385 Ottoman Empire 480, 491, 731–2 Poland 404–5, 406, 410, 410, 411 Russia/USSR 434, 604 Spain 460 Turkey 506–7 Uruguay 526 USA 533, 552, 730 Chile 18, 671 China 104 artificial fibres 137, 574 artisan production of textiles 118, 728 CCP (Chinese Communist Party) 135, 746, 756, 759, 768 child labour 121, 123 cloth production 134, 136–7 comparative manufacturing costs 783 cotton cultivation 639 cotton industry 12, 107–8, 113, 115–19, 128, 129, 133, 134, 629–30 raw cotton consumption 12 spinning 107, 108, 111–12, 116, 586 weaving 108, 109, 117–18, 131–2, 133, 197, 734 criminal gangs 124–5, 604 de-skilling of textile work 135 domestic production of textiles 107–9, 117–19, 127, 131, 133, 137, 592–3, 600, 708, 726–7 ethnicity of textile workers 673–4 exports 105, 112, 113, 114, 115, 116, 117, 119, 132, 134–5, 137–8, 139, 618–19, 645, 727, 731, 734 factory production of textiles 106, 113, 114, 118, 119, 122–3, 131, 132–3, 134, 135–6, 137, 138, 570, 592–3, 618–19 failure to produce capitalist industrial revolution 106–7 fines imposed on workers 696

gender division of labour 103, 120–5, 126, 128, 135, 714, 715 Great Leap Forward 133, 136, 768 guilds 105, 106, 114–15, 128, 753 health of textile workers 124 hereditary artisans 103, 105 history of textile production 103–39, 570, 637 Qing Dynasty (1644–1911) 105–9 nineteenth century 109–12 People’s Republic of China 131–8 Hong Kong 71 housing of textile workers 124, 604, 689 imports 110–11, 113, 131, 429, 568, 570, 572 industrial textile workers 120–5 industrialization of textile production 112–15, 116–17 labour relations 128–30, 132–3, 135–6, 618–19, 721, 768, 771, 788 living conditions 124 manufactory production of textiles 119, 689 maternity rights 128, 135 men, role in textile production 103, 120–1, 714 see also gender division of labour merchants’ role in textile industry 105, 108, 109, 110, 111, 118, 629–30, 727, 734 migrant workers 138, 656–7, 673–4, 676, 677, 743 nationalization of the textile industry 746 night work 124, 125–6, 692 Number Ones 123, 124, 125, 129, 674, 688, 721, 738 piece-rates 120, 127, 133 production statistics 12, 116, 119, 131, 132, 133, 134, 136–7 protectionism 637 putting-out system 108, 117–18, 580, 592, 600, 685 recruitment of labour 123–5, 604, 674, 688, 738 rural textile production 107–9, 117–19, 127, 131, 133, 137 silk industry 103, 106, 112–15, 119, 121, 126–7, 128, 132, 134–5, 137, 138, 571, 591–3, 629 798

Index Netherlands 385, 387, 653, 740, 760 trade unions 385, 387 USA 534, 535, 666, 756 chronological sequence of textile development 622–5, 624, 774–7 CIDOSA (Mexico) 343 CIO (Congress of Industrial Organizations, USA) 554, 555 citizenship 370, 743–4, 751 Clapham, Sir John 244 clientilism 765 cloth as a form of currency: Brazil 83, 88, 661, 695, 716 Cape Verde Islands 658 cloth production: Brazil 87, 97 China 134, 136–7 India 260 Russia/USSR 430, 442, 446 clothiers 726 Great Britain 234–5 clothing industry: Brazil 98 Egypt 181–2, 195 Habsburg Monarchy 59 international trade 574–5 Italy 301–3 Turkey 499, 500–1 CNT (National Workers Union, Uruguay) 525 CNT (Spain) 468, 469, 470, 763–4 Coats 603, 613, 615 coerced labour see slave labour collective bargaining 761–3 Argentina 761 Denmark 156, 157, 159–60, 609, 761, 762–3 Germany 762 Great Britain 238, 242–3, 246, 247, 248–9, 252, 761 Japan 609, 762 Uruguay 521–2 see also labour relations; trade unions collective contracts, Mexico 347 colonialism 4, 5, 659, 669, 675, 677, 686, 766, 786 combing machines, number of 8, 12, 13 COMECON 64, 70 Comisiones Obreras (CC.OO, Spain) 476 Communists (Argentina) 31–2, 36

filatures 112, 113–15, 121, 126 weaving 114–15, 119–20, 121, 122–3, 126, 128, 129–30, 591–3, 733, 734 sisterhoods 129, 130, 720 spinning 116–17, 570, 731 cotton 107, 108, 111–12, 116, 586 ring spinning 734 spindles, number of 12, 113, 116, 133, 573 yarn production 12, 134, 136 strikes 128, 136 see also labour relations taxation 111 textile machinery industry 134 wages 108, 109, 127, 699, 717, 783 water power 106 weaving: cotton 108, 109, 117–18, 131–2, 133, 197, 734 handloom weaving 107–9, 117–19, 570, 699 number of looms 12, 131 number of weavers 120 power-loom weaving 117, 119, 127, 131, 137, 734 silk 114–15, 119–20, 121, 122–3, 126, 128, 129–30, 591–3, 733, 734 women, role in textile production 103, 120–5, 126, 128, 135, 714, 715 wool industry 12 workers, number of 12, 120, 131 working conditions 125–30, 135 working hours 125–6, 127, 692 workshop production of textiles 103, 106, 108, 117, 118–19, 129, 131, 133, 137, 138–9, 591–3, 629, 728, 734 see also Asia China Sericulture Company 131, 132, 134 Christianity: Catholicism: France 653 Germany 229, 760 Ireland 654 Italy 284 Mexico 341, 351, 689, 740 Netherlands 385, 387, 692, 740, 760 Ottoman Empire 590 trade unions 229, 351, 387, 760 USA 535, 666, 756 Protestantism: Ireland 654 799

The Ashgate Companion to the History of Textile Workers Belgium 12 Brazil 12 China 12, 107–8, 113, 115–19, 128, 129, 133, 134, 629–30 Czechoslovakia 12 Denmark 12, 147–8 Egypt 12 France 13 Germany 13, 201–2, 203, 204, 207, 208, 209–12, 209, 210, 213–14, 214, 215, 219–20, 223, 225–6, 226 mechanization 216–18 Great Britain 13, 233, 235–8, 240–1, 242–3, 246–8, 249, 250–1, 252 Habsburg Monarchy 46, 51, 54–5 India 12 Italy 13, 284, 287–9, 296, 299 Japan 12, 307, 312–13, 312, 313, 315, 322, 323, 324, 327, 572, 573 Mexico 12, 338–41, 339, 343–61 Netherlands 13, 366–7, 369, 370 number of workers: Argentina 30, 37 China 120 Egypt 184 Germany 208, 209, 209, 210, 211, 212 Habsburg Monarchy 50 Italy 296, 299 Mexico 339, 340, 349, 353, 357, 357 Netherlands 369, 380, 381, 383 Ottoman Empire 483–4 Poland 13, 400, 401, 402, 407, 408, 413, 414, 414, 418–19, 419 Russia/USSR 13, 423–48, 442, 447 Spain 13, 453–5, 454, 455, 457, 462–71, 473–4, 473 Turkey 499 Uruguay 516 USA 12, 533–4, 536–48, 547, 550–2, 551 Cotton Industry Act (1959), Great Britain 250 Cotton Manufacturing Industry Act (1934), Great Britain 247 cotton pieces, production of 12, 13 cotton printing industry 55 Habsburg Monarchy 57 printing, mechanization of 58 Russia/USSR 424, 425, 427, 429, 434, 436 USA 539, 541 cotton production, measures of 8

Communists (China) 129–30, 135, 746, 756, 759, 768 Communists (Czechoslovakia) 63 Communists (Poland) 412 Communists (Russia) 767–8 competition 700, 778 Confederación Argentina de Industrias Textiles (CAIT) 28 Confederación General del Trabajo (CGT, Argentina) 31 Congress of Egyptian Trade Unions (CETU) 188 Congress of Industrial Organizations (CIO, USA) 554, 555 Consejos de Salarios (Salary Councils, Uruguay) 521, 525 Continental System 66 contracting out see putting-out system contractors: China 124–5, 674, 685–6 India 267–8, 656 Japan 317 Copenhagen (Denmark) 142, 144 Córdoba (Argentina) 18, 20–1 Coromandel (India) 254, 256, 257 corporatism 762 Cortes, Hernan 333, 336 COT (Textile Workers Congress, Uruguay) 525 cotton: cotton gins, Brazil 85, 86 cultivation 562, 569, 638–40 Argentina 25 China 639 Egypt 639–40 India 639 Ottoman Empire 639 Russia/USSR 423, 444 Turkey 499 USA 639 global trade in 561–71, 572–6 1650–1780 562–6 1780–1913 566–71 1913–2000 572–5 long staple 179–80 raw cotton consumption (by country) 12, 13 cotton industry: Argentina 12, 30 Austria 12 800

Index cotton spinning see spinning Cotton Spinning Act 1937 (Great Britain) 248 Cotton Supply Association (Manchester) 23 cotton weaving see weaving counting houses, USA 537 Courtaulds 613, 614 criminal gangs, China 124–5, 604 CROM (Confederación Regional Obrera Mexicana, Mexico) 350, 351, 352, 354–5 Crompton’s mule 236, 238 see also mule spinning CTM (Confederación de Trabajadores de Mexico) 355, 356, 359 CTU (National Conference of Trade Unions, Denmark) 156, 162 cultural traditions in textiles, Brazil 90, 91 currency, cloth as a form of: Brazil 83, 88, 661, 695, 714 Cape Verde Islands 658 customs policy, Argentina 35–6 Czechoslovakia 44 cotton industry 12 Czech Party 60 de-industrialization 72 decline of textile industry 71–2 ethnicity of textile workers 673 exports 62, 64, 65 factory production of textiles 63 flax industry 61 hemp industry 61 history of textile production 61–4, 65, 70–1 import substitution 64 jute industry 61 looms, number of 12 nationalism and labour recruitment 673 nationalization of the textile industry 63 spindles, number of 12, 61 trade unions 63 wool industry 12 workers, number of 12, 71–2 yarn production 12 see also Austria, Habsburg Monarchy

Dansk Tekstilfabricantforening see Textile Manufacturers’ Federation (Denmark) de Vries, Boudien 652–3 de Vries, Jan 776 de-colonization 4, 640, 644, 783 de-globalization 5 de-industrialization 569, 581–4, 595, 632, 634–5, 636, 645, 775 Argentina 39–40 Asia 623, 634, 636 Czechoslovakia 72 India 258, 570–1 Italy 277 Spain 449 Uruguay 523 de-skilling of textile work 601, 617, 711, 712, 743, 747 China 135 Japan 314 Netherlands 739–40 Spain 463 USA 610 decline of textile industry 645, 769–71, 783 Argentina 37–41, 574, 770 Austria 72–3 Brazil 99 Czechoslovakia 71–2 Denmark 168–9 EEC countries 615 Egypt 565, 770 Great Britain 250–2, 573 India 582, 770–1 Italy 277–8, 302 Japan 323–5 Mexico 356–61, 569–70 Netherlands 389–95, 769–70 Ottoman Empire 479, 480–1, 569, 582 Poland 407, 418 Russia/USSR 448–9 Uruguay 517 USA 556–7 Democratic Movement for National Liberation (DMNL, Egypt) 189, 190 Denmark 142 artisan production of textiles 154, 155, 729 Bedaux system 152, 162–4 see also multiple loom system carding 146

Dalmatia (Habsburg Monarchy) 44, 45 Damascus (Ottoman Empire) 478, 484, 490–1, 495, 584–5, 600, 655 Danish Employers’ Confederation 156–7, 169 801

The Ashgate Companion to the History of Textile Workers ring spinning 148, 152 self-actor spinning machines 146, 741 yarn production 12, 148 state subsidy of textile industry 144 steam power 147 strikes 155, 158–9, 161, 162–3, 167–8, 612 see also collective bargaining; labour relations; trade unions textile machinery industry 144–5, 146–7 trade unions 155–64, 167–9, 612, 756, 757, 764–5 see also collective bargaining; labour relations; strikes wages 715 wage lists 155, 696 weavers 161 weaving: automatic looms 151, 166 cotton 147, 151–2, 160–2, 165 handloom weavers 146, 756 multiple loom system 148, 151–2, 158, 160–2, 163, 612, 780 number of looms 12 power-loom weaving 146, 147, 151–2, 737 women, role in textile production 155, 164–8, 165, 588, 739 wool industry 12, 149, 150 workers, number of 145, 147, 148–9, 149, 152 working conditions 160–2 workshop production of textiles 143, 154 zünft system 154, 155, 756 design patterns 90 Deutscher Textilarbeiterverband (DTAV, Germany) 757 DISK (Turkey) 504, 505 DMNL (Democratic Movement for National Liberation, Egypt) 189, 190 domestic production of textiles 579–80, 590–4, 598–9, 600, 626, 707, 708–10, 726–7, 775–6 Brazil 79, 84, 590–1 China 107–9, 117–19, 127, 131, 133, 137, 592–3, 600, 708, 726–7 Denmark 141, 143 Germany 217, 219 Great Britain 233–4, 235, 236, 708–9 Habsburg Monarchy 47–8, 59–61, 594

carpet production 153 collective bargaining 156, 157, 159–60, 609, 761, 762–3 see also labour relations; trade unions cotton industry 12, 147–8 weaving 147, 151–2, 160–2, 165 decline of textile industry 168–9 domestic production of textiles 141, 143 exports 145, 153 factory production of textiles 144–5, 147–8, 152, 162–3 gender division of labour 155, 164–8, 165, 588, 739 guilds 144, 154, 584, 753 history of textile production 141, 143–53, 651 manufactures 143–5 industrialization 146–52 increasing productivity 150–2 post-war climax 152–3 industrialization of textile production 146–52 labour relations 154–69, 609, 612 lockouts 150, 156–7, 169, 609 see also collective bargaining; strikes; trade unions manufactory production of textiles 143–5, 753 men, role in textile production 164–8, 588 see also gender division of labour merchants’ role in textile industry 144 multiple loom system 148, 151–2, 158, 160–2, 163, 612, 780 see also Bedaux system peasant labour 141, 143 piece-rates 157–8, 159–61, 166 pre-industrial textile production 141, 143, 154, 164 production processes 151–2, 165–6 production statistics 12, 145, 152 productivity 150–2 protectionism 150, 152 putting-out system 144 rural textile production 141, 143 ‘September Compromise’ 157, 162, 609 social democracy 149–50, 154, 162, 163, 764 spinning: number of spindles 12 802

