German Immigration and Servitude in America, 1709-1920 [1 ed.] 9781136682513, 9780415610612

Citation preview

German Immigration and Servitude in America, 1709–1920

Germans comprised the largest group of non-­English-speaking European immigrants to English North America in the eighteenth and nineteenth centuries. Their impact was felt primarily in the middle-­colonies and central-­states of America. They made substantial contributions to the population growth, cultural development, social structure, and economic dynamism of this region. Knowing who these immigrants were, how the migration process worked, and how they fit into English America is important for comprehending the American immigrant experience. More quantifiable records were generated for this German migration than for other European arrivals. Exploiting these records, along with using the substantial literary writings and personal correspondences that have survived, for this migration allows for a more complete description and detailed investigation than for any other European transatlantic migration in this era. Besides the magnitude and timing of German immigration, extensive knowledge about the immigrants themselves can be gleaned from these records, such as their educational levels, family structures, age profiles, and occupational backgrounds. The nature and difficulties of the recruitment and shipping process, and the travails of the voyage, including that of crowding, voyage mortality, and post-­voyage mortality, are vividly revealed. This book provides the most comprehensive history of German migration to North America for the period 1709 to 1920 than has been done before. Employing state-­of-the-­art methodological and statistical techniques, the book has two objectives. First it explores how the recruitment and shipping markets for immigrants were set up, determining what the voyage was like in terms of the health outcomes for the passengers, and identifying the characteristics of the immigrants in terms of family, age, and occupational compositions and educational attainments. Second it details how immigrant servitude worked, by identifying how important it was to passenger financing, how shippers profited from carrying immigrant servants, how the labor auction treated immigrant servants, and when and why this method of financing passage to America came to an end. Farley Grubb is Professor of Economics at the Alfred Lerner College of Business and Economics at the University of Delaware, USA.

Routledge explorations in economic history Edited by Lars Magnusson Uppsala University, Sweden

  1 Economic Ideas and Government Policy Contributions to contemporary economic history Sir Alec Cairncross   2 The Organization of Labour Markets Modernity, culture and governance in Germany, Sweden, Britain and Japan Bo Stråth

  7 Production Efficiency in Domesday England, 1086 John McDonald   8 Free Trade and its Reception 1815–1960 Freedom and trade: volume I Edited by Andrew Marrison   9 Conceiving Companies Joint-­stock politics in Victorian England Timothy L. Alborn

  3 Currency Convertibility The gold standard and beyond Edited by Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis

10 The British Industrial Decline Reconsidered Edited by Jean-­Pierre Dormois and Michael Dintenfass

  4 Britain’s Place in the World A historical enquiry into import controls 1945–1960 Alan S. Milward and George Brennan

11 The Conservatives and Industrial Efficiency, 1951–1964 Thirteen wasted years? Nick Tiratsoo and Jim Tomlinson

  5 France and the International Economy From Vichy to the Treaty of Rome Frances M.B. Lynch   6 Monetary Standards and Exchange Rates M.C. Marcuzzo, L. Officer and A. Rosselli

12 Pacific Centuries Pacific and Pacific Rim economic history since the 16th century Edited by Dennis O. Flynn, Lionel Frost and A.J.H. Latham 13 The Premodern Chinese Economy Structural equilibrium and capitalist sterility Gang Deng

14 The Role of Banks in Monitoring Firms The case of the crédit mobilier Elisabeth Paulet 15 Management of the National Debt in the United Kingdom, 1900–1932 Jeremy Wormell 16 An Economic History of Sweden Lars Magnusson 17 Freedom and Growth The rise of states and markets in Europe, 1300–1750 S.R. Epstein 18 The Mediterranean Response to Globalization before 1950 Sevket Pamuk and Jeffrey G. Williamson

23 The Origins of National Financial Systems Alexander Gerschenkron reconsidered Edited by Douglas J. Forsyth and Daniel Verdier 24 The Russian Revolutionary Economy, 1890–1940 Ideas, debates and alternatives Vincent Barnett 25 Land Rights, Ethno Nationality and Sovereignty in History Edited by Stanley L. Engerman and Jacob Metzer 26 An Economic History of Film Edited by John Sedgwick and Mike Pokorny 27 The Foreign Exchange Market of London Development since 1900 John Atkin

19 Production and Consumption in English Households 1600–1750 Mark Overton, Jane Whittle, Darron Dean and Andrew Hann

28 Rethinking Economic Change in India Labour and livelihood Tirthankar Roy

20 Governance, the State, Regulation and Industrial Relations Ian Clark

29 The Mechanics of Modernity in Europe and East Asia The institutional origins of social change and stagnation Erik Ringmar

21 Early Modern Capitalism Economic and social change in Europe 1400–1800 Edited by Maarten Prak

30 International Economic Integration in Historical Perspective Dennis M.P. McCarthy

22 An Economic History of London, 1800–1914 Michael Ball and David Sunderland

31 Theories of International Trade Adam Klug Edited by Warren Young and Michael Bordo

32 Classical Trade Protectionism 1815–1914 Edited by Jean Pierre Dormois and Pedro Lains 33 Economy and Economics of Ancient Greece Takeshi Amemiya 34 Social Capital, Trust and the Industrial Revolution 1780–1880 David Sunderland 35 Pricing Theory, Financing of International Organisations and Monetary History Lawrence H. Officer 36 Political Competition and Economic Regulation Edited by Peter Bernholz and Roland Vaubel 37 Industrial Development in Postwar Japan Hirohisa Kohama 38 Reflections on the Cliometrics Revolution Conversations with economic historians Edited by John S. Lyons, Louis P. Cain, and Samuel H. Williamson 39 Agriculture and Economic Development in Europe since 1870 Edited by Pedro Lains and Vicente Pinilla 40 Quantitative Economic History The good of counting Edited by Joshua Rosenbloom

41 A History of Macroeconomic Policy in the United States John H. Wood 42 An Economic History of the American Steel Industry Robert P. Rogers 43 Ireland and the Industrial Revolution The impact of the Industrial Revolution on Irish industry and society, 1801–1922 Andy Bielenberg 44 Intra-­Asian Trade and Industrialization Essays in memory of Yasukichi Yasuba Edited by A.J.H. Latham and Heita Kawakatsu 45 Nation, State and the Industrial Revolution The visible hand Lars Magnusson 46 A Cultural History of Finance Irene Finel-­Honigman 47 Managing Crises and De-­globalisation Nordic foreign trade and exchange 1919–1939 Edited by Sven-­Olof Olsson 48 The International Tin Cartel John Hillman 49 The South Sea Bubble Helen J. Paul 50 Ideas and Economic Crises in Britain from Attlee to Blair (1945–2005) Matthias Matthijs

51 Bengal Industries and the British Industrial Revolution (1757–1857) Indrajit Ray 52 The Evolving Structure of the East Asian Economic System since 1700 Edited by A.J.H. Latham and Heita Kawakatsu

53 German Immigration and Servitude in America, 1709–1920 Farley Grubb

German Immigration and Servitude in America, 1709–1920 Farley Grubb

First published 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2011 Farley Grubb The right of Farley Grubb to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized 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. Trademark 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 A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Grubb, Farley Ward, 1954–   German immigration and servitude in America, 1709–1920/Farley Grubb. p. cm. Includes bibliographical references. 1. German Americans–History. 2. Immigrants–United States–History. 3. Indentured servants–United States–History. 4. Germany–Emigration and immigration–History. 5. United States–Emigration and immigration–History. 6. German Americans–Statistics. 7. Immigrants– United States–Statistics. 8. Indentured servants–United States– Statistics. 9. Germany–Emigration and immigration–Statistics. 10. United States–Emigration and immigration–Statistics. I. Title. E184.G3G78 2011 973′.0431–dc22 2010053093 ISBN: 978-0-415-61061-2 (hbk) ISBN: 978-0-203-81050-7 (ebk) Typeset in Times by Wearset Ltd, Boldon, Tyne and Wear

Contents



Lists of figures and tables Preface

  1 Introduction PART I

German immigration to America, 1709–1835

xii xvii 1

15

A The migration experience: magnitudes, causes, conveyances, and conditions   2 The flow of immigrants, 1727–1835

17

  3 The transatlantic shipping market

52

  4 Morbidity and mortality on the North Atlantic passage

71

B Immigrant characteristics and human capital   5 Age, occupation, and family composition

89

  6 Literacy: longitudinal patterns and market forces

106

  7 The age structure of German immigrant literacy

131

  8 Educational choice in the era before free public schooling

144

x   Contents PART II

German immigrant servitude in America, 1745–1835

157

A Patterns of servitude among the immigrants   9 The incidence of servitude in transatlantic migration, 1771–1804

159

10 Servant auction records, 1745–1831: the proportion of females among the servants

181

11 The occupational and geographical distribution of immigrant servant labor in the Delaware valley

197

B The market for German immigrant servants 12 Determining the method of entering servitude and modeling its performance

220

13 Servant contract choice and shipper profits

244

14 The auction of German immigrant servants in Philadelphia, 1771–1804

255

15 Debt shifting within German immigrant families

276

C The end of German immigrant servitude in America, 1784–1835 16 Processing German servants at the port of Philadelphia, 1817–31: the documents

303

17 The disappearance of organized immigrant servant markets in the U.S.: five popular explanations reexamined

316

18 The collapse of the German immigrant servant market: timing and causes

341

Contents   xi PART III

Epilogue: German immigration to the U.S., 1820–1920: from founding migration to mass migration

373

19 German immigration to the U.S., 1820–1920: magnitudes, patterns, and relative shares

375

20 German immigrants in the mass migration era

398



Index

424

Figures and tables

Figures   2.1 The flow of German immigrants through the port of Philadelphia, 1727–1835   5.1 Adult male literacy among German immigrants to Pennsylvania   6.1 Patterns of regional literacy growth in rural colonial America   6.2 Supply and demand model of colonial literacy decline and growth caused by changes in population density   7.1 Illiteracy percentage by year: German immigration to Philadelphia, 1727–75   7.2 Age structure of illiteracy: logit regressions, German male immigration to Philadelphia, 1730–54 12.1 Immigrant servant contract entered into by Mary Elizabeth Bauer, 1767 12.2 Immigrant servant contract entered into by Anna Margareta Sahlin, 1785 12.3 Immigrant servant contract entered into by Daniel Kent, 1784 12.4 The basic market model of adult German redemptioner servitude 12.5 The enhanced market model of adult German redemptioner servitude 15.1 Servitude among German immigrant children, 1785–1804 16.1 Passenger contract for German immigrants on board the ship Favorite, captain John Alftan [Alston], bound for Philadelphia from Amsterdam 17.1 The demand hypothesis of the end of European immigrant servitude 18.1 Percentage of servants among German immigrants to Pennsylvania, 1772–1835 18.2 Adult passage fares, contract prices, and per-­year contract prices, 1772–1835

28 97 114 119 132 136 224 225 226 232 236 287 309 334 347 350

Figures and tables   xiii 18.3 The supply and demand for German servants in the Philadelphia market, 1772–1821 19.1 German immigration to the U.S., 1820–1920 19.2 German emigration rates and destination percentages, 1820–1920 19.3 U.S. immigration: ethnic shares, 1820–1920 19.4 Ports of embarkation for German emigrants, 1847–1920 (percentage distribution) 19.5 Origins of German overseas emigrants by region, 1830–1910 19.6 Percentage distribution of German immigrants to the U.S. by port of arrival, 1840–90 19.7 U.S. decennial censuses: foreign-­born percentages, 1850–1920 20.1 Percentage male for select immigrant groups, 1820–1920 20.2 Percentage of immigrants in selected age groups by ethnic category, 1820–1920 20.3 Occupational distribution among adult male immigrants to the U.S., 1820–1920

351 377 383 384 386 388 390 393 401 403 408

Tables   1.1 Percentage distribution of the white population of the United States by nationality and state from the 1790 census   1.2 The raw distribution of the white population of the United States by nationality and state from the 1790 census   1.3 Estimated population growth in the Delaware valley colonies, 1730–75   2.1 Regional distribution of German immigrants’ residence of origin by port of departure, 1798–1808   2.2 Distribution of the ports of origin for German immigrants to Philadelphia, 1727–75   2.3 Distribution of the ports of origin for German immigrants to Philadelphia, 1785–1804   2.4 Regression results: the determination of the variance of German immigration to Pennsylvania, 1727–74   2.5 Seasonal distribution of German immigrants arriving in Philadelphia, 1727–1804   2.6 Port of last clearance for German immigrant vessels arriving in Philadelphia, 1727–75   2.7 Regression results: determination of crowding (passengers-­ per-ton) on vessels carrying German immigrants to Philadelphia, 1727–74   3.1 The activities of the Steadman family in shipping German immigrants to Philadelphia, 1731–54   3.2 Role of ship captains in carrying German passengers to Pennsylvania

2 3 3 22 23 24 30–1 34 36 40 54–5 61

xiv   Figures and tables   3.3 Consignment of German redemptioner cargos in Philadelphia, 1751–74   3.4 Trans-­Atlantic passage fares for adult male Germans, Holland to Philadelphia, 1708–1819   3.5 German immigrants per ship Holland to Philadelphia, 1727–74   4.1 Passage mortality: German immigration to Pennsylvania, 1727–1805   4.2 Debarkation morbidity for men above age 15, German immigration to Pennsylvania, 1727–54   4.3 Estimates of first year German immigrant versus resident mortality: Philadelphia, 1738–62   4.4 Distribution of major descriptive causes of mortality due to infectious disease in Christ-­Church parish, Philadelphia, 1747–75   5.1 Ratio of dependent migrants to adult male migrants   5.2 Social composition of German and English immigrants to Pennsylvania, 1709–1820   5.3 Family composition of German and English immigrants to Pennsylvania, 1727–1820   5.4 Age distribution of adult German and English immigrants to Pennsylvania, 1730–1820   5.5 Occupational distribution of adult male German and English immigrants arriving in Pennsylvania, 1709–1820   6.1 Studies of colonial adult male literacy, 1650–1840   6.2 Studies of European adult male literacy, 1633–1840   7.1 The age structure of illiteracy for male German immigrants to Philadelphia, 1730–54   8.1 Education specified in servant contracts for German immigrant children in Pennsylvania, 1771–1804   8.2 Education and age of German immigrant servants in Pennsylvania, 1787–1804   9.1 The distribution by family status of the incidence of servant contracting among German immigrants to Pennsylvania, 1785–1804   9.2 The incidence of servitude for adult parents by family size: German immigration to Pennsylvania, 1785–1804   9.3 The incidence of servitude by age for adult male English emigrants, December 1773 to April 1776   9.4 The incidence of servitude among German immigrant children to Pennsylvania, 1785–1804   9.5 The incidence of servitude among immigrants to Pennsylvania, 1771–3   9.6 The incidence of immigrant servitude across colonies, English immigration, 1773–6

63 65 66 74 75 76–7 83 91 91 92 95 98–9 108–9 110–11 135 147 151 161 163 164 165 167 172

Figures and tables   xv 10.1 Composition of the four Philadelphia servant records, 1745–1831 10.2 The sexual composition of immigrant servants to Philadelphia by nationality, 1745–1831 10.3 Yearly estimates of the sexual composition of German and Irish servant immigration to Philadelphia, 1729–75 10.4 Age and marital status of female servants among German immigrants to Philadelphia, 1771–1831 11.1 The geographic distribution of immigrant servant purchasers: the Philadelphia market, 1745–1804 11.2 Distribution of immigrant servants and slaves by the percentage of resident population for select counties, 1771–3 11.3 The geographic distribution of Irish immigrant indentured servant purchasers by county of residence: Philadelphia market, 1745–46 11.4 The geographic distribution of British immigrant indentured and redemptioner servant purchasers by county of residence: the Philadelphia market, 1771–3 11.5 The geographic distribution of German immigrant redemptioner servant purchasers by county of residence: Philadelphia market, 1771–3 11.6 The geographic distribution of German immigrant redemptioner servant purchasers by county of residence: Philadelphia market, 1787–1804 11.7 The geographic distribution of British indentured and redemptioner servant purchasers by town of residence: the Philadelphia market, 1771–3 11.8 The geographic distribution of German immigrant redemptioner servant purchasers by town of residence: Philadelphia market, 1771–3 11.9 The geographic distribution of German immigrant redemptioner servant purchasers by town of residence: Philadelphia market, 1787–1804 11.10 The occupational distribution of German immigrant redemptioner servant purchasers: Philadelphia market, 1787–1804 11.11 Comparison of the percentage distribution of Pennsylvanian occupations with those of servant purchasers 13.1 Indentured versus redemptioner servitude in the Philadelphia immigrant market, 1745–73 13.2 Distribution of contract characteristics by contract type in the Philadelphia immigrant servant market, 1745–73 13.3 Average rates of return per voyage for financing German redemptioner immigration, 1802–19 13.4 Comparison of passenger-­per-ton ratios

183 185 188–9 191 199 200 202 204–5 206 207–8 209 210 211 214–15 216 245 246 249 251

xvi   Figures and tables 14.1 Redemptioner servant compensation for single adult male immigrants: Philadelphia, 1771–1804 14.2 The auction of British and German redemptioner immigrants: Philadelphia, 1771–3 14.3 Auction of German redemptioner immigrants: Philadelphia, 1787–1804 15.1 Seven families on the ship Britannia arriving from Rotterdam at Philadelphia, September 18, 1773 15.2 Six more families on the ship Britannia arriving from Rotterdam at Philadelphia, September 18, 1773 17.1 The market for German immigrant servants in Philadelphia, 1785–1831 17.2 Incidence of runaway servants among German immigrant servants in the Philadelphia region, 1785–1820 17.3 Geographic distribution of servant buyers, 1787–1831 17.4 Distribution of buyers and servants by the number of servants purchased per buyer, 1817–31 17.5 Distribution of occupations among buyers of German immigrant servants, 1787–1831, with some comparative occupational distributions for immigrants and residents 18.1 German immigration to Pennsylvania, 1772–1835 19.1 Percentage distribution of the German-­born in the 1860 and 1870 U.S. population censuses by their German region or state of origin 19.2 Top dozen U.S. states with German-­born residents in 1850 19.3 Percentage of German-­born residents in the 1850 U.S. population census in the largest U.S. cities 20.1 Percentage distribution of reported occupations in the U.S. censuses by German-­born residents who were gainfully employed, 1850–1910 20.2 Percentage point difference between German-­born occupational distributions and that for (1) the U.S.-born and for (2) the other foreign-­born in the U.S. censuses

260–1 264–5 266–7 282–3 284–5 318 324 327–8 329 331–2 344–6 389 391 392 412–13 415–17

Preface

What follows is an accumulation of 30 years of research on German immigration and servitude in eighteenth- and early nineteenth-­century America. The inspiration for this research came in a graduate course on the economic history of colonial America taught by David W. Galenson at the University of Chicago in the early months of 1980. Therein, Galenson presented his work on the transatlantic market for British indentured servants (Galenson 1981). He argued that servants who were viewed as potentially more productive in colonial America would negotiate shorter labor contracts at embarkation and, therefore, would not sell for higher prices in America compared with less productive servants. His evidence from British embarkation records showed an empirical association between contract length and various measures of servant human-­capital, measures that were proxies for expected labor productivity in America. Being based on embarkation records, however, this work could not empirically explore how contract prices paid at debarkation influenced the transatlantic servant market. Sometime in early 1980 a classmate, Jenny France, showed me a book she had found in the University of Chicago Library that contained a transcription of records from the mayor’s office of Philadelphia listing servant contract prices for European immigrants landing in Philadelphia between October 1771 and October 1773 (“Record of indentures”). I realized that I could incorporate these servant contract prices into the model of British transatlantic indentured servitude and directly measure the impact they had on that market. This became the core chapter of my 1984 PhD dissertation and was subsequently published as an article in the Journal of Economic History in 1985 (Grubb 1984, 1985). Being on British indentured servitude, this material is not included in this volume. These Philadelphia records also contained a substantial amount of information on German servant contract prices. German immigrant servants used a different method for financing transatlantic migration—the redemption method—rather than the indenture method favored by the British. This led me to expand my initial research on British immigrant servitude to encompass that of German immigrant servitude. I wanted to model and measure how German redemption servitude performed and how it differed from the British indenture method. While pursuing this research, I discovered that surviving records for German immigration to the middle colonies and the early U.S. republic were

xviii   Preface more substantial than for other immigrant groups. This allowed me to explore numerous aspects of German immigration that were ancillary to the servitude aspect of that migration, and to do so using quantitative tools and economic insights. The research evolved in a piecemeal fashion toward a comprehensive study of German immigration and servitude in America, covering all the quantifiable dimensions that I could glean from the surviving records found in America for German immigration during the eighteenth and early nineteenth centuries. This period and ethnic group seemed worthy of investigation. Germans represented the largest non-­English-speaking migration to English North America in this era. How such a substantial foreign-­speaking population was integrated into early America would be interesting to know in light of the current debates regarding U.S. immigration. In addition, this was the founding period of German immigration to America which, in some respects, paved the way for the mass migration of Germans to America later in the nineteenth century. Understanding this founding period, then, seemed important to comprehending the mass migration that would come soon thereafter. Finally, many aspects of British and Irish immigration to America during the eighteenth and early nineteenth centuries are unclear due to a lack of surviving quantifiable records, aspects that can be uncovered by studying the German experience. By appropriate analogy, therefore, exploring the data on German immigration might be the best that can be done to uncover the general transatlantic immigration experience in this period. Records on German immigration during the eighteenth and early nineteenth centuries were not systematically collected by governments or businesses, and what were collected are fragmentary and intermittent in their survival. As such, their use often requires careful deconstruction, judgment regarding their representativeness, and merging with other records to make them valuable and as a check on their trustworthiness. Despite these difficulties many aspects of the German migration experience can be sorted out quantitatively with some confidence. The aspects I initially studied included: the market for redemptioner ser­ vants—including the pricing, profits, and contract length determinants in the market; the immigrant shipping market for transporting Germans from Holland to America; the health outcomes experienced by immigrants during and shortly after the transatlantic voyage; the age, gender, and family composition of the immigrant stream; how immigrant servants were distributed geographically after landing and distributed occupationally among American employers; the literacy and occupational skills brought by immigrants to America; and lastly the dynamic process by which the act of immigration affected the relative education levels of adult immigrants and that of their accompanying children. While many of these initial studies were important ancillary components to comprehending the German immigrant servant market, they began to be crafted to stand on their own as component studies of the German migration experience in general. In preliminary form, these investigations were incorporated into my 1984 PhD dissertation. Most of this research was subsequently revised, expanded, and then published in a series of journal articles over the next