Index dyeing industry 180, 195 employers’ associations 607–8 employment contracts 195 ethnicity of textile workers 667–8 exports 173, 176, 178, 180, 194 factory production of textiles 178–9 flax industry 178, 179, 180, 183 gender division of labour 174, 179, 181, 182, 183, 194, 714 guilds 173, 175–6, 177, 604 hat production 178, 179 history of textile production 171–97 seventeenth and eighteenth centuries 174–6 nineteenth century 177–84 machine manufactured imports and local craft production 180–4 mechanization, trade unions and the nationalist movement 184–7 textile workers’ unions and the nationalist movement 187–90 Free Officers and the trade union movement 190–2 towards a neo-liberal order 192–7 imports 180–1 industrialization of textile production 184–7 jute industry 178 labour relations 186–92, 193, 196, 604, 607–8, 788–9, 788 lockouts 607 see also strikes; trade unions luxury products 180 men, role in textile production 183, 714 see also gender division of labour merchants’ role in textile industry 173, 182 migrant workers 667–8 Muslims 173, 174, 186, 189, 665, 740 nationalization of the textile industry 174, 192 peasant labour 177, 181, 186, 187, 188 piece-rates 177 production statistics 184–5, 194 putting-out system 178, 181 silk industry 176, 183 slave labour 177 spinning: cotton 179, 184–5 number of spindles 12, 184

impact of industrialization 681 India 256, 593, 600, 727 Italy 279, 281, 282, 288 Mexico 333, 335, 591 Poland 399, 400, 403 Russia/USSR 426 Turkey 508, 582 Uruguay 522, 581–2 USA 533 working conditions 680, 684–5, 691, 693, 701 see also peasant labour dormitories see housing of textile workers drapers, Netherlands 371 dress, European 89, 499 DTAV (Deutscher Textilarbeiterverband, Germany) 757 Ducilo S.A. (Argentina) 28, 29 DuPont (USA) 28, 614, 619 dyeing industry: Egypt 180, 195 Russia/USSR 424 East Africa, textile imports from India 563 East Anglia (Great Britain) 231, 233, 234–5, 653 East India Company (Great Britain) 257–8, 566, 629, 639, 656 Eastern European countries, transfer of textile production to 68–9, 71 economic crashes: of 1873 58, 69 of 1929 62, 66 of 1973 68, 69, 323, 644 education of textile workers, Japan 331 EEC countries 393 decline of textile industry 615 protectionism 393 see also EU (European Union) Egypt 172 artisan production of textiles 174, 176, 729 child labour 179 clothing industry 181–2, 195 cotton cultivation 639–40 cotton industry 12 spinning 179, 184–5 weaving 184–5 decline of textile industry 565, 770 803

The Ashgate Companion to the History of Textile Workers yarn production 12 strikes 187, 188, 189, 190–1, 193, 196 see also labour relations; trade unions textile imports from India 563, 565 trade unions 173–4, 184–5, 186–92, 766 see also labour relations; strikes wages 695, 716 spinners 194 weavers 176, 177, 183, 194 weaving 177–8 cotton 184–5 handloom weavers 181, 182, 183, 184 number of looms 12, 184 number of weavers 180, 182, 183, 184 power-loom weaving 179, 184 silk 183 women, role in textile production 174, 179, 181, 182, 194, 714 see also gender division of labour workers, number of 178, 180–1, 182–3, 184, 185, 193–4, 195, 684 working conditions 183, 195–6, 690, 691 working hours 192, 194, 692–3 workshop production of textiles 173, 174–5, 177–9, 180, 599 Egyptian Federation of Trade Unions (ETUF) 192, 194 Eichsfeld (Germany) 200, 204–5 electrical power, Japan 314 Ellena, Vittorio 295–6 employers’ associations: Argentina 611–12 Denmark 156–7 Egypt 607–8 Great Britain 242–3, 606–7, 608–9 Japan 316, 607, 609 and labour relations 605–10 Mexico 607 Netherlands 387, 388, 607 Poland 411 Spain 607 Switzerland 49, 55, 65–6 employment contracts: Egypt 195 Japan 320 Mexico 353 Poland 405

encomienda system 660–1, 736 Argentina 21, 661, 685, 727 Mexico 337, 729 see also slave labour England see Great Britain England method of payment (piece rates) 157 etatism (Turkey) 497, 503, 507 ethnic division of labour, Ottoman Empire 493 ethnicity: definition 648 of textile workers 647–77 Americas 658–62, 676 Argentina 672 Asia 656–7 Brazil 667 Cape Verde Islands 657–8 China 673–4 Czechoslovakia 673 Egypt 667–8 Europe 650–5 France 673 Germany 673 modern period 662–74 Ottoman Empire 493, 655–6, 669–70, 676 sixteenth to eighteenth century 650–62 twentieth century to present 670–3 Uruguay 672 USA 665–6, 671 ETUF (Egyptian Federation of Trade Unions) 192, 194 EU (European Union) 71, 501 Spain’s membership of 472–3 see also EEC Europe: artificial fibres 574 ethnicity of textile workers 650–5 exports 681 guilds 750–1 imports 176 migrant workers 650–5, 663–4, 675–6 protectionism 565, 575 textile imports from Great Britain 567 textile imports from India 563–4, 565 European Recovery Programme 67 European Union see EU exchange rate policies, Uruguay 528 804

Index Germany 202, 204, 217, 219–20, 222–7, 580 Great Britain 235–8, 251, 252, 631 Habsburg Monarchy 49, 50, 54, 55, 56, 57, 59–60, 61, 594 India 261, 262–3, 265, 266–8, 271, 272, 593, 673 Italy 284–5, 287–92, 297, 298 Japan 305, 310–11, 314–15, 315, 317–20, 318, 328–9 Mexico 338–9, 339, 340–2, 343–4, 349, 358, 360, 569–70, 590 Netherlands 379–80, 380, 384 Ottoman Empire 480, 485, 486, 491 Poland 401, 402–4, 405, 406, 407, 407, 409, 412 Russia/USSR 424–6, 427–8, 429–32, 430, 448, 618 Spain 451–2, 453–5, 454, 455–6, 457, 462–4, 472 Uruguay 514, 516, 581–2 USA 533–4, 535–6, 539, 540 see also industrialization of textile production Fall River (USA) 532, 534, 535, 536–9, 540, 540, 541, 542–5, 721–2 families, role in recruitment of labour 710 China 124 family-based production of textiles see domestic production of textiles fashion industry, Italy 301–3 fazendas, textile production 79, 590–1, 685 Federación Argentina de la Industria Textil 37 Federación de la Industria Textil Argentina (FITA) 40 Federación de la Industria Textile Espanola (FITE, Spain) 466–7 Federación de las Tres Clases de Vapor (TCV, Spain) 461–2, 465, 466, 763, 764 Federación Obrera de la Industria Textil (FOIT), Argentina 31 feminization of the workforce 711–15, 723 China 135, 711 Denmark 165, 588 Egypt 181, 582 Mexico 713 Russia/USSR 434, 438 Spain 464 USA 713 Filature Nationale d’Egypte 184–5, 189

export restrictions, Argentina 29 exports 621, 626, 631, 637–8, 642 Argentina 22–3, 29, 33, 36 Austria 65, 66, 70 Brazil 83–4, 92, 94, 98–9, 629 China 105, 112, 113, 114, 115, 116, 117, 119, 132, 134–5, 137–8, 139, 618–19, 645, 727, 731, 734 Czechoslovakia 62, 64, 65 Denmark 145, 153 Egypt 173, 176, 178, 180, 194 Germany 208, 214–16, 214 Great Britain 95, 110–11, 231, 243–4, 249, 634 Habsburg Monarchy 47 India 255–6, 260, 272, 563–5, 570, 572, 573, 575, 582, 635, 639 Italy 278, 286–7, 297, 299, 301 Japan 307, 312–3, 313, 315–16, 322, 323–4, 323, 325, 637 Mexico 358, 359 Netherlands 365, 378, 393 Ottoman Empire 479, 480, 481, 482, 483, 582, 583, 639 Poland 400, 401, 409, 413 Russia/USSR 429, 444 Spain 466, 468 spatial aspects of 621 Turkey 500–2, 507, 508 Uruguay 514, 515, 516, 517, 518, 523, 527, 529–30 USA 534, 546, 568 Fabrikanten Vereniging Enschede (FVE, Netherlands) 387, 388 fabriqueurs (Netherlands) 371 factory production of textiles 579, 583, 590–1, 601–5, 631, 632–6, 684 Argentina 22, 23–4, 26, 28, 29, 30, 31, 38, 39 Asia 622–3 Austria 67–8 Brazil 75, 78, 80, 82, 83, 84–5, 87, 88–90, 92, 93, 94–6, 100–1, 667 China 106, 113, 114, 118, 119, 122–3, 131, 132–3, 134, 135–6, 137, 138, 570, 592–3, 618–19 Czechoslovakia 63 Denmark 144–5, 147–8, 152, 162–3 Egypt 178–9 805

The Ashgate Companion to the History of Textile Workers filatures, silk, China 112, 113–15, 121, 126 fine fabrics: Germany 207–8, 212, 215 USA 535–6 see also luxury products fines imposed on workers 599, 696 China 696 Germany 228 Italy 285 Mexico 342, 345, 350, 696 Russia/USSR 434, 436, 696 USA 541 finishing processes: Germany 220, 223 Habsburg Monarchy 54–5, 59 India 572 Italy 281, 283 Japan 308, 318, 319, 326 Netherlands 391 Russia/USSR 424, 425–6, 427 Spain 460, 467, 468–9 First Five Year Plan (USSR) 444, 445 First Opium War (1841–1842) 110 First World War: impact on Argentina 25 impact on Germany 201 impact on Great Britain 245–6 impact on Italy 298 impact on Poland 407 impact on Russia 439–41 impact on Uruguay 515 FITA (Federación de la Industria Textil Argentina) 40 FITE (Federación de la Industria Textile Española, Spain) 466–7 flax industry: Czechoslovakia 61 Egypt 178, 179, 180, 183 Germany 207, 209 Habsburg Monarchy 51 Italy 283, 297 Poland 399, 400, 404, 406, 407, 413, 414, 418, 419 Russia/USSR 431 see also linen industry FOIT (Workers’ Federation of the Textile Industry, Argentina) 31 forced labour see slave labour foreign investment, Mexico 343–4 foreign workers see migrant workers

framework document 3, 10 France: artificial fibres 614 Catholicism 653 cotton industry 13 ethnicity of textile workers 673 guilds 753 labour relations 788 migrant workers 663, 673, 677 number of combing machines 13 spinning 13 textile industry 566 Turkish textile businesses 508–9 weaving, number of looms 13 wool industry 13 workers, number of 13 Free Officers (Egypt) 190 friendly societies 720 Great Britain 235, 237, 242, 756 Italy 293 fustian industry 567, 627 Germany 201, 207 Great Britain 233 Habsburg Monarchy 47, 51 Italy 283, 287 Netherlands 366, 378 Galicia (Poland) 406, 407 Galicia (Spain) 450, 451, 591, 652 Gamma Holdings NV (Netherlands) 613 Gap textile plant (Belfast) 615–16 garment industry see clothing industry gedik (Ottoman Empire) 487–8 gender division of labour 4, 595, 599, 703–24, 730–3, 777 Argentina 30 Austria 68, 72 Brazil 88, 96 China 103, 120–5, 126, 128, 135, 714, 715 Denmark 155, 164–8, 165, 588, 739 Egypt 174, 179, 181, 182, 183, 194, 714 Germany 199, 213, 223–4, 704 Great Britain 236, 237, 238, 239, 241, 242, 251, 663–4, 714–15 Habsburg Monarchy 48, 52, 56, 708 India 264–5, 269–70, 272, 664, 705 Italy 279, 281–2, 288–9, 295, 300 Japan 310, 311, 318, 319, 320–1, 329–30, 699, 704, 705 and labour relations 719–22 806

Index Mexico 337, 340, 342, 344, 347–8, 349, 351, 353, 359, 590, 705 Netherlands 370, 374–5, 374, 376, 377, 378, 379, 380, 382, 385, 390, 391, 589,704, 705 Ottoman Empire 479–80, 491–3, 494, 589, 714 Poland 403, 404–5, 406, 409, 409, 410, 411, 416, 417 pre-industrial textile production 707–8 Russia/USSR 433–4, 438–40, 443, 445, 446 Spain 451, 453, 456, 458–61, 460, 464, 465, 466–7, 473–5, 588–9 Turkey 502, 503 Uruguay 522, 523–4, 524, 525, 526, 527 USA 534–5, 540, 540, 542–3, 549–50, 552, 659, 666 wage differentials 698, 715–17 General Labour Confederation (CGT, Argentina) 31, 37, 41 General and Municipal Workers’ Union (Great Britain) 251–2 General Trade Union of Textile Workers ‘Unity’ (Netherlands) 387 General Union of Textile Workers (Great Britain) 240, 245, 248 see also National Union of Textile Workers geographical distribution of textile production see spatial division of labour Germany 66–7, 200 artificial fibres 614 camlet production 213, 221 carding 202, 205, 206, 206, 207, 208, 212, 218, 220 Catholicism 229, 760 child labour 223, 224–5 collective bargaining 762 see also labour relations; strikes; trade unions cotton industry 13, 201–2, 203, 204, 207, 208, 209–12, 209, 210, 213–14, 214, 215, 216–18, 219–20, 223, 225–6, 226 spinning 202, 203, 209–10, 209, 211, 216–17, 223 weaving 203, 207, 210–11, 210, 217–18, 219–20 domestic production of textiles 217, 219 ethnicity of textile workers 673 exports 205, 208, 214–16, 214

factory production of textiles 202, 204, 217, 219–20, 222–7, 580 fine fabrics 207–8, 212, 215 fines imposed on workers 228 finishing processes 220, 223 flax industry 207, 209 fustian industry 201, 207 gender division of labour 199, 213, 223–4, 704 guilds 221, 223–4, 227, 752 health of textile workers 227 history of textile production 199–229 markets 214–16 organization of production 219–21 production 207–8 technological development 216–19 workers 221–9 housing of textile workers 227 impact of First World War 201 industrialization of textile production 211, 216–21 labour relations 227–9, 788 see also collective bargaining; strikes; trade unions linen industry 208 living conditions 227 manufactory production of textiles 199, 221, 222, 224 men, role in textile production 199, 213, 223 see also gender division of labour merchants’ role in textile industry 206 migrant workers 222–3, 663, 673, 677 night work, restrictions on 224 noise 227 production statistics 13, 206, 208, 209, 210, 211–14, 220, 221 productivity 210 proto-industrialization 220 putting-out system 221, 227 rural textile production 219, 221, 222 social democracy 228–9, 745, 760 spinning: cotton 202, 203, 209–10, 209, 211, 216–17, 223 number of spinners 209, 210, 212, 213, 214 ring spinning 217, 217 self-actor spinning machines 217, 217, 218 807