Preface   xix half-­decade. These journal articles are reproduced here as Chapters 3–7, 9, 11, 13, and 14. In the process of polishing and expanding this dissertation material into journal articles I was led to augment my research beyond the dissertation. In particular, I wanted to understand the effect of German immigration on the female servant market in the middle colonies and the effect of servant markets on the education of German immigrant children during late colonial and early republican periods. These additional studies were subsequently published as journal articles and are reproduced here as Chapters 8 and 10. I also wanted to expand the research project to cover the period through the end of German immigrant servitude in America, and to solve the puzzle of how German children were used by their immigrant parents in the transatlantic servitude market as well as why they were used this way. To take the research through to the end of organized markets for German immigrant servants in America, besides being something that had not been well studied before, seemed an appropriate termination point for my research into German immigration to America. These additional research projects, originally published as several journal articles and as an introduction to a data transcription book, are reproduced here as Chapters 15–18. Several previously unpublished studies are included in this volume. Besides the introduction in Chapter 1, much of the evidence and analysis presented in Chapter 2 on the flow of German immigrants to America was originally part of my 1984 PhD dissertation, but was never published before now. In part, this was because some of that research relied on a journal article published in 1981 by Marianne S. Wokeck. I presumed that Wokeck was also working on a book, and I did not want to preempt or intrude on her work-­in-progress. She did not publish her book until 1999 (Wokeck 1981, 1999). In the meantime, Hans-­Jurgen Grabbe published a journal article in 1989 that charted German immigration from 1783 through 1820, and Georg Fertig published a chapter in an edited volume in 1994 that presented an econometric estimation of German immigration to colonial America similar to what I had done in my 1984 dissertation. Their published work reduced the novelty of that portion of my dissertation (Fertig 1994: 192–235; Grabbe 1989). I return to that dissertation material here, revising and updating it in Chapter 2. This chapter now serves as a necessary base upon which the rest of the book builds. It also presents evidence and ana­ lysis that is still fresh and new, not to be found elsewhere in the published literature. Chapter 12 presents a previously unpublished graphical model of transatlantic servitude tailored to the German redemption experience. Published studies on German and British immigrant servitude in America typically present mathematical or econometric equation models of servitude (e.g. see Chapter 14). Such models are not necessarily the best way to teach students how servant markets worked, nor are they informative to non-­technical readers. I have used a graphical servant model to teach how transatlantic servant markets performed, in my classes on colonial economic history at the University of Delaware for many

xx   Preface years. While a version of this graphical model was adapted to the transatlantic British convict servant market and published in 2000, a fully enhanced version applied only to the voluntary servant market is presented in Chapter 12 here for the first time (Grubb 2000a). As a heuristic tool, it clearly describes the forces that constrained the market, and it produces numerous relationships that yield testable (falsifiable) propositions that can be empirically investigated. As such, it serves as an excellent introduction to the chapters that follow wherein such pro­ positions about how the German servant market performed are empirically investigated and tested. Finally, the epilogue Chapters 19 and 20 are new. They represent a summary or synthesis of the government data and the published research of other scholars, rather than new or original contributions by me. They are used here to wrap up the book with some comparisons of the founding era of German migration—the original research in the book, with the era of mass migration of Germans to America that arose after 1830 and ended at World War I. Most of the chapters in the book are previously published journal articles. This has certain advantages and certain disadvantages. First, as a professor in an economics department at an American university, promotion and tenure is based on publishing numerous refereed journal articles. Little weight is given to books. Embarking on a 30-year research project to produce a book like the one here, without prior publication of parts as journal articles, would not be possible while still retaining a tenured academic position or being promoted within academia. Having such an academic position is, in turn, necessary to generate the time, income, and access to scholarly resources needed to complete such a lengthy research project. Second, the journal editing and double-­blind refereeing process imposes a high standard on the quality of the research and the presentation of the material. The experience can be quite rough-­and-tumble as well as lengthy. But the iterative process of revising and polishing a manuscript to achieve journal acceptance yields in the end, despite the author’s grief, pain, and frustration, a superior product compared with what is typically acceptable in other venues. Journal articles also impose judiciousness on the author—compactness of presentation and conciseness of argumentation. Lengthy digressions and secondary points in the main text are not tolerated or are at best condensed and confined to footnotes. The material can be both quite dense and speed by quickly, more so than is typical of most books. In addition, the journal refereeing and editing process requires arguments and supporting evidence to be less speculative and more soundly demonstrated and tested than is typical of most books. Third, journal articles are designed to stand on their own as independent pre­ sentations. Thus, most of the chapters in the book can be read and used on their own separately from the rest of the book. At the same time, the book was intentionally organized to flow from chapter to chapter and section to section as a coherent whole, in many cases building on prior chapters to produce a richer understanding of German immigration and servitude than can be gleaned from any single chapter alone. Due to their independent nature as journal articles,

Preface   xxi however, there is some repetition from chapter to chapter, especially in the opening introduction and motivation given in each chapter. However, the substantive analysis and outcome differ from chapter to chapter. Four factors determine the scope of, and period spanned by, the book. First, the book is a study of “German” immigration and servitude to America—meaning Germanic-­speaking immigrants from Germanic-­speaking parts of Europe, though some comparative evidence on other groups is also given. “Germany” as a unified country did not exist until 1871 (Fenske 1980). Second, it is confined to the era when organized markets for German immigrant servants in America existed. Third, it focuses on the use and statistical analysis of quantitative data. Fourth, it relies primarily on evidence on the American side, mostly in English, of German immigration. The book focuses narrowly on the economic factors involved in the actual migration event and much less on the social, political, religious, and cultural conditions in, and their transference between, German communities in Europe and immigrant communities in America pre- and post-­migration. Excellent books by other scholars have covered such material (e.g. see Fogleman 1996; Klepp, Grubb, and Pfaelzer de Ortiz 2006; Roeber 1993; Sabean 1990; Wolf 1976). The book also does not contain all my prior research and journal publications on European immigrant servitude in America, but only those dealing primarily with data on German immigrants. One article not included here may prove instructive and is recommended to those who want to understand more about servant contract design and colonial laws dealing with servant contracting in general, including German servant contracts as a subset therein, namely Grubb (2000b). Also not included here are two articles containing some tangential information on German immigrant servant heights and the matrix of German servant contract lengths (Grubb 1992, 1999). My research into German immigration and servitude is at an end, which explains why this collected volume is being offered now. I have permanently moved on to other projects in a different area of economic history (e.g. see Grubb 2010a, 2010b). Only a few unfinished projects on German immigration and servitude, which I had intended to pursue, have been abandoned. One such project was to show the near impossibility of tracking, in a comprehensive or representative way based solely on nominal linkage, the fate of German immigrants and immigrant servants after they debarked in America. The idea came from looking at the evidence for the ship Britannia, for which there are three different lists of passenger names recorded independently by three different English officials at different times during the migration (“Munstering [mustering] book;” “Record of indentures;” Strassburger 1934: vol. 1, ship# 306). All three lists are for the same group of people as the lists were confined to everyone on the Britannia. Yet, matching individuals across the lists using only their names is quite problematic—suggesting the presence of substantial spelling corruption of German names when spoken to and then written down by English officials. A truly representative or comprehensive tracking of the fate of German immigrants

xxii   Preface and immigrant servants post-­debarkation will have to await a major compilation of the accumulated works of genealogists. The other project abandoned was to verify the facts and assess the veracity and representativeness of the autobiography of the Bavarian immigrant Dr. Christian Boerstler, who arrived in America in 1784, with regard to his migration experience and observation of the German servant market in America (Boerstler 1908—I wish to thank Jim Cawley for bringing this autobiography back to my attention after so many years). It seems like a reasonable account—not all that different from the other firsthand accounts I have read and used. Because of that, and because I never got around to vetting it properly, I did not use Boerstler’s account in the previously published studies included here. But I still recommend it to interested readers. Finally, I am sure there is fruitful research still to be undertaken on this topic beyond that presented in this volume, but it will have to fall to others to carry on the pursuit of those studies. A few words on methodology should be given. Economics provides the guiding theoretical tools for most of the investigations in this book. Economics is a method, not a doctrine. It assumes that individuals systematically try to do what they think is best for themselves, whatever that may be, in a consistent and rational manner in the face of changing constraints that are observable obstacles outside the individual’s immediate control. Changes in constraints lead to systematic changes in observed behavior, as individuals adapt to the new constraints in the endless search for what they think is best for themselves. In short, people make tradeoffs. They assess a potentially realizable outcome against its next best feasible alternative, given the current constraints, and choose one path, thereby sacrificing other paths, namely choosing the path which they deem is best for themselves. The gap between the path chosen and the next best path foregone measures the net gain to, or value of, the observed choice outcome. When constraints change, the tradeoffs that people make change. One important task of economics is to identify those tradeoffs and how people systematically alter their tradeoffs as constraints change. The goal is not to prove that these assumptions or ideas are true or that they are doctrinal conclusions to be reached or preached. They are just ideas (fairy stories) to be used as a method for finding complex patterns among observables, i.e. to discover new history. Economic theory as a set of ideas is just a method that is more or less useful at finding new patterns among observables. It is the actual finding of such new patterns which can be empirically tested and verified that is the measure of the value and success of using economic methodology. Part of the economist’s task is to identify what the relevant observable constraints are that affect the observable choices or behavioral outcomes to be studied—in this case the behavioral choices by individuals involved in the migration and servitude process involving Germans coming to America in the eighteenth and early nineteenth centuries—and then chart as best as can be done how those constraints differed across space and over time. Important constraints could include the prices of all sorts of goods (such as passage fares or remunerations

Preface   xxiii for having a particular skill), personal income and wealth, information, laws, weather, risks, stomach capacity, health, physical stature, skills and education, age, gender, family composition, shipping technologies, the competition for the same desired outcome, the time needed to achieve or consume a given good or pursue a given behavior, and so on. Much of the book, especially the first half focusing on migration in general, is about identifying and sorting out what the constraints were that immigrants, shippers, and immigrant-­employers faced. The real value of economic reasoning, however, is identifying suspected systematic associations of observed differences in behaviors with observed differences in constraints. The economist’s task is to pursue these associations and systematically test for their presence and magnitude. Often this requires quantitative data and formal statistical procedures to disentangle effects, i.e. to isolate each effect from all the others by holding other potentially conflating influences constant. This is done in the service of history. Uncovering systematic patterns in the data not readily apparent to the naked eye is the creation of new history. Much of the book, especially the later part on servitude, uses economic reasoning to find new patterns of observations, test for their presence and magnitude, and thereby create new history and a new historical understanding of German immigration and servitude in early America. This does not mean that literary sources are not used or are not important. In what follows, a substantial amount of literary and textual sources are employed, but they are used more to illustrate what is going on, as an additional guide for how to model the behavior in question, and to enliven the presentation. They are seldom used as proof by themselves, but only as supporting evidence or as a check on outcomes found in the quantitative data. Literary sources are historical evidence, not history. They require careful assessment of their representativeness and their author’s intentions. In what follows, some of the prime literary sources that scholars have often used unquestioningly in the past to describe German immigration and servitude are shown to be unrepresentative, biased, and sometimes downright erroneous. I strived throughout the book to achieve an effective blend of literary and quantitative evidence both for presentation purposes and to achieve a robust pattern of evidential support. Economic theory can be articulated in words, graphs, or equations. One form is no less legitimate or rigorous than another, but one form can be more judicious or clearer in presentation than another depending on the task at hand. In what follows, I sometimes present economic theory linking changing behaviors with changing constraints in the guise of graphic models, sometimes in mathematical models, but most often in words. I try to keep the formalization to a minimum and use words to explain the modeling ideas in their most intuitive form. Even where no formal economic model or reasoning seems present, it may still lurk between the lines guiding how the discussion is organized and what behaviors are considered important. Finally, economic theory in the guise of market analysis is used often in what follows. Markets are not places or things, but theoretical concepts for organizing a mass of individual observed behaviors, those that have some commonality of

xxiv   Preface interest or purpose, into a coherent set of forces and outcomes. Markets are heuristic tools for organizing ideas in order to predict patterns in the data. They are not realistic representations of things that one can see. Markets embed the economic ideas discussed above formally and systematically. They help identify additional constraints that affect individuals, and the conditions under which these constraints may or may not change, when groups of individuals all pursue similar desires in the same relevant vicinity. For example, in the transatlantic shipping market, an increase in migration (an increase in the demand for shipping by immigrants) will not change passage fares (a constraint) if there is excess capacity in the transatlantic shipping market (a perfectly elastic supply of shipping). Assessing whether markets exist or work is not the goal. Such assessments are largely misguided. Markets are theoretical ideas, not things. They always exist and work as ideas in our heads. The goal is to use the idea of markets to identify potential patterns among observable behaviors that can then be tested in the data. Markets are simply more or less useful at finding such patterns (Grubb 2001). I hope in what follows the reader will see that markets are a useful tool for uncovering new and exciting patterns of behavior, i.e., for uncovering some new history about German immigrants arriving in eighteenth- and early nineteenth-­century America. Acknowledgments are listed separately in each chapter’s opening footnotes. I also want to thank Thomas Sutton, Emily Kindleysides, and Brian Holt of Taylor and Francis (Routledge) for their encouragement and assistance in assembling this project, and my wife, Tracy McQueen, for her patience, encouragement, and editorial assistance on the previously unpublished materials. I wish to thank the University of Delaware and department colleagues for their comments and encouragements, and the university administration, especially my department chair, Saul Hoffman, for providing the resource support and department environment that allows research projects like this to thrive. I would be remiss if I did not also thank my thesis advisors from long ago, Gary Becker, Robert Fogel, and David Galenson, for their economic insight, encouragement, and motivating criticism at the inception of this project in the early 1980s. Finally, I want to acknowledge the giants of German immigration and servitude historiography from the distant past, namely Frank R. Diffenderffer, Karl F. Geiser, Cheesman A. Herrick, Abbot E. Smith, and Ralph B. Strassburger. Without their published research in English, this book would not have been possible, or would have been more difficult to complete (Diffenderffer 1897, 1899; Geiser 1901; Herrick 1926; Smith 1947; Strassburger 1934). Farley Grubb Newark, Delaware, USA December 2010

Preface   xxv

References Boerstler, C. (1908) “Autobiography of Bavarian immigrant,” Journal of America History, 2: 457–70. Diffenderffer, F. R. (1897) “The German exodus to England in 1709,” Part 1 of “The German influence in its settlement and development,” Pennsylvania German Society, 7: 255–414. Diffenderffer, F. R. (1899) “The German immigration into Pennsylvania through the port of Philadelphia, and ‘the redemptioners’,” Part 7 of “The German influence in its settlement and development,” Pennsylvania German Society, 10: 1–328. Fenske, H. (1980) “International migration: Germany in the eighteenth century,” Central European History, 13: 332–47. Fertig, G. (1994) “Transatlantic migration from the German-­speaking parts of central Europe, 1600–1800: proportions, structures, and explanations,” in N. Canny (ed.) Europeans on the move: studies on European migration, 1500–1800, New York: Oxford University Press. Fogleman, A. S. (1996) Hopeful journeys: German immigration, settlement, and political culture in colonial America, 1717–1775, Philadelphia: University of Pennsylvania Press. Galenson, D. W. (1981) White servitude in colonial America, New York: Cambridge University Press. Geiser, K. F. (1901) Redemptioners and indentured servants in the colony and commonwealth of Pennsylvania, New Haven, CT: Tuttle, Morehouse, and Taylor Company. Grabbe, H. J. (1989) “European immigration to the United States in the early national period, 1783–1820,” Proceedings of the American Philosophical Society, 133: 190–214. Grubb, F. (1984) “Immigration and servitude in the colony and commonwealth of Pennsylvania: a quantitative and economic analysis,” Chicago: Unpublished PhD Dissertation, University of Chicago. Grubb, F. (1985) “The market for indentured immigrants: evidence on the efficiency of forward‑labor contracting in Philadelphia, 1745–1773,” Journal of Economic History, 45: 855–68. Grubb, F. (1992) “The long-­run trend in the value of European immigrant servants, 1654–1831: new measurements and interpretations,” Research in Economic History, 14: 167–240. Grubb, F. (1999) “Lilliputians and Brobdingnagians, stature in British colonial America: evidence from servants, convicts, and apprentices,” Research in Economic History, 19: 139–203. Grubb, F. (2000a) “The transatlantic market for British convict labor,” Journal of Economic History, 60: 94–122. Grubb, F. (2000b) “The statutory regulation of colonial servitude: an incomplete-­contract approach,” Explorations in Economic History, 37: 42–75. Grubb, F. (2001) “Social science versus social rhetoric: methodology and the Pacific labor trade to Queensland, Australia,” Historical Methods, 34: 5–36. Grubb, F. (2010a) “Testing for the economic impact of the U.S. Constitution: purchasing power parity across the colonies versus across the States, 1748–1811,” Journal of Economic History, 70: 118–45. Grubb, F. (2010b) “U.S. land policy: founding choices and outcomes, 1781–1802,” in D. A. Irwin and R. Sylla (eds.) Founding choices, Chicago: University of Chicago Press.

xxvi   Preface Herrick, C. A. (1926) White servitude in Pennsylvania, Philadelphia: John Joseph McVey. Klepp, S. E., Grubb, F., and Pfaelzer de Ortiz, A., eds. (2006) Souls for sale: two German redemptioners come to revolutionary America: the life stories of John Frederick Whitehead and Johann Carl Büttner, University Park, PA: Pennsylvania State University Press. “Munstering [mustering] book for the Britannia Capt. James Peters from Rotterdam 3rd July 1773,” Harrisburg, PA: Unpublished manuscript, AM 209, Historical Society of Pennsylvania. “Record of indentures of individuals bound out as apprentices, servants, etc. and of German and other redemptioners in the office of the Mayor of the City of Philadelphia, October 3, 1771, to October 5, 1773,” (1907) The Pennsylvania-­German Society, 16: 4–325. Roeber, A. G. (1993) Palatines, liberty, and property, Baltimore: Johns Hopkins Press. Sabean, D. W. (1990) Property, production, and family in Neckarhausen, 1700–1870, New York: Cambridge University Press. Smith, A E. (1947) Colonists in bondage, New York: W. W. Norton. Strassburger, R. B. (1934) Pennsylvania German pioneers, 3 Vols., Norristown, PA: Pennsylvania German Society. Wokeck, M. S. (1981) “The flow and composition of German immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, 105: 249–78. Wokeck, M. S. (1999) The trade in strangers: the beginnings of mass migration to North America, University Park, PA: Pennsylvania State University Press. Wolf, S. G. (1976) Urban village: population, community, and family structure in Germantown, Pennsylvania, 1683–1800, Princeton, NJ: Princeton University Press.

1 Introduction*

The economic, cultural, and political life of America has been often shaped by immigration. This was never more the case than in the colonial and early national periods. To comprehend the evolution of American society, the immigrant experience must be understood. Who immigrated and why, what were  their social origins and characteristics, what resources did they bring with them, what was their journey like, how did they finance that journey, and how did they fit into established American society? These questions guide the investigation that follows. In particular, this book focuses on the migration experience of the Germans, the largest non-­English-speaking European migration to English America during the eighteenth and early nineteenth centuries. The integration of foreign speakers into American society has frequently caused public concern and affected American immigration policy. Germans were the first major wave of European foreign speakers entering English America. Their migration experience offers historical perspective on how Americans dealt with such immigrants and how such immigrants coped with their entry into American society. Subsequent waves of foreign speakers entering English America have continued to the present day. Understanding the past may help us better comprehend the present and deal with the future. Immigrants molded early America. Their nature and the nature of the institutions they brought with them, when filtered by the constraints of their new American environment, formed the basis of American culture, economy, and political values (Fischer 1989; Roeber 1993). The character and social origins of early  European-­Americans, however, have been often disputed.1 Their occupational and family composition, levels of education, proportion who came as bonded servants, among other characteristics, have yet to be conclusively established. America as a melting pot of nations began in the colonial period. While traces of many nationalities can be found, the dominant group among European-­ Americans of non-­English-speaking ancestry were the Germans. Table 1.1 presents the percentage distribution of the white population of the United States by ethnicity and state from the 1790 census—the first U.S. Federal census. While

2   Introduction Table 1.1 Percentage distribution of the white population of the United States by nationality and state from the 1790 census (%) State

English

Scottish

Irish

German

Othera

Maine New Hampshire Vermont Massachusetts Rhode Island Connecticut New York New Jersey Pennsylvania Delaware Maryland Virginia North Carolina South Carolina Georgia

60.0 61.0 76.0 82.0 71.0 67.0 53.0 47.0 35.3 60.0 64.5 68.5 66.0 60.2 57.4

4.5 6.2 5.1 4.4 5.8 2.2 7.0 7.7 8.6 8.0 7.6 10.2 14.8 15.1 15.5

11.7 7.5 5.1 3.9 2.8 2.9 8.1 9.5 14.5 11.7 12.3 11.7 11.1 12.8 15.3

1.3 0.4 0.2 0.3 0.5 0.3 8.2 9.2 33.3 1.1 11.7 6.3 4.7 5.0 7.6

22.5 24.9 13.6 9.4 19.9 27.6 23.7 26.6 8.3 19.2 3.9 3.3 3.4 6.9 4.2

U.S. averages

60.9

8.3

9.7

8.7

12.4

Source: U.S. Bureau of the Census (1975: 1168). Note a This category is composed of Dutch, French, Swedes, Swiss, and non-assignable individuals. The non-assignable group dominates this category in all states except New York, New Jersey, and Delaware, where the Dutch comprise the majority of this category.

the accuracy of the ethnic designators in this census has been disputed, the general magnitudes and their relative patterns are still informative.2 Nothing as comprehensive exists for the colonial period. White Americans of English ancestry were the majority in every state except New Jersey and Pennsylvania. Those of German ancestry were a third of the population of Pennsylvania—almost equal to the English percentage, and approximately 10 percent of the population in the neighboring states of New York, New Jersey, and Maryland. If Dutch, Swiss, and Swedes are included in the German category, this group would likely be the largest single ethnic group in Pennsylvania and the second largest in New York, New Jersey, and Maryland. Elsewhere, the percentage of white Americans of German ancestry lagged well behind those of English, Scottish, and Irish ancestry. Table 1.2 transforms the approximate ethnic percentages from the 1790 census into their raw population totals. Just over half of German-­Americans were concentrated in Pennsylvania. Another 25 percent were in the neighboring states of New York, New Jersey, and Maryland. The rest were scattered across the southern states. Again, if the Dutch, Swiss, and Swedes are included in the German category, the concentration of Americans of Germanic ancestry in the middle states would be higher. The middle states from New York through Maryland, principally Pennsylvania, were overwhelmingly the location of German-­ Americans by the end of the eighteenth century.

Introduction   3 Table 1.2 The raw distribution of the white population of the United States by nationality and state from the 1790 census State

English

Maine New Hampshire Vermont Massachusetts Rhode Island Connecticut New York New Jersey Pennsylvania Delaware Maryland Virginia North Carolina South Carolina Georgia U.S. totals

Scottish

Irish

German

Othera

57,600 86,620 64,600 305,860 46,150 156,110 166,420 79,900 149,672 27,600 134,805 302,770 190,080 84,280 30,422

4,320 8,804 4,335 16,412 3,770 5,126 21,980 13,090 36,464 3,680 15,884 45,084 42,624 21,140 8,215

11,232 10,650 4,335 14,547 1,820 6,757 25,434 16,150 61,480 5,382 25,707 51,714 31,968 17,920 8,109

1,248 568 170 1,119 325 699 25,748 15,640 141,192 506 24,453 27,846 13,536 7,000 4,028

21,600 35,358 11,560 35,062 12,935 64,308 74,418 45,220 35,192 8,832 8,151 14,586 9,792 9,660 2,226

1,882,889

250,928

293,205

264,078

388,900

Source: Derived from the U.S. Bureau of the Census (1975: 25–36, 1168). Note a This category is composed of Dutch, French, Swedes, Swiss, and non-assignable individuals. The non-assignable group dominates this category in all states except New York, New Jersey, and Delaware where the Dutch comprise the majority of this category.

This concentration of German-­Americans in the middle states was the direct outcome of prior immigration. Table 1.3 estimates the contribution of German immigrants to the population growth of the middle colonies of Pennsylvania, New Jersey, and Delaware.3 This contribution was substantial, peaking at 30 percent during the decade of the 1750s. Considering that German immigrants were more family oriented than British and Irish immigrants, the indirect Table 1.3 Estimated population growth in the Delaware valley colonies, 1730–75 Decade

Population increase in Delaware, New Jersey, and Pennsylvania

Estimated German immigration per decade as a percentage of population growth

1730–40 1740–50 1750–60 1760–70 1770–75

58,493 62,883 91,003 82,218 59,666

20.74 27.15 30.42   8.82   7.89

Sources: Derived from U.S. Bureau of the Census (1975: 1168); Wokeck (1981: 260–1); chapter 2, figure 2.1. Notes The increase from 1770–80 was divided in half to derive the 1770–5 estimate. This was necessary because German immigration stopped at the America revolution in 1775.