The Ashgate Companion to the History of Textile Workers wool 204, 206, 218, 220 yarn production 13, 204, 205, 206, 209, 216–17, 217, 218 steam power 202, 219 strikes 228 see also collective bargaining; labour relations; trade unions trade unions 227–9, 757, 758–9, 760 see also collective bargaining; labour relations; strikes Turkish textile businesses 509 wages 225–6, 226 water power 202, 219 weaving 219, 222, 224, 226 cotton 203, 207, 210–11, 210, 217–18, 219–20 handloom weavers 211, 217–18 number of looms 13, 203, 205, 208, 210, 211, 217 number of weavers 210, 211, 212, 214 power-loom weaving 210, 217–18 wool 204, 205, 218, 221 women, role in textile production 213, 223–4, 704 see also gender division of labour wool industry 13, 204–7, 207–8, 209, 212–13, 215–16, 220, 221, 223 number of workers 208, 209, 214 spinning 204, 206, 218, 220 weaving 204, 205, 218, 221 see also worsted industry workers, number of 13, 204, 208–9, 209, 210, 210, 211–12, 213, 214, 217, 219, 221, 222 wool industry 208, 209, 214 working hours 224–5, 226, 692 workshop production of textiles 220, 222 worsted industry 204, 205, 206, 212 see also wool industry Gewerkschaft Textil Bekleidung Leder (Austria) 72 Gierek, Edward 413, 416 Glasgow Master Spinners’ Association (Scotland) 606 global peripheries, transfer of textile production to 68–9 global trade in textiles 561–76, 622–3, 636 1650–1780 562–6 1780–1913 566–71

1913–2000 572–5 globalization 5–6, 643–5, 706, 713, 782, 789–90, 792 and the Central European textile industry 70–3 Mexico 358–61 and USA 556–7 Gordon, Wendy 663–4, 665–6 Grafa 29 Great Britain 232 artificial fibres 614 carding 236, 240, 241 child labour 236, 241, 242, 730 collective bargaining 238, 242–3, 246, 247, 248–9, 252, 761 see also labour relations; strikes; trade unions comparative manufacturing costs 783 cotton industry 13, 233, 235–8, 240–1, 242–3, 246–8, 249, 250–1, 252 exports 567–8, 569–71 imports 567 raw cotton consumption 13, 78, 92, 100 spinning 236–7, 586, 755 trade unions 237–8, 240–1, 242–3, 249, 251, 252 weaving 238, 247–8 cotton prices 22 decline of textile industry 250–2, 573 domestic production of textiles 233–4, 235, 236, 708–9 employers’ associations 242–3, 606–7, 608–9 exports 95, 110–11, 231, 243–4, 249, 634 factory production of textiles 235–8, 251, 252, 631 friendly societies 235, 237, 242, 756 fustian industry 233 gender division of labour 236, 237, 238, 239, 241, 242, 251, 663–4, 714–15 government intervention in textile industry 246, 252 guilds 753 history of textile production 231, 233–52 industry in Lancashire and Yorkshire (c. 1700) 231, 233–5 Industrial Revolution 235–9 mass unionism 240–5 crisis and depression 245–9 808

Index cotton industry 237–8, 240–1, 242–3, 249, 251, 252, 786–7 wool industry 235, 239, 240, 243–5, 248–9, 252 see also collective bargaining; labour relations; strikes wages 241–2, 245, 247, 717, 783 wage lists 241, 244–5, 247, 696 weavers 241, 247–8 weaving: automatic looms 246, 247, 250 cotton 238, 247–8 handloom weavers 235, 236, 238, 239, 756 looms, number of 13 wool 238 women, role in textile production 236, 237, 238, 239, 241, 242, 251, 663–4, 714–15 wool industry 13, 233, 234–5, 238–9, 239, 240, 243–5, 246, 248–9, 251, 252 raw wool consumption 13 spinning 238–9 weaving 238 see also worsted industry wool-combers 239 workers, number of 13, 240, 615 worsted industry 233, 234–5, 238, 243 see also wool industry Great Leap Forward (China) 133, 136, 768 Greater Poland 399, 401 Green Gang (China) 124–5, 604 Green, Nancy 7, 676–7 guest workers 392, 500, 508, 675, 677, 743 guilds 584–5, 599, 689–90, 707–8, 749–53, 776–7 China 105, 106, 114–15, 128, 753 Denmark 144, 154, 584, 753 differences from trade unions 756–7 Egypt 173, 175–6, 177, 604 Europe 750–1 France 753 functions of 751 and gender 719 Germany 221, 223–4, 227, 752 Great Britain 753 Habsburg Monarchy 47, 49, 752 Italy 277, 284 Japan 308, 309, 316 Mexico 338

post-war decline 250–2 impact of First World War 245–6 import substitution 233, 636 imports 105 Industrial Revolution 171, 173, 235–9, 423 labour relations 237, 238, 240–5, 246, 247–9, 250–1, 252, 603, 607, 608–9, 788, 789 lockouts 606, 608 see also collective bargaining; strikes; trade unions machine-smashing 238, 239, 726 men, role in textile production 236, 237, 239, 241 see also gender division of labour merchants’ role in textile industry 606 migrant workers 251, 663–4 to Brazil 96 modernization of textile production 250–1 night work 696 restrictions on 718, 723 paternalism 244 peasant labour 235 piece-rates 236, 244–5 piecers 237, 240, 695, 732 protectionism 565, 568 spinning 631, 730–1, 732–3 cotton 236–7, 586, 755 mechanization of spinning process 53 mule spinning 238, 239, 241, 244, 712, 730, 732, 733, 755, 787 number of spindles 13 ring spinning 217, 238, 250 self-actor spinning machines 237, 239, 240, 246 wool industry 238–9 yarn production 13 steam power 236 strikes 239, 243 see also collective bargaining; labour relations; trade unions textile exports: to Argentina 22 to Brazil 78, 94–5 to China 110–11 to India 258 to Portugal 80 textile imports, from India 563 trade unions 757, 764–5 809

The Ashgate Companion to the History of Textile Workers Netherlands 366, 366, 370–2, 373, 375, 377–8, 594 Ottoman Empire 479–80, 486–9, 491–2, 584–5, 656, 752 Poland 399 Spain 452–3, 752 Gustafsson, Bo 750, 751 Gutsherrschaft (Habsburg Monarchy) 45 Gutswirtschaft (Habsburg Monarchy) 45

night work, restrictions on 406 peasant labour 56 protectionism 53 proto-industrialization 48–52 putting-out system 48–9, 56, 61, 594 recruitment of labour 52 rural textile production 47–8, 56 serfdom 52, 654, 690 silk industry 46, 51, 59 spinning 53, 58, 586 cotton 54, 55 trade unions 58, 60 water power 48 weaving 56, 633 handloom weaving 54, 56, 57, 59, 60–1, 699 power-loom weaving 54, 55, 739 women, role in textile production 48, 52, 56 see also gender division of labour wool industry 46, 50, 51 workers, number of 50 working conditions 57, 406 workshop production of textiles 48, 50, 52, 54, 56, 59, 654, 728 see also Austria, Czechoslovakia HAK-IŞ (Turkey) 504, 760 handicraft production see domestic production handloom weaving 586, 633, 634 China 107–9, 117–19, 570, 699 Denmark 146, 756 Egypt 181, 182, 183, 184 Germany 211, 217–18 Great Britain 235, 236, 238, 239, 756 Habsburg Monarchy 54, 57, 59, 699 India 255, 257–8, 260–5, 266, 270–1, 272, 273, 583, 593, 600, 673, 695, 700, 705, 735, 746 Italy 297 Netherlands 381, 381 Ottoman Empire 485 Russia/USSR 425–6, 427, 428, 429, 684 Spain 451, 456, 458, 462, 591, 728 USA 531, 533 hang (China) 105 Hanover (Germany) 200, 203, 205, 205, 207, 217 Hargreaves’ spinning jenny 236, 370, 567, 730, 779

Habsburg Monarchy 44 artisan production of textiles 47–8, 51, 52, 728 clothing industry 59 cotton industry 46, 51, 54–5 cotton printing 57 spinning 54, 55 domestic production of textiles 47–8, 59–61, 594 exports 47 factory production of textiles 49, 50, 54, 55, 56, 57, 59–60, 61, 594 finishing processes 54–5, 59 flax industry 51 fustian industry 47, 51 gender division of labour 48, 52, 56, 708 guilds 47, 49, 752 history of textile production 43–61, 566, 585, 594 rural subsistence and urban crafts 1650–1720 47–8 proto-industrialization 1720–1780 48–52 response to British challenge 1780–1873 52–8 changing regional and sectoral patterns 1873–1918 58–61 industrialization of textile production 54–5, 58, 59 knitting industry 59, 61, 69 labour relations 57–8, 60 lace-making 55, 59 linen industry 46, 51 living conditions 57 machine-smashing 58 manufactory production of textiles 49–50, 51, 52, 53, 54, 56, 594, 654, 752 men, role in textile production 48, 56 see also gender division of labour migrant workers 52, 654–5 810

Index hat production: Egypt 178, 179 Italy 279 headmen (India) 257 health of textile workers 602, 683 Argentina 31 China 124 Germany 227 Italy 288–9 Japan 320, 321 Mexico 336–7 Poland 405, 420 USA 534 hemp industry: Czechoslovakia 61 Italy 282, 283, 297 Poland 399, 400, 404 Spain 454 home production see domestic production of textiles Hong Kong (China) 71 hours of work see working hours household production see domestic production of textiles housing of textile workers 603–4, 668, 686, 688–9, 695, 718 Brazil 97, 668, 689 China 124, 604, 689 Germany 227 Italy 294 Japan 320–1, 329, 330, 331, 603–4, 689, 691, 718 Mexico 341, 342, 350, 358, 668, 689 Netherlands 689 Ottoman Empire 492, 494, 503 Russia/USSR 433, 689 Spain 463, 466, 467, 689 USA 551, 552, 666, 668 Howard, Robert 541, 545 Huguenots 202, 653–4 Hungary 43, 44, 46, 68–9 hydroelectric power, Mexico 343

Imperial Chemical Industries (ICI, British producer of synthetic fibres) 28, 251, 614 import substitution 25, 623, 631, 636, 640, 641–3, 644, 645, 779 Argentina 25, 28, 33, 36, 40 Brazil 8, 81, 98 Czechoslovakia 64 Great Britain 233, 636 Turkey 500, 507 Uruguay 511, 515–17, 522, 528, 529 import taxes see protectionism imports: China 110–11, 113, 131, 429, 568, 570, 572 Egypt 180–1 Mexico 333 Ottoman Empire 480 USA 556 Inalcik, Halil 568, 570 India 254 artificial fibres 574 artisan production of textiles 258–65, 266, 274 cloth production 260 comparative manufacturing costs 783 cotton industry 12 cotton cultivation 639 raw cotton consumption 12 spinning 259–60 weaving 257–8, 260–5, 266, 271 de-industrialization 258, 570–1 decline of textile industry 582, 770–1 domestic production of textiles 256, 593, 600, 727 employment practices 262–3 exports 255–6, 260, 272, 563–5, 570, 572, 573, 575, 582, 635, 639 factory production of textiles 261, 262–3, 265, 266–8, 271, 272, 593, 673 finishing processes 572 gender division of labour 264–5, 269–70, 272, 664, 705 history of textile production 6, 253, 255–74, 572, 574, 582, 583, 593 early modern trade 255–8 Industrial Revolution and the artisans 258–65 Industrial revolution and the mill workers 266–70, 266

I.C. Modeweg (Denmark) 145, 146, 152 ICI (Imperial Chemical Industries, British producer of synthetic fibres) 28, 251, 614 identities of textile workers 725–47 IMF 194, 500 immigrant workers see migrant workers 811

The Ashgate Companion to the History of Textile Workers indianas factories (Spain) 453, 454, 455 indigenous populations, assimilation of, Brazil 82–3, 89–90 indigenous textile production 727 Americas 658 Argentina 19–20, 21, 661, 727 Brazil 78–9, 86, 87, 88–90, 93, 661–2 Chile 671 Mexico 333, 335, 604, 660, 661, 727, 729 Uruguay 526 Indonesia: artificial fibres 574 textile production 615 industrial espionage 53–4 Industrial Revolution 4 Great Britain 171, 173, 235–9, 423 Industrial Workers of the World (IWW, USA) 549–50 industrialization of textile production 577–81, 622, 623, 624, 625, 627, 631–8, 641, 705–6, 778–9 Argentina 23–4, 34 Austria 68 Brazil 84, 95–7 China 112–15, 116–17 Denmark 146–52 Egypt 184–7 Germany 211, 216–21 Great Britain 235–9 Habsburg Monarchy 54–5, 58, 59 impact on domestic production of textiles 681 India 266–70 Italy 285–95 Japan 310, 311, 312, 314–17 and labour relations 601–9 Mexico 338–46 Netherlands 378–89 Poland 400, 402 Russia/USSR 428–9 Spain 455–62 see also factory production of textiles ‘industrious production’ 776, 791 ‘industrious revolution’ 776 internal migration see migrant workers International Programme on the Elimination of Child Labour (IPEC) 506 International Textile, Garment and Leather Workers’ Federation (ITGLWF) 616

from autarky to reintegration 270–2 impact of British competition 570–1, 635–6 industrialization of textile production 266–70 jute industry 267, 270 labour relations 268–70, 271, 788 see also strikes; trade unions maternity rights 270 mechanized spinning 111 men, role in textile production 265, 272, 664, 705 see also gender division of labour merchants’ role in textile industry 261–2, 263, 274 migrant workers 82, 261, 264, 265, 267–8, 269, 270, 593, 656, 664, 673, 676 network migration 656 night work, restrictions on 270, 718, 723 peasant labour 257, 267, 270 production statistics 12, 266 recruitment of labour 267–8, 656 rural textile production 264 silk industry 264 spinning 111, 259–60, 731 cotton 12, 259–60 strikes 269, 271 trade unions 268–9, 758, 765–6 see also labour relations; strikes wages 263, 268, 274, 783 weaving 257, 260–5, 266, 593, 735 cotton 257–8, 260–5, 266, 271 handloom weavers 255, 257–8, 260–5, 266, 270–1, 272, 273, 583, 593, 600, 673, 695, 700, 705, 735, 746 looms, number of 12, 259, 266 power-loom weaving 255, 260, 262, 263, 271, 272 silk 264 women, role in textile production 264–5, 269–70, 272, 664 see also gender division of labour wool industry 12 workers, number of 12, 266 working conditions 564–5 working hours 261 workshop production of textiles 256, 273, 593 see also Asia 812