4   Introduction contribution of German immigration through American-­born offspring to subsequent population growth was also higher than for other ethnic groups (Chapters 5 and 9). The impact of German immigration on the population of Pennsylvania was noted at the time. For example, in 1728 Governor Thomas estimated that Germans constituted three-­fifths of the population (Jacobs 1897: 148). By 1760, Pennsylvanians of German heritage were estimated to still be in the majority (Diffenderffer 1899: 97–106). By 1790, they were still the dominant ethnic group in Pennsylvania, though no longer a majority of the population (Table 1.1; footnote 2). Both the long-­run trend and the short-­run fluctuation in the region’s economic activities were linked to the ebb and flow of German immigration (Nash 1979: 161–97, 233–63; Smith 1981). For example, the variance in acreage warranted (brought under cultivation) in Pennsylvania during the eighteenth century closely matches the variance in German immigration through the port of Philadelphia (Lemon 1972: 68; Chapter 2, Figure 2.1). The regional labor market was dependent on the continued arrival of German immigrants to provide a fresh source of hired labor. German immigration was an essential part of the cultural and economic development of this region during the eighteenth and early nineteenth centuries. Sustained efforts to recruit German immigrants to British North America began with the founding of Pennsylvania by William Penn in 1682. Penn himself was half Dutch and had made several apostolic journeys to the Rhineland prior to receiving his colonial grant in 1682. He naturally turned to Germans to help populate his new colony. He established agents in Holland, the most important being Benjamin Furly, who spread enticing accounts of the new colony, some written by Penn himself. Penn also sold Pennsylvania lands to merchants in Krefeld and Frankfort to help recruit German colonists. While some German colonists went, their numbers were small, under 200, compared with what would follow in the eighteenth century. These early immigrants, however, provided favorable accounts, which influenced Germans in later years to choose Pennsylvania as their favored New World destination.4 German emigration to British North America began in earnest in 1709 when approximately 14,000 desperately poor Rhinelanders, escaping starvation and the ravages of the War of the Spanish Succession, fled to Holland and then to England. They petitioned the British government for relief and assistance. With this assistance some 3,800 were transported to Ireland, 3,200 to the colony of New York, 650 to the Carolina colonies, and some 2,000 returned to Germany. The rest, those who had not perished during the ordeal, made their home in England.5 While Pennsylvania was not a serious contender for this initial wave of German immigrants to English America, it became the ultimate beneficiary. The Germans who arrived in New York experienced hard times and ill treatment, which prompted many to move overland to Pennsylvania. The letters they sent back to Germany diverted German emigrants bound for the New World to Penn-

Introduction   5 sylvania thereafter. The Swedish botanist and traveler, Peter Kalm, while visiting the colonies in 1748, recounted these events in his journal: The Germans not satisfied with being themselves removed from New York, wrote to their relations and friends and advised them if ever they intended to come to America not to go to New York, where the government had shown itself so inequitable. This advice had such influence that the Germans, who afterwards emigrated in great numbers to North America, constantly avoided New York and kept going to Pennsylvania. (Kalm 1937: vol. 1, 142–3) After 1709, German immigration to the Delaware valley increased. Many Germans who ultimately made their homes in Maryland, New Jersey, Virginia, and Delaware landed initially in Pennsylvania. Philadelphia became the primary port of entry and remained so throughout the eighteenth and into the early nineteenth century (Chapter 2). As such, the surviving records from the port of Phil­ adelphia, and from Pennsylvania’s colonial and commonwealth governments, form the evidential basis for studies of German immigration to America. Concentrating on Pennsylvania captures both the dominant and representative experience of Germans immigrating to America in this period. The proliferation of Germans in Pennsylvania alarmed the English inhabitants and attracted the attention of the provincial council and the governor. Addressing the provincial council in 1717, Governor William Keith, stated, that great numbers of foreigners, from Germany, strangers to our language and Constitutions, having lately been imported into this Province, daily dispersed themselves immediately after Landing, without producing any Certificates, from whence they came or what they were; & as they seemed to have first Landed in Britain, & afterwards to have left it without any License from the Government, or so much as their knowledge, so in the same manner they behave here, without making the least application to himself or to any of the Magistrates; That as this Practice might be of very dangerous Consequences, since by the same method any number of foreigners from any nation whatever, as well Enemys as friends, might throw themselves upon us; The Governor, therefore, thought it requisite that this matter should be considered . . ., & ’tis Ordered thereupon, that all masters of vessels who have lately imported any of these foreigners be summoned to appear at this Board, to render an Account of the Number and character of their Passengers respectively from Britain; That all those who are already Landed be required by a Proclamation, to be issued for that purpose, to Repair within the space of one month to some Magistrate, particularly to the Recorder of this City [Philadelphia], to take such Oath appointed by Law as are Necessary to give assurances of their being well affected to his Majesty and his government (Diffenderffer 1899: 34–6; Strassburger 1934: vol. 1, xvii–xviii)

6   Introduction The provincial council did not permanently implement this proposal until 1727, when it required shippers to submit passenger lists of German immigrants and oaths of allegiance to be administered to these German arrivals.6 Continuing fears that German immigration would suborn English culture, language, and political control in Pennsylvania kept these requirements intact and diligently administered throughout the remainder of the colonial era. Benjamin Franklin expressed these continuing fears in letter to a friend in 1753: I am perfectly of your mind, that measures of great temper are necessary touching the Germans, and I am not without apprehensions, that, through their indiscretion, or ours, or both, great disorder may one day arise among us. Those who came hither are generally the most stupid of their nation, and as ignorance is often attended with great credulity, when knavery would mislead it, and with suspicion when honesty would set it right; and few of the English understand the German language, and so cannot address them either from the press or pulpit, it is almost impossible to remove any prejudice they may entertain. . . . Not being use to liberty, they know not how to make modest use of it. . . . They are under no restraint from ecclesiastical government; they behave, however, submissively enough at present to the civil government, which I wish they may continue to do, for I remember when they modestly declined intermeddling with our elections; but now they come in droves and carry all before them, except in one or two counties. Few of their children in the country know English. . . . The signs in our streets have inscriptions in both languages, and some places only in German. They begin, of late, to make all their bonds and other legal instruments in their own language, which (though I think it ought not to be), are allowed in our courts, where the German business so increases, that there is continual need of interpreters, and I suppose in a few years, they will also be necessary in the Assembly, to tell one-­half of our legislators, what the other half says. In short, unless the stream of importation could be turned from this to other colonies, as you very judiciously propose, they will soon outnumber us, that all the advantages we will have, will in my opinion, be not able to preserve our language, and even our government will become precarious. (Diffenderffer 1899: 108–10; Sparks 1838: vol. 7, 71–3) It was exactly because the Germans were such a large foreign-­speaking group, thereby threatening English culture and political control, that Pennsylvania authorities kept a more thorough documentation of German immigration than for any other arriving group. Prior to 1727, the exact number of German arrivals is difficult to estimate, but after 1727 a fairly exact accounting can be made. These surviving passenger lists and oaths of allegiance provide the most detailed, comprehensive, and continuous source of information available for any single migration stream to British North America during the eighteenth century. Analyzing German immigration to Pennsylvania, therefore, may offer insights into the migration experience for other ethnic groups for whom records were not as well kept.7

Introduction   7 Upon landing in Philadelphia, the most likely fate awaiting a German immigrant was to spend a number of years as a bonded servant. Being sold into servitude was the principal means by which European immigrants secured passage to British colonial America when they had insufficient cash to meet this expense. This was a prominent feature of European immigration to the New World throughout the seventeenth and eighteenth centuries. Without transatlantic servitude, America would not have been populated as fast or grown as rapidly as it did. The continual arrival of European servants to Pennsylvania and New Jersey in the eighteenth century was one reason slavery never took root north of the Chesapeake.8 This fact, in turn, influenced the economic, cultural, and political course of the U.S. through the nineteenth century. During the colonial period between one-­half and two-­thirds of all European transatlantic immigrants arriving south of New England came as servants. This practice of financing the transatlantic journey through servitude began shortly after the founding of Jamestown, around 1620, and continued through 1820.9 Germans were similar to the other major ethnic groups migrating to British colonial America, namely the English, Scottish, and Irish, in relying on a market for voluntary servant labor contracts in America to finance that migration. In the latter half of the eighteenth century, Philadelphia was the largest market for European immigrant servants. The reliance on white servant labor was eclipsed by black African slaves in colonies south of Pennsylvania. Colonies north of Pennsylvania never attracted many European immigrant servants.10 Studying immigrant servitude in Pennsylvania and New Jersey provides insight into the last major episode of European servitude in North America. The Philadelphia servant market was also unique among British possessions in America in that the local government periodically kept detailed records of the auction of arriving servants (Chapter 10). In fact, the eighteenth- and early nineteenth-­century Philadelphia servant market provides the only evidence in large enough quantities to allow for a systematic analysis of how servant contracting and immigrant servant auctions in America worked. As such, studying immigrant servitude in the Delaware valley has wider implications regarding how the institution of European transatlantic servitude functioned throughout English America. The existence of major bodies of evidence on immigrant servant auctions in Philadelphia is, in part, a result of the size of German immigration. The political power garnered by German-­Pennsylvanians, through associations such as the German Society of Pennsylvania as well as through the German press in America, was instrumental in causing the Pennsylvania government to establish a registry of German immigrant servant contracts in 1785—the last entry being in 1831. This was done primarily to protect immigrants from fraudulent transactions. This registry is the largest, most continuous, and most important body of information on European immigrant servant contract sales in the New World.11 America as a destination for German emigrants was fairly insignificant in the  eighteenth century—more of a sideshow than a prominent lure. Internal migration within Germany, seasonal and permanent labor migration to the

8   Introduction Netherlands, recruitment by the Dutch East India Company (Verenigde Nederlandse Geoctroyeerde Oostindische Compaigne or VOC) for workers on their ships going to, and in their settlements in, the East Indies (Indonesia), and migration to lands in eastern Europe and Russia involved substantially more German emigrants than did emigration to America.12 David Eltis (2002: 17) concluded that for “every German that moved to the Americas, nine migrated in the opposite direction.” At best, German emigration to America in the eighteenth century was only a fifth of that going to eastern Europe. Germans migrated southeast to the Balkans, east to Hungary, Poland, and the Crimea, but predominantly north and east into Russia, where Germans became the majority of new settlers. Attracted by Empress Catherine the Second’s lenient immigration policy for Germans and her promise to grant religious tolerance, German settlers came in the thousands to homestead in Russia. Between 1764 and 1774 over 100 German colonies were founded in the middle and lower Volga region. These were in addition to the 100,000 Germans who already lived in the Russian empire by the mid-­eighteenth century. Similarly, the VOC, using recruiting agents that scoured Germany, sent many Germans overseas to work in the East Indies. The numbers they recruited at times rivaled those migrating to eastern Europe. The likelihood of running into a recruiter for the VOC in Germany was far greater than running into a recruiter for emigrant shippers headed to America.13 Many indigenous factors are thought to have propelled emigration out of Germany in the eighteenth and early nineteenth centuries. In general, lands from whence Germans emigrated, primarily the southern Rhineland, were poorer than other central European regions. Poverty was the consequence of a variety of causes: the lingering aftereffects of the Thirty Years War, a heavy load of intrusive bureaucracies and high taxes, rapidly rising populations, natural harvest failures caused by draughts, and incessant wars that led to death, disease, and the destruction of property and crops. Inheritance laws in some Germanic areas that disallowed the division of property and favored the eldest son were another factor that contributed to a steady flow of emigrants. Rising urbanization, increasing faster than the natural rate of increase, meant that migration from the countryside to nearby cities was on the increase. This, in turn, made the next step of external emigration psychologically easier. Finally, religious freedom was still an issue in the eighteenth century. The major Christian denominations— Roman Catholic, Lutheran, and Reformed—were mutually intolerant. Minority faiths—Mennonite, Amish, Moravian, and other pietist sects—had little or no protection under the laws of any European state. Both religious persecution and persistent poverty caused thousands to migrate.14 In eighteenth-­century Germanic states, patriarchal family structures predominated both by law and custom. Fathers ruled over their wives and children. The urge to migrate was often affected by family considerations, such as by overbearing fathers or by abusive step-­fathers. Family circumstances, however, were not a uniquely uniform determinant in the decision to emigrate as the contrasting memoirs of John Frederick Whitehead and Johann Carl Büttner demonstrate.15

Introduction   9 When German emigrants chose America as their destination, they overwhelmingly chose Philadelphia as their port of entry, at least from the 1720s into the 1820s. Given that most Germans on the move in this period did not go to America, the choice to go to America had to be conscious and deliberate, not just the movement of people mindlessly swept away in the general direction of the herd. It is that conscious deliberate choice of America, that notion that something must be different about these immigrants and this choice, that provides meat for analytical investigation and gives meaning to the study of German immigration to America. Questions about immigration naturally lend themselves to answers that require quantification. What were the magnitude and flow of German immigration, the characteristics and social origins of these immigrants, the pricing and profits of German immigrant servant contracts, the voyage and post-­voyage experiences of these immigrants, and so on? Without reasonable samples of quantifiable data, the answers to such questions would rely only on literary sources and contemporary opinions—evidence that is prone to exaggerate the atypical, be unrepresentative, and contain embedded biases. Literary sources are historical evidence, not history. Quantitative research is not without its pitfalls. The statistical misuse and misrepresentation of quantifiable evidence and the failure to adequately control, sort, and assess the biases in the data, as well as explore the robustness of the findings, are ever-­present problems. Every other researcher may aspire to praise, the quantitative researcher can only hope to escape reproach, and even this negative recompense has been granted to very few. Yet, quantitative research is the only way to narrow the range of possibilities and separate history from historical fiction. The analysis that follows employs detailed quantitative evidence and advanced statistical techniques for processing that evidence. It also organizes and interprets this evidence from an economic point of view. Economic theory provides a simple, consistent, and integrated way of sorting historical evidence and identifying patterns in the data that are not otherwise readily obvious.16 The boundaries of this study are circumscribed by surviving bodies of quantifiable data. The principal ones are the collections of German passenger lists and servant auction records for the colony and state of Pennsylvania. German passenger lists from 1727 through 1808 are collected in Strassburger (1934). After 1800, German passenger lists can be gleaned from the “Philadelphia Customs House records.” Four bodies of servant auction records for the port of Philadelphia have survived, namely for 1745–6, 1771–3, 1785–1804, and 1817–31—the last two being exclusively for German arrivals (Chapter 10). These data sets were merged by hand—name by name, ship by ship, and then often merged again by hand with other ancillary data sets to form the evidentiary basis for the research that follows. The book is organized into two major parts, each with several subsections. Part I covers the German immigrant experience from 1727 to 1835. The first half

10   Introduction of Part I investigates the magnitudes, causes, conveyances, and conditions of that migration. The flow of immigrants is charted. The causes of emigration and the availability of transatlantic immigrant shipping is assessed. Whether shipping markets were monopolized is addressed. Finally, the health outcomes of the transatlantic voyage are measured. The second half of Part I investigates the immigrants themselves, including their age, gender, occupation, and family compositions. This section also addresses how the act of migration itself influenced the literacy levels of German immigrants and the subsequent education of their offspring. Part II explores German immigrant servitude in three sections. The first section of Part II establishes the patterns of servitude among the immigrants, namely the percentage of immigrants entering servitude in America, the geographical and occupational diffusion of servants among (the demand for servants by) American employers, the nature of the surviving servant auction records, and the size of the female portion of the servant market. The second section of Part II analyzes the market for immigrant servants, including how the contracting method was determined, whether shippers earned excessive profits on transporting servants, how servant contracts were priced in the American auction, and how German parents used their children in the servant market to finance the family’s migration. The third section of Part II documents when, and explains why, the market for German immigrant servants in the U.S. came to an end. Part III is a brief two-­chapter conclusion. Its primary purpose is to serve as an epilogue that contrasts the founding migration era of 1709–1835, the era dominated by immigrant servitude, with the subsequent era of mass migration of Germans to America that arose after 1835 and ended with World War I, an era dominated by free passengers and free immigrant labor.

Notes   * Adapted and updated from the preface and chapter 1 of Farley Grubb, “Immigration and servitude in the colony and commonwealth of Pennsylvania: a quantitative and economic analysis,” PhD Thesis, University of Chicago, 1984 [previously unpublished].   1 For example, see the debates and alternative views in Campbell (1978, 1979); Cohn (2009: 98–124); Diffenderffer (1899: 107–38); Galenson (1978, 1979).   2 For example, see Akenson (1984a, 1984b); McDonald and McDonald (1980, 1984); O’Brien (1908–9); Purvis (1984a, 1984b). These debates indicate that a higher proportion of Germans and a lower proportion of English were in the Pennsylvania population than shown in tables 1.1 and 1.2, thus making Germans the single largest ethnic group in Pennsylvania in 1790.   3 See also the analysis in Fogleman (1995: 3–10).   4 See Diffenderffer (1899: 5–20); Fenske (1980: 333); Geiser (1901: 8–17); Herrick (1926: 48–56); Köllman and Marschalck (1973: 512–13); Wokeck (1999: 45).   5 See Diffenderffer (1897: 255–414); Fenske (1980: 336); Jacobs (1897: 29–150); Knittle (1937); Statt (1995: 121–65).   6 For examples of these oaths, a discussion of their implementation, and the history of these records, see Strassburger (1934: vol. 1, vii–6); Wokeck (1981: 249–60).   7 See Diffenderffer (1899: 97–106); Geiser (1901: 23–7); Smith (1947: 320); Strassburger (1934: vol. 1, xix–xx); Wokeck (1981: 260; 1999: 45).

Introduction   11   8 See Galenson (1981a, 1984); Grubb (1992); Smith (1947); chapter 9. For a sampling of the various hypotheses about the transition from white immigrant servant labor to black African slave labor in colonial America, see Coelho and McGuire (1997); Dunn (1972: 47–83); Galenson (1981b); Grubb and Stitt (1994); Hanes (1996); Menard (1977); Morgan (1975: 295–315); Smith (1947: 26–35); Wood (1974); chapter 11. A common aspect in all these hypotheses is the role played by the character and supply of European immigrant bound labor.   9 See Galenson (1984); Smith (1947: 3–16, 335–7); chapters 9 and 18. 10 See note 8. 11 See Geiser (1901: 66–70); Herrick (1926: 305–6); chapter 10. 12 See Fenske (1980); Klepp, Grubb, and Pfaelzer de Ortiz (2006: 7–9, 174–7, 213–16); Walker (1964: 31–41, 250). 13 See Benz (1996: 47); Bruijn, Gaastra, and Schoffer (1979–87); Fenske (1980); Fertig (1994: 203); Gestrich (2000: 77); Hellie (2002: 317–18); Lucassen (1994: 153–91); Wellenreuther (2000: 3–35). 14 See Beiler (1997: 73–87); Berkner (1976: 71–95); de Vries (1984: 272–3); Fertig (1994: 192–235); Fogleman (1996: 15–65); Jütte (1996: 377–85); O’Reilly (1998: 101–121); Statt (1995); Wegge (1998, 1999); Wellenreuther (2000: 3–35); Wokeck (1999: 1–17). 15 See Berkner (1976: 71–95); Klepp, Grubb, and Pfaelzer de Ortiz (2006); Sabean (1976: 96–111; 1990). 16 See the preface and Grubb (2001) for more on my methodological interpretation of economic theory and how it is used as a tool for finding new patterns of behavior and discovering new history within quantitative labor and migration data.

References Akenson, D. H. (1984a) “Why the accepted estimates of the ethnicity of the American people, 1790, are unacceptable,” William and Mary Quarterly, 41: 102–19. Akenson, D. H. (1984b) “Commentary,” William and Mary Quarterly, 41: 125–9. Beiler, R. J. (1997) “Distributing aid to believers in need: the religious foundations of transatlantic migration,” Pennsylvania History, 64, no. 5: 73–87. Berkner, L. K. (1976) “Inheritance, land tenure and peasant family structure: a German regional comparison,” in J. Goody, J. Thirsk and E. P. Thompson (eds.) Family and inheritance: rural society in western Europe, 1200–1800, New York: Cambridge University Press. Benz, E. (1996) “Population change and the economy,” in S. Ogilvie (ed.) Germany: a new social and economic history. Volume II, 1630–1800, New York: Arnold. Bruijn, J. R., Gaastra, F. S., and Schoffer, I. (1979–1987) Dutch-­Asiatic shipping in the 17th and 18th centuries, vols. 1 and 2, The Hague: Nijhoff. Campbell, M. (1978) “Mildred Campbell’s response,” William and Mary Quarterly, 35: 525–40. Campbell, M. (1979) “Replies,” William and Mary Quarterly, 36: 277–86. Coelho, P. R. P. and McGuire, R. A. (1997) “African and European bound labor in the British new world: the biological consequences of economic choices,” Journal of Economic History, 57: 83–115. Cohn, R. L. (2009) Mass migration under sail: European immigration to the antebellum United States, New York: Cambridge University Press. de Vries, J. (1984) European urbanization 1500–1800, Cambridge, MA: Harvard University Press. Diffenderffer, F. R. (1897) “The German exodus to England in 1709,” part 1 of

12   Introduction “The  German influence in its settlement and development,” Pennsylvania German Society, 7: 255–414. Diffenderffer, F. R. (1899) “The German immigration into Pennsylvania through the port of Philadelphia, and ‘the redemptioners’,” part 7 of “The German influence in its settlement and development,” Pennsylvania German Society, 10: 1–328. Dunn, R. S. (1972) Sugar and slaves, New York: W. W. Norton. Eltis, D. (2002) “Introduction,” in D. Eltis (ed.) Coerced and free migration: global perspectives, Stanford: Stanford University Press. Fenske, H. (1980) “International migration: Germany in the eighteenth century,” Central European History, 13: 332–47. Fertig, G. (1994) “Transatlantic migration from the German-­speaking parts of central Europe, 1600–1800: proportions, structures, and explanations,” in N. Canny (ed.) Europeans on the move: studies on European migration, 1500–1800, New York: Oxford University Press. Fischer, D. H. (1989) Albion’s seed: four British folkways in America, New York: Oxford University Press. Fogleman, A. S. (1995) “Immigration, German immigration and 18th-century America,” in E. Reichman, L. J. Rippley, and J. Nagler (eds.) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­American Center. Fogleman, A. S. (1996) Hopeful journeys: German immigration, settlement, and political culture in colonial America, 1717–1775, Philadelphia: University of Pennsylvania Press. Galenson, D. W. (1978) “Middling people or common sort? The social origins of some early Americans reexamined,” William and Mary Quarterly, 35: 499–524. Galenson, D. W. (1979) “The social origins of some early Americans: rejoinder,” William and Mary Quarterly, 36: 264–77. Galenson, D. W. (1981a) White servitude in colonial America, New York: Cambridge University Press. Galenson, D. W. (1981b) “White servitude and the growth of black slavery in colonial America,” Journal of Economic History, 41: 39–48. Galenson, D. W. (1984) “The rise and fall of indentured servitude in the Americas: an economic analysis,” Journal of Economic History, 44: 1–25. Geiser, K. F. (1901) Redemptioners and indentured servants in the colony and commonwealth of Pennsylvania, New Haven, CT: Tuttle, Morehouse, and Taylor Company. Gestrich, A. (2000) “German religious emigration to Russia in the eighteenth and early nineteenth centuries,” in H. Lehmann, H. Wellenreuther, and R. Wilson (eds.) In search of peace and prosperity: new German settlements in eighteenth century Europe and America, University Park, PA: Pennsylvania State University Press. Grubb, F. (1992) “The long-­run trend in the value of European immigrant servants, 1654–1831: new measurements and interpretations,” Research in Economic History, 14: 167–240. Grubb, F. (2001) “Social science versus social rhetoric: methodology and the Pacific labor trade to Queensland, Australia,” Historical Methods, 34: 5–36. Grubb, F. and Stitt, T. (1994) “The Liverpool emigrant servant trade and the transition to slave labor in the Chesapeake, 1697–1707: market adjustments to war,” Explorations in Economic History, 31: 376–405. Hanes, C. (1996) “Turnover cost and the distribution of slave labor in Anglo-­America,” Journal of Economic History, 56: 307–29.