Index International Workers’ Association (Denmark) 154, 155 International Working Men’s Association 759 Ireland: Catholicism 654 migrant workers 535, 536, 539, 540, 542, 654, 663, 664, 665, 666, 675, 676, 714 Protestantism 654 Istanbul (Ottoman Empire) 478, 488 Italy 276 artificial fibres 299–300 artisan production of textiles 281 carding 279, 280–1, 291 Catholicism 284 child labour 281, 284, 288–9, 295 clothing industry 301–3 comparative manufacturing costs 783 cotton industry 13, 284, 287–9, 296, 299 de-industrialization 277 decline of textile industry 277–8, 302 domestic production of textiles 279, 281, 282, 288 exports 278, 286–7, 297, 299, 301 factory production of textiles 284–5, 287–92, 297, 298 fashion industry 301–3 fines imposed on workers 285 finishing processes 281, 283 flax industry 283, 297 friendly societies 293 fustian industry 283, 287 gender division of labour 279, 281–2, 288–9, 295, 300 guilds 277, 284 hat production 279 health of textile workers 288–9 hemp industry 282, 283, 297 history of textile production 275, 277–303 pre-industrial period 277–85 development of the factory system 284–5 early phases of industrialization 285–95 late nineteenth and twentieth centuries 295–303 interwar period 298–300 housing of textile workers 294 impact of First World War 298

industrialization of textile production 285–95 jute industry 297 knitting industry 280 linen industry 283, 297 living conditions 294 manufactory production of textiles 281, 282 men, role in textile production 281–2, 289, 295 see also gender division of labour merchants’ role in textile industry 282, 283, 285, 286–7, 289 night work 296 paternalism 292–5, 298, 301 peasant labour 278, 282, 307 pre-industrial textile production 277–85 production statistics 279, 280, 296, 297, 299 protectionism 291 proto-industrialization 278 recycled textiles 279 rope industry 297 rural textile production 289 spinning 13 mule spinning 291 number of spindles 13, 299 self-actor spinning machines 290 wool 290–1 strikes 298 wages 281–2, 296, 300, 783 weaving 281–2 handloom weavers 297 number of looms 13, 296, 297 number of weavers 283 power-loom weaving 295, 297 wool 197 women, role in textile production 279, 281, 282, 288–9, 295, 300 wool industry 13, 279–82, 289–92, 297, 299 number of workers 292, 296, 299 spinning 290–1 weaving 197 wool-combers 282, 284 workers, number of 13, 279, 283, 292, 295, 296, 297, 299, 300 working conditions 284–5, 288–9 working hours 284, 296 workshop production of textiles 288 813

The Ashgate Companion to the History of Textile Workers labour relations 316–17, 607, 609 see also collective bargaining; strikes; trade unions labour shortage 327–8 manufactory production of textiles 309, 691 men, role in textile production 310, 319 see also gender division of labour merchants’ role in textile industry 307 migrant workers 317, 657, 674, 676 multinational textile companies 615 night work 319 restrictions on 117, 311, 328, 718, 723 paternalism 321, 603–4, 605, 611, 668–9, 674, 689, 718 piecers 688 pre-industrial textile production 305, 307–10 production statistics 12, 312, 325 productivity 327–8 protectionism 637 proto-industrialization 309 recruitment of labour 317–21, 687, 688 rural textile production 308–9, 311 silk industry 113, 115, 307, 308, 312, 312, 313, 315, 323, 571, 585 spinning 111, 586, 731 cotton 314, 315, 327 number of spindles 12 number of spinners 318 ring spinning 314, 731 yarn production 12 steam power 314 strikes 321 see also labour relations; trade unions sumptuary laws 307 textile exports 312–14, 313, 315–16, 322, 323 to China 111 textile production in China 116–17, 131 trade unions 316, 327, 328, 329, 758 see also labour relations; strikes wages 320, 699, 717, 783 weaving: automatic looms 314, 327 cotton 315, 327 number of looms 12 number of weavers 318 power-loom weaving 314

ITGLWF (International Textile, Garment and Leather Workers’ Federation) 616 Ivanovo-Voznesensk (Russia/USSR) 422, 424, 425, 427, 428, 430, 431, 432–3, 434, 437, 440, 441, 444, 445, 448, 741, 768 IWW (Industrial Workers of the World, USA) 549–50 Jacquard looms 55, 59, 314, 464, 557 Japan 306 artificial fibres 311, 324, 325, 574, 614 artisan production of textiles 308, 309 bureaucratization of work 611 carding 309, 327 child labour 319 collective bargaining 609, 762 see also labour relations; strikes; trade unions comparative manufacturing costs 783 cotton industry 12, 307, 312–13, 312, 313, 315, 322, 323, 324, 327, 572, 573 de-skilling of textile work 314 decline of textile industry 323–5 education of textile workers 331 electrical power 314 employers’ associations 316, 607, 609 employment contracts 320 exports 307, 312–13, 313, 315–16, 322, 323–4, 323, 325, 637 factory production of textiles 305, 310–11, 314–15, 315, 317–20, 318, 328–9 finishing processes 308, 318, 319, 326 gender division of labour 310, 311, 318, 319, 320–1, 329–30, 699, 704, 705 guilds 308, 309, 316 health of textile workers 320, 321 history of textile production 305, 307–32, 568–9, 593, 630, 637 1650–1850s 305, 307–10 1850–1937 310–21 1945–2000 322–31 firms and production 326–8 housing of textile workers 320–1, 329, 330, 331, 603–4, 689, 691, 718 industrialization of textile production 310, 311, 312, 314–17 knitting industry 318 814

Index La Constància (Spain) 467–8, 469 La Industrial (Uruguay) 518–19, 520, 521, 526, 672 labour codes, Mexico 349 labour costs, reduction of 777–85 Labour Party: Great Britain 243 Habsburg Monarchy 60 labour relations 597–619, 693–4, 787–90, 788, 792 Argentina 24, 31–2, 36–7, 40–1, 611–12 Australia 788 Austria 70 Belgium 788 Canada 788 China 128–30, 132–3, 135–6, 618–19, 721, 768, 771, 788 Denmark 154–69, 609, 612 Egypt 186–92, 193, 196, 604, 607–8, 788–9, 788 and employers’ associations 605–10 France 788 and gender 719–22 Germany 227–9, 788 Great Britain 237, 238, 240–5, 246, 247–9, 250–1, 252, 603, 607, 608–9, 788, 789 Habsburg Monarchy 57–8, 60 India 268–70, 271, 788 industrial textile production 601–9 Japan 316–17, 607, 609 Mexico 342, 344–56, 358–61, 607, 608, 788 and multinational textile companies 613–16 Netherlands 384, 386–9, 391–2, 607 Ottoman Empire 494–5 Pakistan 788 Poland 405–6, 411, 788 Russia/USSR 435–9, 439–41, 604–5, 608, 618, 788–9, 788 Spain 457–8, 461–71, 475–6, 475, 603, 607, 788 Turkey 504–5 Uruguay 521–2, 524–6, 612 USA 534–5, 536–8, 540–1, 542–5, 546, 549–50, 550–1, 553–6, 602, 788, 789–90 see also collective bargaining; strikes; trade unions

women, role in textile production 310, 311, 318, 319, 320–1, 329–30, 699, 704, 705 wool industry 12, 311, 323 workers, number of 12, 315, 318, 323 working conditions 311, 319–20 working hours 319–20, 328–9, 692 workshop production of textiles 309, 315, 315 see also Asia jennies see spinning jennies Jews: Czechoslovakia 61, 673, 736 Egypt 174, 665, 740 Germany 63 Netherlands 375, 653, 740 Ottoman Empire 488, 493, 665, 669 Poland 412, 671 Turkey 503, 740 USA 671 Jiangnan (China) 122 JSA (Japan Spinners’ Association) 324, 326–7, 607 jute industry: Czechoslovakia 61 Egypt 178 India 267, 270 Italy 297 Poland 404, 406, 407 Kafr al-Dawwar (Egypt) 172, 185, 191, 192, 193, 196 Kaufsystem 49, 282, 451, 680, 685, 693, 701 Khurunfish (Cairo, Egypt) 177–8, 667, 684, 691, 695 Kingdom of Poland 401, 402 knitting industry: Argentina 23, 27–8, 30, 32, 34 Habsburg Monarchy 59, 61, 69 Italy 280 Japan 318 Netherlands 368n12 Spain 454, 545 Turkey 500, 501 Uruguay 514 Koninklijke Nederlandse Textiel Unie (KNTU, Netherlands) 394 Korean War (1950–53) 322 La Aurora (Uruguay) 516, 518 815

The Ashgate Companion to the History of Textile Workers Egypt 607 Great Britain 606, 608 Mexico 345, 607 Netherlands 386, 387–8, 607 Spain 465, 466, 607 Uruguay 525, 526 USA 544, 546, 551, 602 see also labour relations; strikes; trade unions Łodz (Poland) 398, 401, 403, 405–6, 407–11, 412, 416–18, 420, 671 Lombardo, Vicente 354, 355 Lombardy (Habsburg Monarchy) 43, 44, 45, 52 looms: automatic 246, 641, 691 Denmark 151 Great Britain 246, 247, 250 Japan 314, 327 Mexico 343 Netherlands 382 Russia/USSR 428 Uruguay 519–20 USA 553 frame looms 262 iron gear looms 117–18 Jacquard looms 55, 59, 314, 464, 557 number of looms; Argentina 12, 38 Australia 12 Austria 12 Belgium 12 Brazil 12 China 12, 131 Czechoslovakia 12 Denmark 12 Egypt 12, 184 France 13 Germany 13, 203, 205, 208, 210, 211, 217 Great Britain 13 India 12, 259, 266 Italy 13, 296, 297 Japan 12 Mexico 12, 336, 337, 339, 343, 569 Netherlands 13, 381, 383, 384 Ottoman Empire 490 Poland 13 Russia/USSR 13, 426 Spain 13, 451, 452, 453, 473

labour shortage 630–1 Japan 327–8 Netherlands 390–1 lace-making, Habsburg Monarchy 55, 59 lanari (Italy) 284 Lancashire (Great Britain): cotton industry 233, 235–8, 240, 246–8, 250–1, 571, 573, 609, 679, 681, 682, 686, 690, 695, 697, 699, 715, 717, 726, 730–1, 732, 733, 734, 736, 737, 741, 744, 747, 755, 789 trade unionism 237–8, 240–1, 242–3, 249, 251, 252, 741, 755, 757, 761, 786–7 migrant workers in USA 536, 537, 538, 539, 540, 541, 542, 543, 544–5, 547, 551, 687 Preston studies 663–4 Landelijke Federatie van Textielarbeiders (Netherlands) 387 Lanoficio Rossi (Italy) 275, 291–2, 294–5, 296 see also Rossi family Latin America 568, 637–8 see also Argentina; Brazil; Mexico; Uruguay Lawrence (USA) 532, 535, 539, 540, 549–50 Leiden (Netherlands) 364, 365–6, 367, 368, 369, 371, 372, 376, 390–1, 652–3, 784 Leitenberger, Johann Josef 51, 53, 566 linen industry: Germany 208 Habsburg Monarchy 46, 51 Italy 283, 297 Netherlands 366 Russia/USSR 423, 424, 425, 442, 447 Spain 451, 454 see also flax industry Linussio, Giacomo 283 Linz (Habsburg Monarchy) 44, 57, 655 Linzer Wollzeugfabrik 50, 54 Lister, Samuel 243, 244 living conditions 688–9, 718 Brazil 97 China 124 Germany 227 Habsburg Monarchy 57 Italy 294 Netherlands 376–7 Russia/USSR 433 lockouts 606, 607, 608, 609, 786 Denmark 150, 156–7, 169, 609 816

Index maquiladora programme (Mexico) 358–9 Marshall Plan 67, 393 Marx, Karl 601, 759 Marzotto group (Italy) 301–2 masculinization of the workforce 124, 711, 713 maternity rights 717, 718, 723 Bangladesh 615 China 128, 135 India 270 Mawdsley, James 241, 243 maximum prices policy see price control mechanization see industrialization of textile production men, role in textile production 4, 584, 589–90, 703–24 Brazil 88, 96 China 103, 120–1, 714 Denmark 164–8, 588 Egypt 183, 714 Germany 199, 213, 223 Great Britain 236, 237, 239, 241 Habsburg Monarchy 48, 56 India 265, 272, 664, 705 Italy 281–2, 289, 295 Japan 310, 319 Mexico 337, 340, 342, 344, 349, 351, 353, 590, 705 Netherlands 374–5, 374, 376, 377, 378, 380, 382, 391, 589 Ottoman Empire 479–80, 491–3 Poland 403 Russia/USSR 433, 443, 445 Spain 451, 453, 456, 458–61, 460, 464, 465, 466–7, 588–9 Turkey 503 Uruguay 524, 524, 526, 527 wage differentials 715–17 see also gender division of labour Mendels, Franklin 578, 579 mercaders (Spain) 452 merchants’ associations, China 105 merchants, role in textile industry 580–1, 591–3, 594, 598–600, 622, 626–7, 628, 631, 637, 685, 689, 690, 735, 751, 785 China 105, 108, 109, 110, 111, 118, 629–30, 727, 734 Denmark 144 Egypt 173, 182 Germany 206

Uruguay 519 USA 12 vertical 86, 87 see also handloom weaving; powerloom weaving; weaving Love, J. Spencer 553 Lowell (USA) 532, 533, 534, 539, 540, 547–8, 612–13, 665–6, 668, 671, 757 Lower Austria 44, 54, 54, 55 Lucassen, Leo 652–3 Luddism see machine-smashing luxury products 56, 700, 791 ban on production in Brazil 78, 91–2, 94, 100 Egypt 180 Ottoman Empire 479, 735 see also fine fabrics Macara, Charles 243 machine-smashing: Great Britain 238, 239, 726 Habsburg Monarchy 58 Ottoman Empire 495 machinery see textile machinery McKinley Tariff 243 Mahalla al-Kubra (Egypt) 172, 183, 186–7, 189, 193, 196 man-made fibres see artificial fibres Manchester Factory, Copenhagen 144, 145 Manchester (Great Britain) see Lancashire Manningham mill strike (1890–91) 243, 244, 245 Manual Trades Workers Union (Egypt) 184, 185 manufactory production of textiles 593, 626, 627–32, 633, 636, 638, 640, 681, 684, 689, 691, 692, 696, 779 Brazil 95, 693, 696 China 119, 689 Denmark 143–5, 753 Germany 199, 221, 222, 224 Habsburg Monarchy 49–50, 51, 52, 53, 54, 56, 594, 654, 752 Italy 281, 282 Japan 309, 691 Netherlands 371 Poland 400, 402 Russia/USSR 423 see also proto-industrialization 817