Introduction   13 Heavner, R. O. (1978) Economic aspects of indentured servitude in colonial Pennsylvania, New York: Arno Press. Hellie, R. (2002) “Migration in early modern Russia,” in D. Eltis (ed.) Coerced and free migration: global perspectives, Stanford: Stanford University Press. Herrick, C. A. (1926) White servitude in Pennsylvania, Philadelphia: John Joseph McVey. Jacobs, H. E. (1897) “The German emigration to America, 1709–1740,” Pennsylvania German Society, 8: 29–150. Jütte, R. (1996) “Poor and poverty relief,” in S. Ogilvie (ed.) Germany: a new social and economic history. Volume II, 1630–1800, New York: Arnold. Kalm, P. (1937) Travels in North America, 2 vols, New York: Wilson-­Erickson, Adolph B. Benson trans. and ed. Klepp, S. E., Grubb, F., and Pfaelzer de Ortiz, A., eds. (2006) Souls for sale: two German redemptioners come to revolutionary America: the life stories of John Frederick Whitehead and Johann Carl Büttner, University Park, PA: Pennsylvania State University Press. Knittle, W. A. (1937) Early eighteenth century Palatine emigration, Philadelphia: Dorrance and Company. Köllman, W. and Marschalck, P. (1973) “German Emigration to the United States,” Perspectives in American History, 7: 497–554. Lemon, J. T. (1972) The best poor man’s country, Baltimore: The John’s Hopkins Press. Lucassen, J. (1994) “The Netherlands, the Dutch, and long-­distance migration in the late sixteenth to early nineteenth centuries,” in N. Canny (ed.) Europeans on the move: studies on European migration, 1500–1800, New York: Oxford University Press. McDonald, F. and McDonald, E. S. (1980) “The ethnic origins of the American people, 1790,” William and Mary Quarterly, 37: 179–99. McDonald, F. and McDonald, E. S. (1984) “Commentary,” William and Mary Quarterly, 41: 129–35. Menard, R. R. (1977) “From servants to slaves: the transformation of the Chesapeake labor system,” Southern Studies, 16: 355–75. Morgan, E. S. (1975) American slavery, American freedom, New York: W. W. Norton. Nash, G. B. (1979) The urban crucible, Cambridge, MA: Harvard University Press. O’Brien, M. J. (1908–9) “First census of the United States. Some pointed comments on the manner of taking same and the results thereof. An interesting paper on a subject not heretofore touched by the society,” Journal of American Irish Historical Society, 8: 209–16. O’Reilly, W. (1998) “Conceptualizing America in early modern central Europe,” Pennsylvania History, 65, no. 5: 101–21. “Philadelphia Customs House records, 1800–1883,” Washington, DC: Unpublished manuscript held at the National Archives, microfilm #425, rolls 16–50. Purvis, T. L. (1984a) “The European ancestry of the United States population, 1790,” William and Mary Quarterly, 41: 85–101. Purvis, T. L. (1984b) “Commentary,” William and Mary Quarterly, 41:119–25. Roeber, A. G. (1993) Palatines, liberty, and property, Baltimore: Johns Hopkins Press. Sabean, D. W. (1976) “Aspects of kinship behavior and property in rural western Europe before 1800,” in J. Goody, J. Thirsk and E. P. Thompson (eds.) Family and inheritance: rural society in western Europe, 1200–1800, New York: Cambridge University Press. Sabean, D. W. (1990) Property, production, and family in Neckarhausen, 1700–1870, New York: Cambridge University Press.

14   Introduction Smith, A. E. (1947) Colonists in bondage, New York: W. W. Norton. Smith, B. G. (1981) “Struggles of the ‘lower sort’: the lives of Philadelphia’s laboring people, 1750–1800,” Los Angeles: PhD Dissertation, UCLA. Sparks, J. (ed.) (1838) The works of Benjamin Franklin 8 vols, Boston: Tappen and Whittmore. Strassburger, R. B. (1934) Pennsylvania German pioneers, 3 vols., Norristown, PA: Pennsylvania German Society. Statt, D. (1995) Foreigners and Englishmen: the controversy over immigration and population, 1660–1760, Newark, DE: University of Delaware Press. U.S. Bureau of the Census (1975) Historical statistics of the United States colonial times to 1970: bicentennial edition, parts 1 and 2, Washington D.C.: U.S. Government Printing Office. Walker, M. (1964) Germany and the emigration, 1816–1885, Cambridge, MA: Harvard University Press. Wegge, S. A. (1998) “Chain migration and information networks: evidence from nineteenth-­century Hesse-­Cassel,” Journal of Economic History, 58: 957–86. Wegge, S. A. (1999) “To part or not to part: emigration and inheritance institutions in nineteenth-­century Hesse-­Cassel,” Explorations in Economic History, 36: 30–55. Wellenreuther, H. (2000) “Contexts for migration in the early modern world: public policy, European migrating experiences, transatlantic migration, and the genesis of American culture,” in H. Lehmann, H. Wellenreuther, and R. Wilson (eds.) In search of peace and prosperity: new German settlements in eighteenth century Europe and America, University Park, PA: Pennsylvania State University Press. Wokeck, M. S. (1981) “The flow and composition of German immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, 105: 249–78. Wokeck, M. S. (1999) The trade in strangers: the beginnings of mass migration to North America, University Park, PA: Pennsylvania State University Press. Wood, P. (1974) Black majority, New York: W. W. Norton.

Part I

German immigration to America, 1709–1835

2 The flow of immigrants, 1727–1835*

For the late colonial and early republican periods, German-­American histori­ ography is driven by the magnitude, variance, and timing of German immigration.1 For Germans arriving through the port of Philadelphia, the quantifiable contents of surviving ship manifests from 1727 through 1835 are used to describe this immigration. The magnitude and origins of this migration, why Germans chose Pennsylvania, and the causes of the variance in this migration over time are explained. Aspects of the migration experience, such as the length of the journey, the season chosen for migration, and the availability of transatlantic shipping services, are also addressed. In 1728, Governor Thomas of Pennsylvania estimated that Germans constituted three-­fifths of the colony’s population (Jacobs 1897: 148). The Pennsylvania government, alarmed by this rising tide of German immigration, ordered ship captains importing foreigners, beginning in September of 1727, to submit passenger lists. Foreign male immigrants above the age of 15 were also required to sign two oaths—one of allegiance to the British Crown and one of abjuration to the authority of the Pope and the House of Stuart. While only a few captains’ passenger lists still exist, the oath lists—having been entered in bound books— have survived largely intact. It is from these lists that the most authoritative estimates of German immigration to colonial Pennsylvania have been derived.2 These lists provide one of the most comprehensive recordings of immigration to colonial America from the American side. Since only German adult males signed the oaths, the problem of estimating total German immigration from the oath lists amounts to inferring the number of women and children accompanying these adult male migrants. The Strassburger (1934: vol. 1) collection of these lists for the colonial period includes 324 ship arrivals, spans the years 1727 to 1775, and yields 26,016 names from the administered oaths. For 178 of the 324 vessels, a total accounting of passengers was also recorded. Comparing these total passenger figures with the number of signers on the corresponding oaths, Strassburger (1934: vol. 1, xxxi) found a ratio of five passengers for every two signers, implying a total immigration of 65,040 for this period. Marianne Wokeck (1981) refined Strassburger’s estimate by adjusting for those ships among the 178 who recorded the number of passengers in terms of

18   German immigration to America, 1709–1835 bodies as opposed to freights.3 In addition, Wokeck found 21 ships not enumerated by Strassburger. Only seven, however, had information on the number of passengers which could be added to the total. Finally, Wokeck allowed the ratio of passengers to oath signers to vary from a high of 3.7 in 1727 to a low of 2.0 in 1775, to account for the apparent decline in the proportion of families among German immigrants. Total German immigration for 1727–75 estimated by Wokeck was 79,107. Therefore, a reasonable conclusion would be that not more than 80,000 or less than 65,000 Germans immigrated to colonial Pennsylvania between 1727 and 1775. Another 26,261 Germans immigrated to North Amer­ ican through other ports, such as Baltimore, making the portion of German immigrants landing in Philadelphia about 75 percent of total German immigration to British North America in this period.4 The 79,107 Germans entering Pennsylvania through Philadelphia between 1727 and 1775 is substantial compared with the magnitude of other European groups arriving in British North America. While evidence on other migrations is sketchier, guesses about their magnitudes are informative. For example, the total number of British convicts transported to all of British North America between 1718 and 1775 was about 50,000. Total non-­convict English immigration to all of British North America numbered about 50,000 and non-­convict Irish and Scots immigration numbered about 100,000 in this period.5 Total European immigration to British North America between 1730 and 1775 has been estimated to be as high as 532,000 (Fogleman 1995: 5; 1996: 157). German immigration through Philadelphia accounts for 15 percent of that total. Turning to comparisons with other migrations into the Delaware valley, the estimated number of immigrants from all Irish ports landing on the Delaware between 1730 and 1775 is 52,000, far less than the 77,214 Germans arriving during these years. Total British emigration to the New World between December 1773 and March 1776, the most comprehensive British emigration record in the late colonial period and a localized peak in British emigration, was just below 10,000, with only 1,400 of this total going to Pennsylvania. By contrast, over a similar time span (December 1772 to March 1775) German immigration to Pennsylvania was 2,650 (Bailyn 1986: 92, 206–7; Wokeck 1999: 45–6, 172–3). By any measure, German immigration into the Delaware valley was substantial in the late colonial period. Germans were the single largest ethnic group arriving in the region from Europe, and their immigration was relatively large even compared with total European immigration to all of British North America. As such, German immigrants were a major contributing force to the population growth as well as to the economic, cultural, and social development of the Delaware valley during the middle 50 years of the eighteenth century (Fogleman 1995: 3–10; 1996: 69–153; Roeber 1993; Wolf 1976). As will be shown in Chapters 5 and 9, because German immigration was more family oriented than other migrations, their contribution to the subsequent population growth in the Delaware valley was even greater than what the relative magnitude of their migration alone would indicate.

The flow of immigrants, 1727–1835   19 Post-­revolution, Pennsylvania law required only one list of passengers from captains of vessels carrying German immigrants. This list was submitted to the Office of Registry of German Passengers. The surviving lists were collected by Strassburger (1934: vol. 3). They span from 1785 to the beginning of 1808 and list a total of 10,165 German immigrants. Strassburger, however, indicates that some passenger lists did not survive.6 The importance of missing lists can be ascertained by comparing the names from surviving passenger lists with the names of German immigrant servants recorded in “Book A of redemptioners.” Because German immigrant vessels carried a fair number of German servants who were required to register their labor contracts soon after landing, servant contracts which cannot be matched to the names of new immigrants on surviving passenger lists indicate the presence of missing or deficient lists.7 For 1785 through 1804 there are a total of 7,837 immigrants named in the Strassburger (1934: vol. 3) lists, of which 3,511 were matched to names in “Book A of redemptioners” yielding a servant percentage of 44.8.8 However, 433 new contracts in “Book A of redemptioners” could not be matched to the passenger lists. These non-­matches were caused either by the anglicized corruption of German names causing matches to be missed, or by the presence of missing or deficient passenger lists.9 If all 433 non-­matches are due to missing passenger lists, and if these missing ships imported servant and non-­servant German immigrants in the proportion found above, then an additional 967 immigrants must be added to the 7,837 found in Strassburger, yielding a total of 8,804 German immigrants for 1785–1804.10 A similar merging of information and cross-­corroboration of ships and passenger counts was made, in conjunction with the Strassburger (1934: vol. 3) lists and “Book A of redemptioners,” using the surviving health officers’ and pilots’ reports on vessels arriving at the port of Philadelphia for 1789–94 and 1793–1808, respectively, and the lists of passengers in the Philadelphia Customs House records for 1800–83 (“Account of vessels;” “Pilot’s report;” “Philadelphia Customs House records”). After merging all lists and eliminating redundant and overlapping information, total German immigration through the port of Philadelphia for 1785–1804 was 12,221 and for 1805–8 was 3,709—totaling 15,930 for 1785–1808. These final totals are 36 percent above that found in the Strassburger (1934: vol. 3) lists alone, 28 percent above that obtained by merging only the Strassburger lists with the names in “Book A of redemptioners,” and 27 percent above the most recent estimate in the prior published literature (Grabbe 1989: 192). For 1809–16, the only surviving passenger lists are from the Philadelphia Custom House records. They yield a total of 1,511 German immigrants landing in Philadelphia across these years. Most of this number comes from the 1,304 who arrived in 1816. Immigration from 1809 through 1815 was suppressed by the continental blockage during the Napoleonic Wars, those wars themselves, and the War of 1812. For 1817–35, three surviving records are combined to determine German immigration into the Delaware valley, namely the Philadelphia Custom House records, “Book C of redemptioners,” and the Federal passenger lists which began