The Ashgate Companion to the History of Textile Workers mature labour regime, the good years 1927–46 353–6 mature labour regime, the bad years 1946–76 356–8 globalization 1976–358–61 housing of textile workers 341, 342, 350, 358, 668, 689 hydroelectric power 343 impact of British imports 569–70 imports 333 indigenous textile production 333, 335, 604, 660, 661, 727, 729 industrialization of textile production 338–46 labour codes 349 labour relations 342, 344–56, 358–61, 607, 608, 788 lockouts 345, 607 see also strikes; trade unions men, role in textile production 337, 340, 342, 344, 349, 351, 353, 590, 705 see also gender division of labour merchants’ role in textile industry 333 migrant workers 343, 668, 670–1 network migration 670 obrajes (workshops) 333, 335–7, 591, 604, 660, 661 paternalism 341–2, 689 production statistics 12, 360 protectionism 638 putting-out system 600 slave labour 337, 604, 660, 661, 686 spinning: number of spindles 12, 339, 340, 343, 569 number of spinners 340 yarn production 12 strikes 342, 344–5, 346–7, 348, 350 see also labour relations; trade unions trade unions 342, 344–5, 346–7, 349–51, 353–6, 359, 360, 757, 765 see also labour relations; strikes transport developments 343 wages 340, 351, 357, 360, 783 water power 339, 341 weaving 336, 337, 733–4 automatic looms 343 number of looms 12, 336, 337, 339, 343, 569 number of weavers 340

Great Britain 606 Habsburg Monarchy 46, 48–9 India 261–2, 263, 274 Italy 282, 283, 285, 286–7, 289 Japan 307 Mexico 333 Netherlands 363, 368, 371–2, 728, 784 Ottoman Empire 483–4, 486, 489–91, 493–4, 665, 742 Poland 399, 400 Spain 451–2, 752 Turkey 508 Methuen Treaty 80–1 METI (Ministry of Economy, Trade and Industry, Japan) 325 Mexico 334 absentee ownership 669 artificial fibres 357–8, 729 artisan production of textiles 333, 335, 338 carding 335, 336 Catholicism 341, 351, 689, 740 child labour 340, 349 collective contracts 347 comparative manufacturing costs 783 cotton industry 12, 338–41, 339, 343–61 decline of textile industry 356–61, 569–70 domestic production of textiles 333, 335, 591 employers’ associations 607 employment contracts 353 encomienda system 337, 729 exports 358, 359 factory production of textiles 338–9, 339, 340–2, 343–4, 349, 358, 360, 569–70, 590 fines imposed on workers 342, 345, 350, 696 foreign investment 343–4 gender division of labour 337, 340, 342, 344, 347–8, 349, 351, 353, 359, 590, 705 globalization 358–61 guilds 338 health of textile workers 336–7 history of textile production 333, 335–61 industrialization 1821–1910 338–46 textile workers and the Mexican revolution 1910–1927 346–53 818

Index USA 535, 536, 537, 538, 539, 540, 541, 542–3, 545, 547, 549, 556, 665–6, 671, 687 see also network migration; race military consumption of textiles 82, 98, 412, 423, 538, 661 mills see factory production of textiles Minas Gerais (Brazil) 81, 87, 89, 90, 92, 94, 95, 96, 658n44, 660 Misr Spinning and Weaving Company (Egypt) 185–7, 189, 190, 196, 690 MITI (Ministry of International Trade and Industry, Japan) 322, 324 modernization of textile production, Great Britain 250–1 mohair industry 566 Ottoman Empire 483, 492 Moldavia (Habsburg Monarchy) 45 Monschau (Germany) 200, 212, 212, 227 Morones, Luis N. 350, 352, 354, 355, 356 Moscow (Russia) 422, 425, 427, 428, 430, 431, 432, 434 see also Central Industrial Region (Russia) mule spinning 567, 588, 679, 688, 691, 695, 697, 730, 756 Great Britain 238, 239, 241, 244, 712, 730, 732, 733, 755, 787 Italy 291 Netherlands 381, 732 Russia/USSR 429, 433, 434, 436 Spain 456, 457, 458, 459, 464, 735 USA 536, 537–8, 539, 540, 541, 542–3, 544–5, 731, 732 see also Crompton’s mule; spinning multinational textile companies 613–6 multiple loom system 691 Denmark 148, 151–2, 158, 160–2, 163, 612, 780 Netherlands 382 see also Bedaux system Muntanya (Spain) 450, 462, 463, 463, 465 museums 73 Muslims: Egypt 173, 174, 186, 189, 665, 740 Ottoman Empire 492, 493, 590, 665, 740 Turkey 503, 504, 760

women, role in textile production 337, 340, 342, 344, 347–8, 349, 353, 359, 590 see also gender division of labour wool industry 12 workers, number of 12, 339, 339, 340, 343, 349, 353, 357, 357 working conditions 336–7, 345–6 workshop production of textiles 333, 335–7, 336, 338, 341, 344, 591, 660, 729 see also Latin America MFA (Multi-Fibre Arrangement) 325, 393 migrant workers 645, 647–77, 742–3 Americas 658–62, 676 Argentina 24, 665, 687 Asia 656–7 Austria 68 Brazil 96, 665, 667, 687 Britain 251, 663 Cape Verde Islands 657–8 China 138, 656–7, 673–4, 676, 677, 743 Egypt 667–8 Europe 650–5, 663–4, 675–6 France 663, 673, 677 Germany 222–3, 663, 673, 677 Great Britain 251, 663–4 Habsburg Monarchy 52, 654–5 Huguenots 202, 653–4 India 261, 264, 265, 267–8, 269, 270, 593, 656, 664, 673, 676 internal migration 663–5, 672–3, 686–7 internal/international migration 665–8 Ireland 535, 536, 539, 540, 542, 654, 663, 664, 665, 666, 675, 676, 714 Japan 317, 657, 674, 676 Mexico 343, 668, 670–1 modern period 662–74 Netherlands 363, 375–6, 383–4, 652–3 guest workers 392, 677, 743 Ottoman Empire 494, 665, 676, 695–6 paternalistic migration 668–9, 670–1 Poland 402–3 Russia/USSR 431, 438, 447, 687 sixteenth to eighteenth century 650–62 Spain 461, 470, 473–4 Turkey, guest workers 500, 508, 675, 677, 743 twentieth century to present 670–4 Uruguay 526, 665, 672

NAFTA (North American Free Trade Agreement) 359, 573, 576, 671 Naples (Italy) 43 819

The Ashgate Companion to the History of Textile Workers Nasser, Gamal Abdel 174, 190, 191 National Federation of Textile Workers (Netherlands) 387 National Union of Textile Workers (Great Britain) 249 see also General Union of Textile Workers National Workers’ Union (Uruguay) 525 nationalism 760 and recruitment of labour 673–4 Nationalist Party (Egypt) 184, 185 nationalization of the textile industry: China 746 Czechoslovakia 63 Egypt 174, 192 Poland 412 Turkey 497, 499 native workers see indigenous workers NCWS (National Committee of Workers and Students, Egypt) 188 Nederlandsche Handel Maatschappij (NHM) 378, 379 neringen (Netherlands) 371 Netherlands 364 artificial fibres 614 artisan production of textiles 728 Austrian Netherlands 43, 45–6, 52 bureaucratization of work 611 Catholicism 385, 387, 692, 740, 760 child labour 376, 377, 380, 382, 385 cotton industry 13, 366–7, 369, 370 spinning 378–9, 381, 381, 382, 383, 383 weaving 381, 381, 382, 383 de-skilling of textile work 739–40 decline of textile industry 389–95, 769–70 drapers 371 employers’ associations 387, 388, 607 exports 365, 378, 393 factory production of textiles 379–80, 380, 384 finishing processes 391 fustian industry 366, 378 gender division of labour 370, 374–5, 374, 376, 377, 378, 379, 380, 382, 385, 390, 391, 589, 704, 705 guilds 366, 366, 370–2, 373, 375, 377–8, 594 history of textile production 363, 365–95, 566

1650–1810 365–78 1810–1950 378–89 1950–2000 389–95 housing of textile workers 689 industrialization of textile production 378–89 knitting industry 368n12 labour costs 784 labour relations 384, 386–9, 391–2, 607 lockouts 386, 387–8, 607 see also strikes; trade unions labour shortage 390–1 linen industry 366 living conditions 376–7 manufactory production of textiles 371 men, role in textile production 374–5, 374, 376, 377, 378, 380, 382, 391, 589 see also gender division of labour merchants’ role in textile industry 363, 368, 371–2, 728, 784 migrant workers 363, 375–6, 383–4, 392, 652–3, 677, 743 network migration 672 noise 385 paternalism 603 peasant labour 383 pre-industrial textile production 365–78 production statistics 367 productivity 381, 381 Protestantism 385, 387, 653, 740, 760 putting-out system 371, 600 spinning: cotton 378–9, 381, 381, 382, 383, 383 mule spinning 381, 732 number of spindles 13, 381, 383, 384 number of spinners 369, 374, 377, 380, 383 ring spinning 380, 382, 391 self-actor spinning machines 380, 381, 382 wool 370, 374–5, 374, 379, 380, 382, 383 yarn production 13 steam power 379, 381 strikes 386, 386, 388–9 see also labour relations; trade unions textile imports from India 563 trade unions 386–9, 391–2, 757, 760 see also labour relations; strikes Turkish textile businesses 508 820

Index wages 382, 385–6, 392, 784 weaving: automatic looms 382 cotton 381, 381, 382, 383 handloom weavers 381, 381 multiple loom system 382 number of looms 13, 381, 383, 384 number of weavers 368, 369, 377, 380, 383 power-loom weaving 381, 384–5 wool 376, 379, 380, 383 women, role in textile production 370, 374–5, 374, 376, 377, 378, 379, 380, 382, 385, 390, 391, 704, 705 wool industry 13, 365–6, 366, 367–8, 369, 370–2, 373, 374–8, 377 number of workers 368, 369, 377, 380, 383 spinning 370, 374–5, 374, 379, 380, 382, 383 weaving 376, 379, 380, 383 wool-combers 375–6, 728 workers, number of 13, 368, 369, 377, 378, 380, 381, 383 working conditions 376–7, 383–6, 392 working hours 385, 692, 693 workshop production of textiles 371 Netherlands Textile Employers’ Federation 607 network migration 663, 669–70, 672 Argentina 672 Flemish workers 653 India 656 Mexico 670 Netherlands 672 Ottoman Empire 669–70 Spain 672 see also migrant workers New Deal (USA) 554 New Economic Policy (NEP, USSR) 442, 443, 767 New England (USA) 533–5, 546, 547, 549, 568, 783 New Lanark mills (Scotland) 603, 668 NHM (Netherlands Trade Company) 378, 379 NICs (newly industrializing countries) 623, 641–2, 643, 644 night work 692, 700 Brazil 692

China 124, 125–6, 692 Great Britain 696 Italy 296 Japan 319 restrictions on: Argentina 37 Brazil 96 Germany 224 Great Britain 718, 723 Habsburg Monarchy 406 India 270, 718, 723 Japan 117, 311, 328, 718, 723 Poland 405, 406, 411 Russia/USSR 434, 718 noise: Germany 227 Netherlands 385 non-voluntary migration see slave labour Norris, Thomas 536–7 North American Free Trade Agreement (NAFTA) 359, 573, 576, 671 North-Brabant (Netherlands) 369, 370, 566 Number Ones (China) 123, 124, 125, 129, 674, 688, 721, 738 nylon see artificial fibres obradores (artisan production of textiles, Mexico) 333, 335, 338, 591, 729 obrajes (workshops): Argentina 21, 661, 684 Mexico 333, 335–7, 591, 604, 660, 661 Oldham Cotton Masters’ Association (Great Britain) 606 Operation Dixie (USA) 554–6 Organización Sindical Española (OSE, Spain) 471, 476 Oriental Trade Company (Habsburg Monarchy) 50, 626–7 Orizaba (Mexico) 334, 354 Ósaka (Japan) 306, 307, 572 OSE (Organizacion Sindical Espanola, Spain) 471, 476 Ottoman Empire 478 carpet production 482–3, 486, 489, 491, 493, 495, 583, 589, 623, 626, 665, 732, 735, 742 Catholicism 590 child labour 480, 491, 731–2 cotton industry 483–4 cotton cultivation 639 821

The Ashgate Companion to the History of Textile Workers decline of textile industry 479, 480–1, 569, 582 ethnicity of textile workers 493, 655–6, 669–70, 676 exports 479, 480, 481, 482, 483, 582, 583, 639 factory production of textiles 480, 485, 486, 491 gender division of labour 479–80, 491–3, 494, 589, 714 guilds 479–80, 486–9, 491–2, 584–5, 656, 752 history of textile production 477, 479–95, 566, 569, 582, 583, 742 decline of manufacturing? 479–85 organization of production 486–95 housing of textile workers 492, 494, 503 imports 480 labour relations 494–5 luxury products 479, 735 machine-smashing 495 men, role in textile production 479–80, 491–3 see also gender division of labour merchants’ role in textile industry 483–4, 486, 489–91, 493–4, 665, 742 migrant workers 494, 665, 676, 695–6 network migration 669–70 Muslims 492, 493, 590, 665, 740 putting-out system 489–90 recruitment of labour 493–4, 665 silk industry 484–5, 492 slave labour 586 spinning 480, 731–2 strikes 495 textile imports from India 563 trade with Habsburg Monarchy 50 trade unions 495 weaving 484, 485, 494, 735 handloom weavers 485 number of looms 490 women, role in textile production 480, 491–3, 494, 589, 714 wool industry 483 workers, number of 477, 485 workshop production of textiles 485, 487, 490–1, 492, 494, 497 Owen, Robert 603, 668 oyabun-kobun system (Japan) 310, 669

Paisley (Scotland) 603, 615, 664 Pakistan, labour relations 788 paternalism 603–4, 718 Brazil 97, 668 Great Britain 244 Italy 292–5, 298, 301 Japan 321, 603–4, 605, 611, 668–9, 674, 689, 718 Mexico 341–2, 689 Netherlands 603 paternalistic migration 668–9, 670–1 Scotland 603 Spain 603, 668, 718 USA 603 Paterson (USA) 532, 112n29, 535, 545, 550, 553, 666, 671 Patria-Piering 69 peasant labour: Denmark 141, 143 Egypt 177, 181, 186, 187, 188 Great Britain 235 Habsburg Monarchy 56 India 257, 267, 270 Italy 278, 282, 307 Netherlands 383 Poland 400, 402, 403 Russia/USSR 423–6, 431–3, 437–8 Spain 461 see also domestic production of textiles pelaires (Spain) 452–3 People’s Republic of China see China Permanent Court of Arbitration (Denmark) 160, 162 Perón, Juan Domingo, and textile industry 34–7, 722 Perrotine machine 427 Persian Gulf, textile imports from India 563 Petrograd see St Petersburg piece-rates 695–6, 697, 780 China 120, 127, 133 Denmark 157–8, 159–61, 166 Egypt 177 Great Britain 236, 244–5 Russia/USSR 434 piecers 691 Great Britain 237, 240, 695, 732 Japan 688 Spain 458, 588, 687n23 Pla (Spain) 450, 462, 463, 465, 468 822

Index yarn production 13, 413, 418–19, 419 strikes 405–6, 411, 416–18 see also labour relations trade unions 405–6, 411, 416–18 see also labour relations wages 405 weaving 406, 407 number of looms 13 number of weavers 403 women, role in textile production 403, 404–5, 406, 409, 409, 410, 411, 416, 417 wool industry 13, 399–400, 401, 402, 407, 408, 413, 414–15, 414, 418–19 419 workers, number of 13, 403, 404, 407, 408, 415–16, 415, 420 working conditions 406 working hours 404–5, 406, 411 workshop production of textiles 399, 400 political ideologies 745–6 polyester see artificial fibres Pombal, Marques de 81, 90, 91 Portugal 80–1, 651, 652 Alvara (Law) 78, 91–2 Napoleonic invasion of 93 sumptuary laws 90 textile imports from India 563 see also Atlantic Islands Posthumus, N.W. 365, 367, 368, 376 Potato Famine 1845 (Ireland) 663, 664, 665, 666 power-loom weaving 54, 586, 691, 712, 732 China 117, 119, 127, 131, 137, 734 Denmark 146, 147, 151–2, 737 Egypt 179, 184 Germany 210, 217–18 Great Britain 236, 238, 239, 586 Habsburg Monarchy 54, 55, 739 India 255, 260, 262, 263, 271, 272 Italy 295, 297 Japan 314 Netherlands 381, 384–5 Russia/USSR 433 Spain 456, 459, 463–4, 586 USA 533 see also looms; weaving Prato (Italy) 276, 279, 290, 299

PNR (Partito Nacional Revolucianario, Mexico) 354, 355 Poland 45, 398 child labour 404–5, 406, 410, 410, 411 cotton industry 13, 400, 401, 402, 407, 408, 413, 414, 414, 418–19, 419 decline of textile industry 407, 418 domestic production of textiles 399, 400, 403 employment contracts 405 exports 400, 401, 409, 413 factory production of textiles 401, 402–4, 405, 406, 407, 407, 409, 412 flax industry 399, 400, 404, 406, 407, 413, 414, 418, 419 gender division of labour 403, 404–5, 406, 409, 409, 410, 411, 416, 417 guilds 399 health of textile workers 405, 420 hemp industry 399, 400, 404 history of textile production 397, 399–420 1650–1815 399–400 1815–1918 401–7 1918–1945 407–11 1945–1990 412 1990–2000 418–20 impact of First World War 407 industrialization of textile production 400, 402 jute industry 404, 406, 407 labour relations 405–6, 411, 788 see also strikes; trade unions manufactory production of textiles 400, 402 men, role in textile production 403 see also gender division of labour merchants’ role in textile industry 399, 400 migrant workers 402–3 nationalization of the textile industry 412 night work, restrictions on 405, 406, 411 peasant labour 400, 402, 403 production statistics 413, 414, 414, 418–19, 419 silk industry 4–7 social democracy 405 spinning 407 number of spindles 13 823

The Ashgate Companion to the History of Textile Workers protectionism 5, 70, 572, 573, 632, 638, 641, 677, 770, 789 against British exports 568 Argentina 17, 24 Austria 65 Brazil 77, 94, 638 China 637 Denmark 150, 152 EC 393 Europe 565, 575 Great Britain 565 Habsburg Monarchy 53 Italy 291 Japan 637 Mexico 638 Spain 451, 466, 469, 472 Uruguay 514, 517, 527–9 USA 533, 548, 548, 549 Protestantism: Ireland 654 Netherlands 385, 387, 653, 740, 760 trade unions 385, 387 USA 534, 535, 666, 756 proto-industrialization 4, 577–81, 594–5, 622, 624, 627, 647, 774–6 Americas 658–9 Brazil 78, 81–4, 100 Germany 220 Habsburg Monarchy 48–52 Italy 278 Japan 309 USA 659 see also manufactory production of textiles Prussia (Germany) 200, 203, 204, 215, 217, 217, 223 putting-out system 580–1, 595, 600, 626, 627, 628–30, 685, 701, 776 Brazil 600 China 108, 117–18, 580, 592, 600, 685 Denmark 144 Egypt 178, 181 Germany 221, 227 Habsburg Monarchy 48–9, 56, 61, 594 Mexico 600 Netherlands 371, 600 Ottoman Empire 489–90 Russia/USSR 426, 427–8, 600 Spain 600 Turkey 600

pre-industrial textile production 3–4, 622, 704, 707–8, 711, 714, 726, 750, 751, 774 Brazil 77, 84 Denmark 141, 143, 154, 164 gender division of labour 707–8 Italy 277–85 Japan 305, 307–10 Netherlands 365–78 Spain 449, 451, 452, 453 Preston (Lancashire, Great Britain) 232, 663–4 price control, Argentina 34–5 privatization of textile industry: Russia/USSR 448 Turkey 499 production chains 621–2, 638, 645 production processes: Argentina 29–30 cotton 29–30 Denmark 151–2, 165–6 production statistics: Argentina 12, 27, 33, 34, 38, 39, 573 Australia 12 Austria 12 Belgium 12 Brazil 12, 87, 94–5, 97 China 12, 116, 119, 131, 132, 133, 134, 136–7 Czechoslovakia 12 Denmark 12, 145, 152 Egypt 12, 184–5, 194 Germany 13, 206, 208, 209, 210, 211–14, 220, 221 India 12, 266 Italy 13, 279, 280, 296, 297, 299 Japan 12, 312, 325 Mexico 12, 360 Netherlands 13, 367 Poland 13, 413, 414, 414, 418–19, 419 Russia/USSR 13, 427, 429, 430, 431, 442, 446, 447 Spain 13, 452, 453, 453 Turkey 12 Uruguay 12 USA 12 productivity: Denmark 150–2 Germany 210 Japan 327–8 Netherlands 381, 381 824

Index Roman Catholicism see Catholicism Romania, transfer of textile production to 68–9 Rooms-Katholieke Fabrieksarbeidersbond (Netherlands) 387 rope industry, Italy 297 Rossi family (Italy) 279, 290, 291 Alessandro Rossi 291–5 see also Lanoficio Rossi Royal Factories (Spain) 451, 652 Ruben, I.H. 147 rural textile production 708 China 107–9, 117–19, 127, 131, 133, 137 Denmark 141, 143 Germany 219, 221, 222 Habsburg Monarchy 47–8, 56 India 264 Italy 289 Japan 308–9, 311 Spain 449 Russia/USSR 422 bureaucratization of work 612 carding 445 child labour 434, 604 cloth production 430, 442, 446 cotton industry 13, 423–48, 442, 447 cotton cultivation 423, 444 cotton printing 424, 425, 427, 429, 434, 436 spinning 427, 433, 436 decline of textile industry 448–9 domestic production of textiles 426 dyeing industry 424 exports 429, 444 factory production of textiles 424–6, 427–8, 429–32, 430, 448, 618 fines imposed on workers 434, 436, 696 finishing processes 424, 425–6, 427 flax industry 431 gender division of labour 433–4, 438–40, 443, 445, 446 history of textile production 421, 423–41, 441–8, 448, 566 from beginnings to 1842 421, 423–6 1842–1914 427–39 war and revolutionary period 439–41 Soviet era 441–8 housing of textile workers 433, 689 impact of First World War 439–41

race, and textile workers 736–7 see also migrant workers ‘race to the bottom’ 774, 777–85, 787, 788, 790 USA 556, 713 raw cotton consumption (by country) 12, 13 raw wool consumption (by country) 12, 13 rayon see artificial fibres recruitment of labour 687–8 China 123–5, 604, 674, 688, 738 Habsburg Monarchy 52 India 267–8, 656 Japan 317–21, 687, 688 and nationalism 673–4 Ottoman Empire 493–4, 665 ‘paternalistic migration’ 668–9, 670–1 role of families in 710 China 124 USA 666 recycled textiles, Italy 279 reeling (of silk) 112, 113–15, 122–3, 485 regolamenti (Italy) 285 religion, and textile workers 653–4, 740–1 relocation see migrant workers relocation of textile industry 781–4, 786 Rentengrundherrschaft 45 Rhineland (Germany) 200, 201, 205, 205, 206, 206, 214, 219, 220, 222, 224, 229 Rhode Island (USA) 533, 547 ribbon industry 55 ring spinning 641, 712, 730, 732, 733, 779 China 734 Denmark 148, 152 Germany 217, 217 Great Britain 217, 238, 250 Japan 314, 731 Netherlands 380, 382, 391 Russia/USSR 429, 433, 434, 436, 444 Spain 457, 464, 466, 467 USA 217, 543, 545, 552 Rio de Janeiro (Brazil) 76, 81, 87, 92, 93, 94, 96, 98, 667, 730 Rio de la Plata (Argentina) 17, 18, 21 Rio Negro (Brazil) 76, 87, 88, 89, 92, 695 RMMS (Rashtriya Mill Mazdoor Sangh, India) 269, 270, 271, 766, 770 Roberts, Richard 237 Roca-Runciman Agreement 28 Roman Catholic Factory Workers’ Union (Netherlands) 387 825

The Ashgate Companion to the History of Textile Workers working conditions 434–5, 438 working hours 434, 438, 692 workshop production of textiles 684

industrialization of textile production 428–9 labour relations 435–9, 439–41, 604–5, 608, 618, 788–9, 788 see also strikes; trade unions linen industry 423, 424, 425, 442, 447 living conditions 433 manufactory production of textiles 423 men, role in textile production 433, 443, 445 see also gender division of labour migrant workers 431, 438, 447, 687 night work, restrictions on 434, 718 peasant labour 423–6, 431–3, 437–8 piece-rates 434 privatization of textile industry 448 production statistics 427, 429, 430, 431, 442, 446, 447 putting-out system 426, 427–8, 600 serfdom 655 silk industry 442, 447 spinning: cotton 427, 433, 436 mule spinning 429, 433, 434, 436 number of spindles 13, 427, 441 ring spinning 429, 433, 434, 436, 444 self-actor spinning machines 433 yarn production 13, 425, 427, 428, 430 strikes 435–41, 444 see also labour relations; trade unions trade unions 435–9, 440–1, 443, 757, 766–8 see also labour relations; strikes transport developments 428–9 wages 434, 699 weaving 425–6, 427–8, 433, 436, 734, 735 automatic looms 428 handloom weavers 425–6, 427, 428, 429, 684 number of looms 13, 426 number of weavers 428 power-loom weaving 433 women, role in textile production 433–4, 438–9, 439–40, 443, 445, 446 wool industry 13, 431, 442, 447 workers, number of 13, 425, 428, 429–30, 430, 448

St Lambertus (Roman Catholic Trade Union, Netherlands) 387 St Petersburg (Russia) 422, 427, 435–6, 439–40 Salary Councils (Uruguay) 521, 525 Santiago del Estero (Argentina) 18, 21, 22 Sao Paulo (Brazil) 76, 79, 94, 95, 98 sardars (India) 267–8 Saxon Oberlausitz (Germany) 579–80, 589–90 Saxony (Germany) 200, 201, 203, 204, 205, 205, 206, 206, 211, 214, 216–17, 217, 219, 220, 225 Sbabi Spinning and Weaving Company (Egypt) 185 Schio (Italy) 276, 279, 282, 290 Lanoficio Rossi 275, 291–2, 294–5, 296 Schwechater Baumwollmanufaktur (Austria) 50, 54 Scientific Management 42, 152, 443, 547, 598, 609–13 see also Taylorism Scotland 603, 606, 609, 615, 664, 668 SDAP (Sozialdemokratische Arbeiterpartei, Habsburg Monarchy) 60 Second World War: impact on Austria 66–7 impact on Czechoslovakia 63 impact on Egypt 188 impact on Great Britain 250 impact on USA 554–5 sector foreign trade companies (SFTC, Turkey) 508 Sedalana (Argentina) 28, 35 self-actor spinning machines 588, 755 Denmark 146, 741 Germany 217, 217, 218 Great Britain 237, 239, 240, 246 Italy 290 Netherlands 380, 381, 382 Russia/USSR 433 Spain 457, 458, 459, 464, 588 USA 537 self-exploitation 589, 590, 677, 693 Sella family (Italy) 290–1 September Compromise (Denmark) 157, 162, 609 826

Index socialism 758–60 Socialist Algemene Bond van Textielarbeiders De Eendracht (Netherlands) 387 Socialists (Argentina) 31–2, 36 Society of Muslim Brothers (Egypt) 189 Solidarity Trade Union (Poland) 417 South America, textile imports from India 564 South Korea, artificial fibres 574 South-East Asia, textile imports from India 563 Southern Germany 205, 206, 206 Southern Netherlands see Austrian Netherlands Soviet Union see Russia/USSR Sozialdemokratische Arbeiterpartei (SDAP, Habsburg Monarchy) 60 Sozialpartnerschaft 70 Spain 450 artificial fibres 472 artisan production of textiles 603, 728, 752 carding 451, 452, 454 child labour 460 comparative manufacturing costs 783 conquest of Argentina 19–20 cotton industry 13, 453–5, 454, 455, 457, 462–71, 473–4, 473 de-industrialization 449 de-skilling of textile work 463 employers’ associations 607 exports 466, 468 factory production of textiles 451–2, 453–5, 454, 455–6, 457, 462–4, 472 finishing processes 460, 467, 468–9 gender division of labour 451, 453, 456, 458–61, 460, 464, 465, 466–7, 473–5, 588–9 guilds 452–3, 752 hemp industry 454 history of textile production 449, 451–76, 566, 586, 651–2 1650–1814 449, 451–5 industrialization 1814–1885 455–62 offensive against the cotton textile trades 1885–1939 462–71 from dictatorship to democracy 1939–2000 471–6 housing of textile workers 463, 466, 467, 689

serfdom: Habsburg Monarchy 52, 654, 690 Russia/USSR 655 sericulture see silk industry SFTC (sector foreign trade companies, Turkey) 508 Shanghai (China) 104, 113, 117, 119, 120, 121–3, 124–5, 129, 131, 138, 604 Shanghai Cotton Cloth Mill 115–16 Shubra al-Khayma (Egypt) 185, 188–9, 190, 192, 193 Sicily 43 Silesia (Germany/Habsburg Monarchy) 44, 45, 200, 205, 206, 220 Silesian Weavers’ Uprising 57 silk industry 562–3, 571 Argentina 28, 30 China 103, 106, 112–15, 119, 121, 126–7, 128, 132, 134–5, 137, 138, 591–3, 629 Egypt 176, 183 global trade 566 Habsburg Monarchy 46, 51, 59 India 264 Italy 278, 286–7, 295, 296, 299 Japan 113, 115, 585 Ottoman Empire 484–5, 492 Poland 407 Russia/USSR 442, 447 Spain 452, 454 USA 535, 550, 666 Silkeborg (Denmark) 142, 162–3, 612 Sino-Japanese War (1894–5) 116 sisterhoods, China 129, 130, 720 slave labour 95, 563, 564, 587, 627, 637, 650, 675 Americas 658–9 Brazil 79–80, 87–8, 96, 600, 604, 658, 659, 667, 685, 730 Cape Verde Islands 657–8 Egypt 177 Mexico 337, 604, 660, 661, 686 Ottoman Empire 586 USA 534, 659, 660, 666 see also encomienda Slovakia 63–4, 70 social democracy 759–60 Denmark 149–50, 154, 162, 163, 764 Germany 228–9, 745, 760 Poland 405 ‘Social Partnership’ (Austria) 70 827

The Ashgate Companion to the History of Textile Workers wool industry 13, 452, 454 workers, number of 13, 451, 452, 453, 455, 456, 460, 463, 473–4, 473, 474 working hours 460, 467–8, 692 workshop production of textiles 451, 474, 476, 591 spatial division of labour 621–45 1700–1820 627–32 1820/1830–1914/1918 632–40 1918–1973/1980 640–3 before 1650/1700 626–7 spindles, number of: Argentina 12, 26, 27, 38 Australia 12 Austria 12 Belgium 12 Brazil 12 China 12, 113, 116, 133, 573 Czechoslovakia 12, 61 Denmark 12 Egypt 12, 184 France 13 Germany 13, 203, 206, 209, 209, 211, 216–17, 217, 218, 219 Great Britain 13 India 12 Italy 13, 299 Japan 12 Mexico 12, 339, 340, 343, 569 Netherlands 13, 381, 383, 384 Poland 13 Russia/USSR 13, 427, 441 Spain 13, 457, 464, 473 Turkey 12 USA 12, 546 spinners: earnings: Brazil 88 China 108, 109, 127 Egypt 194 Russia/USSR 434 number of: Germany 209, 210, 212, 213, 214 Japan 318 Mexico 340 Netherlands 369, 374, 377, 380, 383 USA 542 spinning 585–6, 628–9, 632, 634, 712, 730–3 Asia 634 China 116–17, 570, 731

industrialization of textile production 455–62 knitting industry 454, 545 labour relations 457–8, 461–71, 475–6, 475, 603, 607, 788 lockouts 465, 466, 607 see also strikes; trade unions linen industry 451, 454 men, role in textile production 451, 453, 456, 458–61, 460, 464, 465, 466–7, 588–9 see also gender division of labour merchants’ role in textile industry 451–2, 752 migrant workers 461, 470, 473–4 network migration 672 paternalism 603, 668, 718 peasant labour 461 piecers 458, 588, 687n23 pre-industrial textile production 449, 451, 452, 453 production statistics 452, 453, 453 protectionism 451, 466, 469, 472 putting-out system 600 rural textile production 449 silk industry 452, 454 spinning 13, 456, 458, 459, 463–4 cotton 454 mule spinning 456, 457, 458, 459, 464, 735 number of spindles 13, 457, 464, 473 ring spinning 457, 464, 466, 467 self-actor spinning machines 457, 458, 459, 464, 588 steam power 456, 462, 463 strikes 466–8, 475–6, 475 see also labour relations; trade unions trade unions 461–2, 465–71, 476, 763–4 see also labour relations; strikes wages 783 weaving 454, 456–7, 459, 464, 735 handloom weavers 451, 456, 458, 462, 591, 728 number of looms 13, 451, 452, 453, 473 power-loom weaving 456, 459, 463–4, 586 women, role in textile production 451, 453, 456, 458–61, 460, 464, 465, 466–7, 473–5, 588–9 828

Index strikes: Argentina 31–2, 32, 36–7, 41 China 128, 136 Denmark 155, 158–9, 161, 162–3, 167–8, 612 Egypt 187, 188, 189, 190–1, 193, 196 and gender 720–2 Germany 228 Great Britain 239, 243 India 269, 271 Italy 298 Japan 321 Mexico 342, 344–5, 346–7, 348, 350 Netherlands 386, 386, 388–9 Ottoman Empire 495 Poland 405–6, 411, 416–18 Russia/USSR 435–41, 444 Spain 466–8, 475–6, 475 Turkey 504–5 Uruguay 525 USA 534–5, 536–7, 538, 540–1, 542–5, 549–50, 552–4, 555, 602, 721, 789–90 see also labour relations; trade unions Subei (China) 104, 122, 673 Sudamtex S.A. 28, 29, 35, 39, 41 Sümerbank (Turkey) 497, 503, 507 sumptuary laws: Japan 307 Portugal 90 Swadeshi ideology (India) 261 Switzerland, textile entrepreneurs 49, 55, 65–6 ‘synchronicity of the non-synchronous’ 623, 624, 625, 774–7 syndicalism 31–2, 387, 759 Syndicalists (Argentina) 31–2 synthetic fibres see artificial fibres Syria (Ottoman Empire) 484

cotton 567, 779 Argentina 26–7, 27, 38 Brazil 84, 85 China 107, 108, 111–12, 116, 586 Egypt 179, 184–5 Germany 202, 203, 209–10, 209, 211, 216–17, 223 Great Britain 236–7, 586, 755 Habsburg Monarchy 54, 55 India 259–60 Japan 314, 315, 327 Netherlands 378–9, 381, 381, 382, 383, 383 Russia/USSR 427, 433, 436 Spain 454 gender division of labour, Netherlands 374–5, 374 Great Britain 631, 730–1, 732–3 Habsburg Monarchy 53, 58, 586 hand-spinners, China 111–12 India 111, 259–60, 731 Japan 111, 586, 731 Ottoman Empire 480, 731–2 Poland 407 Spain 456, 458, 459, 463–4 spinning jenny 236, 370, 567, 730, 779 USA 533, 537–8, 540–1, 545 wool: Argentina 26–7 Germany 204, 206, 218, 220 Great Britain 238–9 Italy 290–1 Netherlands 370, 374–5, 374, 379, 380, 382, 383 see also cotton industry; mule spinning; ring spinning; self-actor spinning machines; wool industry; yarn production state subsidy of textile industry, Denmark 144 Statute of Artificers (Great Britain) 231 steam power: Denmark 147 Germany 202, 219 Great Britain 236 Japan 314 Netherlands 379, 381 Spain 456 steam reeling (of silk) 112, 113–15 Stichting Textielvak (Netherlands) 392

tall looms (Japan) 307, 585 Talleres Coghlan (Argentina) 28 tariff protection see protectionism Taylor, F.W. 163, 520, 610 Taylorism 443, 518, 520, 522, 610–13, 614, 696 see also Scientific Management TCV (Three Steam Classes, Spain) 461–2, 465, 466, 763, 764 TEKSIF (Textile and Knitting Workers’ Federation, Turkey) 504, 506, 758, 761 829

The Ashgate Companion to the History of Textile Workers telares sueltos (domestic production of textiles, Mexico) 333, 335, 591 textile firms, and labour relations 597–619 Textile Industry Employers Association (Poland) 411 Textile and Knitting Workers’ Federation (TEKSIF, Turkey) 504, 506, 758, 761 textile machinery, export ban by Great Britain 425, 426, 427, 533 textile machinery industry: Argentina 28 China 134 Denmark 144–5, 146–7 USA 548 Textile Manufacturers’ Federation (Denmark) 156–7 Textile Policy (India) 271, 273 textile trade see exports; global trade, imports Textile Workers’ Congress (COT, Uruguay) 525 Textile Workers’ Federation (Denmark) 155–6, 158–62, 163, 167–9, 612, 764, 770 Textile Workers’ Union of America (TWUA) 554, 555 Textile Workers’ Union (Austria) 72 Textile Workers’ Union (UOT, Argentina) 31, 36, 761 Textile Workers’ Union (Uruguay) 521–2 Textillösung Ost (Austria) 69 Textron (USA) 548, 550, 553 Thailand, artificial fibres 574 Third World countries, transfer of textile production to 68–9, 71, 783–4 ‘Thread and Money’ system 629 Thuringia (Germany) 200, 203, 205, 205, 206, 206 Tianjin (China) 104, 120 Tilburg (Netherlands) 364, 365, 368, 369, 371, 372, 377, 379, 390, 394, 784 Tilly, Louise 790–1 Tools Act (Britain) 144 tops production, Uruguay 516, 517, 518, 519, 529, 530 tourism, Argentina 20 trade in textiles see exports; global trade, imports Trade Union of Textile Industry Workers (Poland) 416

trade unions 129–30, 605–9, 616, 754–71, 786–7 Argentina 24, 31–2, 36–7, 40–1, 611–12, 765 Austria 70 ‘Business Unionism’ 760–1 Catholicism 229, 351, 387, 760 Christian-based 760, 761 clientilism 765 Czechoslovakia 63 Denmark 155–64, 167–9, 612, 756, 757, 764–5 differences from guilds 756–7 Egypt 173–4, 184–5, 186–92, 766 and gender 719–22, 755–6 Germany 227–9, 757, 758–9, 760 Great Britain 757, 764–5 cotton industry 237–8, 240–1, 242–3, 249, 251, 252, 786–7 wool industry 235, 239, 240, 243–5, 248–9, 252 Habsburg Monarchy 58, 60 India 268–9, 758, 765–6 Japan 316, 327, 328, 329, 758 Mexico 342, 344–5, 346–7, 349–51, 353–6, 359, 360, 757, 765 Netherlands 386–9, 391–2, 757, 760 in non-capitalist states 766–8 Ottoman Empire 495 Poland 405–6, 411, 416–18 politics and ideology 758–61, 789 Protestantism 385, 387 recent developments 769–71 rise of unionism 755–7 Russia/USSR 435–9, 440–1, 443, 757, 766–8 Spain 461–2, 465–71, 476, 763–4 Turkey 504–5, 758, 760, 761 union growth 757–8 unions and the state 763–9 Uruguay 521, 524–6 USA 534, 536, 538, 544–5, 554–5, 602, 757, 760–1 see also collective bargaining; labour relations; strikes Trades Union Congress (TUC), Great Britain 240 transport developments: Mexico 343 Russia/USSR 428–9 USA 534 830

Index TWUA (Textile Workers Union of America) 554, 555

transport facilities 58 Transport and General Workers’ Union (Great Britain) 252 transport revolution 5 trapiches see obradores trouser fabrics 207 TUC (Trades Union Congress, Great Britain) 240 Tucuman (Argentina) 21, 22 TÜRK-IŞ (Turkey) 504, 505, 758, 761 Turkey 498 artificial fibres 574 carpet production 497, 501, 503–4, 507, 508 child labour 506–7 clothing industry 499, 500, 500–1 cotton industry 499 domestic production of textiles 508, 582 exports 500–2, 507, 508 gender division of labour 502, 503 guest workers 500, 508, 675, 677 history of textile production 497, 499–509 import substitution 500, 507 knitting industry 500, 501 labour relations 504–5 see also trade unions men, role in textile production 503 see also gender division of labour merchants’ role in textile industry 508 migrant workers 500, 508, 675, 677, 743 Muslims 503, 504, 760 nationalization of the textile industry 497, 499 privatization of textile industry 499 production statistics 12 putting-out system 600 spinning 12 strikes 504–5 trade unions 504–5, 758, 760, 761 wages 504, 505–6 women, role in textile production 502 see also gender division of labour workers, number of 502, 504 workshop production of textiles 508, 582, 600 Turner, Ben 245 Twente (Netherlands) 200, 364, 369, 370, 378–9, 381, 382, 384–5, 388–9, 390, 393–4, 566, 600, 603

Unión Obrera Textil (Textile Workers Union, Uruguay) 521–2 Unión Obrera Textil (UOT, Argentina) 31, 36, 761 unions see trade unions Unitas (Netherlands) 387 United Nations 99 United Textile Factory Workers’ Association (Great Britain) 241, 757 United Textile Workers (UTW-AFL, USA) 545, 549, 550–1, 552–4, 760 UOT (Textile Workers’ Union, Argentina) 31, 36, 761 Uriburu government, Argentina 27–8 Uruguay 512 bureaucratization of work 612 carding 516, 518, 519 child labour 526 collective bargaining 521–2 see also labour relations; trade unions cotton industry 516 de-industrialization 523 decline of textile industry 517 domestic production of textiles 522, 581–2 ethnicity of textile workers 672 exchange rate policies 528 exports 514, 515, 516, 517, 518, 523, 527, 529–30 factory production of textiles 514, 516, 581–2 gender division of labour 522, 523–4, 524, 525, 526, 527 history of textile production 511, 513–30, 513, 581 early textile industry 514–15 period of import substitution 515–17 crisis and transformation 517–18 organization of production 518–22 workers 523–7 relationship with the state 527–30 impact of First World War 515 import substitution 511, 515–17, 522, 528, 529 imports 515, 516 indigenous textile production 526 831

The Ashgate Companion to the History of Textile Workers and globalization 556–7 health of textile workers 534 history of textile production 531, 533–57, 568, 573, 629 origins in New England 533–5 industrialization and change in the Northeast 535–8 impact and aftermath of American Civil War 1861–1865 538–41 unionization 542–5 countering Southern competition 546–8 woollen and silk industries 548–50 Southern textile industry 550–3 1934 General Textile Strike 553–4 ‘Operation Dixie’ and the postwar period 554–6 reconstruction during globalization 556–7 housing of textile workers 551, 552, 666, 668 imports 556 labour relations 534–5, 536–8, 540–1, 542–5, 546, 549–51, 553–6, 602, 788, 789–90 lockouts 544, 546, 551, 602 men, role in textile production 535, 540, 540 see also gender division of labour migrant workers 535, 536, 537, 538, 539, 540, 541, 542–3, 545, 547, 549, 556, 665–6, 671, 687 paternalism 603 production statistics 12 protectionism 533, 548, 548, 549 Protestantism 534, 535, 666, 756 proto-industrialization 659 ‘race to the bottom’ 556, 713 recruitment of labour 666 silk industry 535, 550, 666 slave labour 534, 659, 660, 666 spinning 533, 537–8, 540–1, 545 mule spinning 536, 537–8, 539, 540, 541, 542–3, 544–5, 731, 732 number of spindles 12, 546 number of spinners 542 ring spinning 217, 543, 545, 552 self-actor spinning machines 537 strikes 534–5, 536–7, 538, 540–1, 542–5, 549–50, 552–4, 555, 602, 721, 789–90

knitting industry 514 labour relations 521–2, 524–6, 612 men, role in textile production 524, 524, 526, 527 see also gender division of labour migrant workers 526, 665, 672 production statistics 12 protectionism 514, 517, 527–9 Salary Councils 521, 525 strikes 525 see also labour relations; trade unions tops production 516, 517, 518, 519, 529, 530 trade unions 521, 524–6 see also labour relations wages 522, 526–7 weaving 519–20 women, role in textile production 522, 524, 524, 525, 526, 527 see also gender division of labour wool industry 514, 515, 516, 517 workers, number of 514, 516, 524 working conditions 515 workshop production of textiles 514, 515 see also Latin America USA 532 artificial fibres 552–3, 574, 614 bureaucratization of work 611, 612–13 carding 531 carpet production 553, 557 Catholicism 535, 666, 756 child labour 533, 552, 730 comparative manufacturing costs 783 cotton cultivation 639 cotton industry 12, 533–4, 536–48, 547, 550–2, 551 cotton printing 539, 541 weaving 533, 543 de-skilling of textile work 610 decline of textile industry 556–7 domestic production of textiles 533 ethnicity of textile workers 665–6, 671 exports 534, 546, 568 factory production of textiles 533–4, 535–6, 539, 540 fine fabrics 535–6 fines imposed on workers 541 gender division of labour 534–5, 540, 540, 542–3, 549–50, 552, 659, 666 832

Index see also labour relations; trade unions textile exports, to China 111, 131 textile imports, from India 564 textile machinery industry 548 trade unions 534, 536, 538, 544–5, 554–5, 602, 757, 760–1 see also labour relations; strikes transport developments 534 wages 540–1, 783 water power 533 weaving 545 automatic looms 553 cotton 533, 543 handloom weavers 531, 533 number of looms 12 number of weavers 542, 545 power-loom weaving 533 women, role in textile production 534–5, 540, 540, 542–3, 549–50, 552, 659, 666 wool industry 12, 548–50, 548 workers, number of 12, 535, 540, 542, 545, 546, 549, 551, 555 working conditions 534 working hours 692 workshop production of textiles 659 Yankee textile workers 535, 537, 538, 666, 736 USSR see Russia/USSR UTW-AFL (United Textile Workers, USA) 545, 549, 550–1, 552–4, 760

Vorlande (Austria) 45, 52 Wafd (Egypt) 185, 187, 188, 190 wages 694–700, 701, 779–80, 783–4, 783 Brazil 83, 88, 661, 695, 716 China 699, 717, 783 Denmark 715 wage lists 155, 696 Egypt 695, 716 gender-based differentials 698, 715–17, 780–81 Germany 225–6, 226 Great Britain 241–2, 245, 247, 717, 783 wage lists 241, 244–5, 247, 696 India 263, 268, 274, 783 Italy 281–2, 296, 300, 783 Japan 320, 699, 717, 783 Mexico 340, 351, 357, 360, 783 Netherlands 382, 385–6, 392, 784 Poland 405 Russia/USSR 434, 699 Spain 783 Turkey 504, 505–6 Uruguay 522, 526–7 USA 540–1, 783 Waldviertel (Austria) 70–1 Waltham (USA) 533, 534 war, impact on textile industries 744–5 see also First World War; Second World War water frames 144, 145, 236, 244, 434, 454, 457, 533, 567 see also Arkwright, Richard water power: China 106 Germany 202, 219 Habsburg Monarchy 48 Mexico 339, 341 USA 533 water, use in textile processing 48 weavers: earnings: Brazil 88 China 108, 109, 127 Denmark 161 Egypt 176, 177, 183, 194 Great Britain 241, 247–8 Russia/USSR 434 hereditary 103 number of:

Valencia (Spain) 450, 452, 652 value of production, Argentina 34 Vejle (Denmark) 142, 158–9 Venetia (Habsburg Monarchy) 44, 45–6 Veracruz (Mexico) 334, 339, 341, 341, 348, 349, 349, 350 Vereniging Katoen-, Rayon- en Linnenindustrie (KRL, Netherlands) 394 Verlagsystem 49, 252, 452 Verleger (putting-out agents) 46 Vienna (Habsburg Monarchy) 44, 65, 72 silk industry 46, 51, 55, 56, 59 Vietnam, textile production 616 Vladimir (Russia) 427, 431, 433 see also Central Industrial Region (Russia) Vorarlberg (Habsburg Monarchy) 44, 49, 54, 55, 65–6, 69–70, 72 833

The Ashgate Companion to the History of Textile Workers Spain 454, 456, 456–7, 459, 464, 735 USA 545 wool: Germany 204, 205, 218, 221 Great Britain 238 Italy 197 Netherlands 376, 379, 380, 383 see also cotton industry, wool industry; looms; power-loom weaving West Africa, textile imports from India 564 Westphalia (Germany) 200, 201, 202, 220, 229 Wevers- en Spinnersbond Vooruit (Netherlands) 387 Whitley Councils (Great Britain) 248, 249 Wirtschaftsherrschaft 45, 48 women, role in textile production 4, 584, 589–90, 703–24, 737–8, 780–1, 785–6 Argentina 30 Austria 68, 72 Brazil 88, 96 China 103, 120–5, 126, 128, 135, 714, 715 Denmark 155, 164–8, 165, 588, 739 Egypt 174, 179, 181, 182, 194, 714 Germany 213, 223–4, 704 Great Britain 236, 237, 238, 239, 241, 242, 251, 663–4, 714–15 Habsburg Monarchy 48, 52, 56 India 264–5, 269–70, 272, 664 Italy 279, 281, 282, 288–9, 295, 300 Japan 310, 311, 318, 319, 320–1, 329–30, 699, 704, 705 Mexico 337, 340, 342, 344, 347–8, 349, 353, 359, 590 Netherlands 370, 374–5, 374, 376, 377, 378, 379, 380, 382, 385, 390, 391, 704, 705 Ottoman Empire 480, 491–3, 494, 589, 714 Poland 403, 404–5, 406, 409, 409, 410, 411, 416, 417 Russia/USSR 433–4, 438–40, 443, 445, 446 Spain 451, 453, 456, 458–61, 460, 464, 465, 466–7, 473–5, 588, 588–9 Turkey 502 unionization 167–8 Uruguay 522, 524, 524, 525, 526, 527 USA 534, 534–5, 540, 540, 542–3, 549–50, 552, 659, 666

Argentina 30 China 120 Egypt 180, 182, 183, 184 Germany 210, 211, 212, 214 Italy 283 Japan 318 Mexico 340 Netherlands 368, 369, 377, 380, 383 Poland 403 Russia/USSR 428 USA 542, 545 Weavers’ and Spinners’ Trade Union Forward (Netherlands) 387 weaving 585–6, 628–9, 633–4, 634, 733–5 Asia 634 cotton 779 China 108, 109, 117–18, 131–2, 133, 197, 734 Denmark 147, 151–2, 160–2, 165 Egypt 184–5 Germany 203, 207, 210–11, 210, 217–18, 219–20 Great Britain 238, 247–8 India 257–8, 260–5, 266, 271 Japan 315, 327 manufacturing costs 782–3, 783 Netherlands 381, 381, 382, 383 USA 533, 543 Egypt 177–8 Germany 219, 222, 224, 226 Habsburg Monarchy 56, 633 handloom weaving: Argentina 19 China 107–9, 117–19 Great Britain 235 Habsburg Monarchy 54, 56, 59, 60–1 India 257, 260–5, 266, 593, 735 Italy 281–2 Jacquard looms 55, 59, 314, 464, 557 mechanization of 58, 59 Mexico 336, 337, 733–4 Ottoman Empire 484, 485, 494, 735 Poland 406, 407 Russia/USSR 425–6, 427–8, 433, 436, 734, 735 silk: China 114–15, 119–20, 121, 122–3, 126, 128, 129–30, 591–3, 733, 734 Egypt 183 India 264 834

Index Workers’ Federation of the Textile Industry (FOIT), Argentina 31 workers, number of: Argentina 21, 25–6, 30, 30, 33, 34, 37, 39 Austria 12, 54, 71–2 Belgium 12 Brazil 12, 98, 99 China 12, 120, 131 Czechoslovakia 12, 71–2 Denmark 145, 147, 148–9, 149, 152 Egypt 178, 180–1, 182–3, 184, 185, 193–4, 195, 684 France 13 Germany 13, 204, 208–9, 209, 210, 210, 211–12, 213, 214, 217, 219, 221, 222 Great Britain 13, 240, 615 India 12, 266 Italy 13, 279, 283, 295, 296, 297, 299, 300 Japan 12, 315, 318, 323 Lower Austria 54 Mexico 12, 339, 340, 343, 353, 357, 357 Netherlands 13, 368, 369, 378, 381 Ottoman Empire 477, 485 Poland 13, 403, 404, 407, 408, 415–16, 415, 420 Russia/USSR 13, 425, 428, 429–30, 430, 448 Scotland 615 Spain 13, 451, 452, 453, 455, 456, 460, 463, 473–4, 473, 474 Turkey 502, 504 Uruguay 514, 516, 524 USA 12, 535, 540, 542, 545, 546, 549, 551, 555 working conditions 679–701, 717–18 Argentina 24, 31, 40–1 China 125–30, 135 Denmark 160–2 domestic production of textiles 680, 684–5, 691, 693, 701 Egypt 183, 195–6, 690, 691 Habsburg Monarchy 57, 406 India 564–5 Italy 284–5, 288–9 Japan 311, 319–20 Mexico 336–7, 345–6 Netherlands 376–7, 383–6, 392 Poland 406 Russia/USSR 434–5, 438 Uruguay 515

wage differentials 715–17 see also gender division of labour wool industry 562, 571 Argentina 12, 23–4, 30 Australia 12 Austria 12 Belgium 12 Brazil 12 China 12 Czechoslovakia 12 Denmark 12, 149, 150 France 13 Germany 13, 204–7, 207–8, 209, 212–13, 215–16, 220, 221, 223 mechanization 218–19 global trade 22, 561, 565–6 Great Britain 13, 233, 234–5, 238–9, 239, 240, 243–5, 246, 248–9, 251, 252 trade unions 239, 240, 243–5, 248–9, 252 Habsburg Monarchy 46, 51 India 12 Italy 13, 279–82, 289–92, 297, 299 Japan 12, 311 measures of production 8 Mexico 12 Netherlands 13, 365–6, 366, 367–8, 369, 370–2, 373, 374–8, 377 number of workers: Argentina 30 Germany 208, 209, 214 Habsburg Monarchy 50 Italy 292, 296, 299 Mexico 353, 357 Netherlands 368, 369, 377, 380, 383 Ottoman Empire 483 Poland 13, 399–400, 401, 402, 407, 408, 413, 414–15, 414, 418–19, 419 raw wool consumption (by country) 12, 13 Russia/USSR 13, 431, 442, 447 Spain 13, 452, 454 Uruguay 514, 515, 516, 517 USA 12, 548–50, 548 see also worsted industry wool-combers: Great Britain 239 Italy 282, 284 Netherlands 375–6, 728 gender division of labour 375 835

The Ashgate Companion to the History of Textile Workers Uruguay 514, 515 USA 659 working conditions 692–3 see also artisan production of textiles workspaces 682–6 World Bank 194, 500, 578 worsted industry: Germany 204, 205, 206, 212 Great Britain 233, 234–5, 238, 243 see also wool industry Württemberg (Germany) 200, 202, 203, 204, 206, 217, 217, 222–3, 224, 225

USA 534 workshop production of textiles 692–3 workspaces 682–6 working hours 692–3 Argentina 36, 37, 39 Brazil 89, 96, 98, 100, 692, 693 China 125–6, 127, 692 Egypt 192, 194, 692–3 Germany 224–5, 226, 692 India 261 Italy 284, 296 Japan 319–20, 328–9, 692 Netherlands 385, 692, 693 Poland 404–5, 406, 411 Russia/USSR 434, 438, 692 Spain 460, 467–8, 692 USA 692 workshop production of textiles 579, 590, 599, 628, 689–90, 728–30, 779 Americas 657–8, 675 Brazil 79, 730 China 103, 106, 108, 117, 118–19, 129, 131, 133, 137, 138–9, 591–3, 629, 728, 734 Denmark 143, 154 Egypt 173, 174–5, 177–9, 180, 599 Germany 220, 222 Habsburg Monarchy 48, 50, 52, 54, 56, 59, 654, 728 India 256, 273, 593 Italy 288 Japan 309, 315, 315 Mexico 333, 335–7, 336, 338, 341, 344, 591, 660, 729 Netherlands 371 Ottoman Empire 485, 487, 490–1, 492, 494, 497 Poland 399, 400 Russia/USSR 684 Spain 451, 474, 476, 591 Turkey 508, 582, 600

Yangzi Delta (China) 108, 109, 113 Yankee textile workers (USA) 535, 537, 538, 666, 736 yarn production 12, 13 Argentina 12, 26, 27, 573 China 12, 134, 136 Denmark 12, 148 Germany 13, 204, 205, 206, 209, 216–17, 217, 218 Ottoman Empire 480 Poland 13, 413, 418–19, 419 prices 567 Russia/USSR 13, 425, 427, 428, 430 see also spinning Yarur, Don Juan 671 Yorkshire (Great Britain): wool industry 233, 234–5, 238–9, 240, 246, 251, 726, 730, 732, 737, 740, 741, 744, 779 migrant workers in USA 537, 549, 736 trade unionism 239, 240, 243–5, 248–9, 252, 608–9, 739, 741 YWCA 130, 720, 740–1 zangfang system (China) 629 Zensen (Japan) 327, 328, 329 zünft system (Denmark) 154, 155, 756

836