20   German immigration to America, 1709–1835 in 1820. Merging these three sources in a cross-­corroborative fashion similar to that done for 1785–1808, and adding in German immigrant ships missing from these records found in alternative sources and newspaper reports, yields a total of 12,443 German immigrants landing in the Delaware valley between 1817 and 1835.11 Comparing the magnitude of this German immigration to the U.S. during the early republic with other European migrations is difficult because the evidence on other migrations is sketchier. The best comparison uses the “Philadelphia Custom House records.” For 1800–11, German immigration through Philadelphia was estimated here to be 9,331. The Custom House records report 19,895 arrivals in these years, with 2,775 of this total being from British ports and 3,402 being from Irish ports. Thus between 1800 and 1811 Germans were the single largest ethnic group among European immigrants landing in Philadelphia—comprising around 47 percent of total arrivals landing in Philadelphia in this period.12 For 1812–23, German immigration into the Delaware valley was estimated here to be 8,808. The Customs House records report 34,734 arrivals, with 9,191 of this total being from British ports and 12,375 being from Irish ports (some via Canadian ports). Thus, between 1812 and 1823 Germans were the third largest ethnic group landing in Philadelphia behind the Irish and the British—comprising only 25 percent of total arrivals to Philadelphia in this period. The years 1816–19, however, were considerably different from the years 1820–3. For 1816–19, a total of 8,305 German immigrants arrived in the Delaware valley compared with 22,662 total arrivals. Whereas for 1820–3 a total of 396 German immigrants arrived in the Delaware valley compared with 9,832 total arrivals (“Philadelphia Custom House records;” Chapter 18). The share of German immigrants among total arrivals to Philadelphia fell from 37 percent to 4 percent between 1816–19 and 1820–3. Beginning in 1820 the U.S. government required passenger lists from all captains landing immigrants at U.S. ports. While the lists from the 1820s suffer from undercounting, the relative ethnic patterns of this undercounting are not fully known (Bromwell 1969; Cohn 2009; Grubb 1990). As such, relative comparisons using this evidence may still be somewhat informative. For 1820–35, total German immigration through Philadelphia estimated here is 5,442. The Federal passenger lists report a total of 438,596 immigrants arriving in the U.S. in this period. German immigration through Philadelphia is only 1.2 percent of this total. Some of this small percentage is due to the lower share of Germans in total immigration post-­1819 compared with pre-­1820. Total German immigration (counting all arrivals from Germany, Prussia, Holland, and Switzerland) listed in the Federal passenger lists for 1820–35 is only 61,098, or 14 percent of total immigration (Bromwell 1969: 21–88; Chapter 18). Most of this small percentage, however, is due to Philadelphia losing its prominence as a port of entry. Between 1820 and 1835, only 10 percent of all immigrants to the U.S., and only 9 percent of all German immigrants to the U.S., entered through Philadelphia. German immigrants represented only 12 percent of all immigrants entering through the port of Philadelphia in these years. Not

The flow of immigrants, 1727–1835   21 until the rise of mass migration after 1835 does the share of Germans in total U.S. immigration increase again. Philadelphia, however, would never return to the level of prominence as a port of entry that it enjoyed pre-­1820.13 In summary, total German immigration (nonmilitary) into the Delaware valley is estimated to be 79,107 for 1727–75, zero for 1776–84, 12,221 for 1785–1804, 3,709 for 1805–8, 1,511 for 1809–16, and 12,443 for 1817–35. The grand total for 1727 through 1835 is 108,991. The colonial years account for 73 percent of this total. German immigration to the Delaware valley averaged 1,614 per year from 1727 through 1775, but only 586 per year from 1785 through 1835. The residence of origin or even the regions from whence Germans emigrated is not known for most immigrants. The most common identifying origin mentioned during the colonial period was the Palatinate. This, however, was often used as a blanket term by the English when referring to Germans in general.14 For the colonial period none of the passenger lists provide the origin for individual immigrants. Several lists, on the other hand, do mention some general regions of origin in their introduction. Those most frequently mentioned, after Palatinate, were Wurttemberg and Switzerland. Other origins mentioned were Alsace, Basil, Darmstadt, Durlach, Erbach, Franconia, Hapsburg, Hanau, Hesse, Isenberg, Lorain, Manheim, Nassau, Rothenstein, Swabia, Triers, Westphalia, and Zwezbreckt. One vessel from Hamburg listed its passengers as being from Hamburg, Hanover, and Saxony. Quantifying the distribution of immigrants among these origins is not possible with this evidence, though it appears that the middle and upper Rhine valley (southwest Germany) was the dominant region of emigration (Fenske 1980: 342; Walker 1964: 1–41). Genealogists have also shown that many German colonists originated from this region (Yoder 1980, 1981).15 Post-­revolution, 12 ships arriving between 1798 and 1808 list the origin of each passenger (Strassburger 1934: vol. 3).16 Two ships list both village and general region while four list only the village and six list only the general region of origin. Of the 1,240 immigrants whose origins were listed, 36 percent were from Wurttemberg or Baden, 13 percent from Switzerland, 10 percent from Hessen, 5 percent from the Hanover-­Hamburg region, and 27 percent had villages which could not be located with relative certainty. The general origin of German immigrants during the two decades after the American revolution, therefore, appears similar to that found for the colonial period, being dominated by the middle and upper Rhine valley. Eleven of the 12 ships recorded their port of departure; the odd ship listed 243 of its 245 passengers as being from Wurttemberg. For these 11 vessels, the distribution of immigrant origins by port of departure is reported in Table 2.1. The two decades after the American revolution witnessed a rise in the proportion of ships and immigrants sailing from German ports relative to Dutch ports. This appears to be a new emigration and not just a deflection of the traditional upper Rhine valley migration toward German ports following the French occupation of Holland and the west bank of the Rhine (Schlosser 1844: vols 5, 6, and 7). While firm conclusions are mitigated by the large number of unidentified origins and

22   German immigration to America, 1709–1835 Table 2.1 Regional distribution of German immigrants’ residence of origin by port of departure, 1798–1808 Regions of origin for German immigrants to Pennsylvania

Number of immigrants leaving through Dutch portsa

Number of immigrants leaving through German portsb

Wurttemberg-Baden Hessen Westfalen Pfalz Hanover-Hamburg Saxony Prussia Bavaria Switzerland Holland Othersc Unknown

199 81 18 17 8 7 13 21 161 9 8 306

23.5% 9.6 2.1 2.0 0.9 0.8 1.5 2.5 19.0 1.1 0.9 36.1

3 37 1 0 58 3 5 0 6 0 1 33

2.0% 25.2 0.7 0.0 39.5 2.0 3.4 0.0 4.0 0.0 0.7 22.5

Totals

848

100.0%

147

100.0%

Source: Strassburger (1934: vol. 3). Notes a There were five ships in total. Four sailed from Amsterdam, and one listed its port simply as Holland. b There were six ships in total. Two sailed from Hamburg, two sailed from Tonningen, one sailed from Bremen, and one sailed from Fredericstad. c The other places of origin were two immigrants from Bohemia, two from Sweden, two from France, one from Northern Italy, one from Hungary, and one from Alsace.

the smallness of the sample, it appears that Dutch ports serviced the upper reaches of the Rhine valley, particularly Wurttemberg-­Baden and Switzerland, whereas German ports serviced primarily the north German plain and Hessen region. The emigration through German ports, therefore, appears to represent a new demand from a different region of Germany for emigration to America. Six ship manifests reported the origins of their passengers at the village level but only one, the ship Margaret which sailed from Amsterdam and arrived in Philadelphia in 1804 with 269 immigrants, appears to have had a high concentration of immigrants from a given neighborhood. On this ship 72 percent of immigrants whose village could be located (39 percent of the total immigrants listed) were from villages within 25 miles of Stuttgart. The top three villages were Mohringen with 36, Ehningen with 21, and Markgroningen with 15 immigrants. Whether or not this pattern was caused by a jointly planned community migration is hard to determine. Although the origins of most immigrants are not known with certainty, the ports from which the transatlantic voyages commenced are known. For the colonial period (1727–75), Rotterdam dominated other ports—representing over 80 percent of the ship arrivals and over 90 percent of the adult male immigrants (see Table 2.2). London was second in terms of ship arrivals with a little over

The flow of immigrants, 1727–1835   23 Table 2.2 Distribution of the ports of origin for German immigrants to Philadelphia, 1727–75 Ports of departure

Number of ships

Number of signersa

Ratio of signers to ships

Rotterdam London Amsterdam Lisbon Hamburg

255 31 12 10 9

80.4% 9.8 3.8 3.2 2.8

22,259 625 856 89 566

91.2% 2.6 3.5 0.4 2.3

87.20 20.16 71.33 8.90 62.89

Totals

317

100.0%

24,395

100.0%

76.96

Source: Strassburger (1934: vol. 1). Seven vessels with uncertain origins, two from Boston, and one from South Carolina were excluded. Note a Signers were adult males over the age of 15 who were required by law to sign the oath of abjuration to the Pope and the House of Stuart and the oath of allegiance to the British Crown.

9 percent. However, the average number of adult males per ship was over three times lower on London vessels than for immigrant ships departing Dutch or German ports. Germans securing passage to America by first migrating to London would appear to be of rather incidental importance to London shippers compared with Rotterdam shippers (e.g. Muhlenberg 1942: vol. 1, 22–4). Amsterdam and Hamburg are the only other significant ports providing just over 3 and 2 percent, respectively, of German immigrant shipping services to Pennsylvania. The dominance of Dutch ports in shipping German immigrants continued in the post-­colonial era though to a lesser extent, dropping from over 90 percent to 64 percent (see Table 2.3). This period also saw a shift from Rotterdam to Amsterdam as the chief Dutch port. This may have been a consequence of the French, Dutch, Prussian, and Austrian conflict over the navigation of the Scheldt in 1784–5 as well as the factional turmoil in Holland which resulted in conflicts with the French and Prussians after 1786 (Schlosser 1844: vol. 5, 351–96). A second feature of the post-­colonial period was the rise in the importance of German ports, particularly Hamburg and Bremen, which together accounted for roughly 49 percent of the ship arrivals and 32 percent of the immigrants between 1785 and 1808. As suggested earlier, this appears to be a new source of German emigration to America, perhaps resulting from the anticipation of the French conflict spilling across the Rhine into northern Germany as it finally did in May of 1803, when French troops took Hanover and the rest of northern Germany west of the Elba River (Schlosser 1844: vol. 7, 365–71). Less specialization in immigrant shipping in these German ports compared with that observed in Dutch ports is also apparent. The ratio of immigrants to ships is approximately twice as high for immigrant vessels leaving Dutch ports. Immigrant cargo may have been only of incidental interest to shippers based in German ports compared with shippers in Dutch ports. Overall, with respect to shipping German immigrants, throughout the entire century shippers based in

24   German immigration to America, 1709–1835 Table 2.3 Distribution of the ports of origin for German immigrants to Philadelphia, 1785–1804 Ports of departure Amsterdam Hamburg Rotterdam Bremen Tonningen London Lubeck Fredericstad Copenhagen Totals

Number of ships

Number of immigrants

Ratio of immigrants to ships

57 57 7 10 2 1 1 1 1

41.6% 41.6 5.1 7.3 1.5 0.7 0.7 0.7 0.7

3,657 1,981 792 259 144 93 9 4 3

52.7% 28.5 11.4 3.7 2.1 1.3 0.1 0.1 0.1

64.16 34.75 113.14 25.90 72.00 93.00 9.00 4.00 3.00

137

100.0%

6,942

100.0%

50.67

Source: Strassburger (1934: vol. 3). Note These totals do not include ten ships with 892 immigrants whose port of origin could not be determined.

Dutch ports who carried passengers had a higher degree of specialization or concentration in human cargo relative to non-­human cargo compared with shippers based in English and German ports. Explaining the direction and variance of immigration is easier than explaining the exact magnitude of immigration. The economic theory of migration compares the discounted stream of expected future net benefits from staying home with the discounted stream of expected future net benefits of living in the relevant alternative country minus the expected cost of migration. For permanent location choices, the larger magnitude determines the choice of residence.17 Constructing an applicable economic model from this theory is an arduous task, owing to a lack of evidence on the desired variables. Historians have typically attributed the causes of German emigration to colonial America primarily to the differences in religious persecution, inheritance issues, border restrictions, and periodic wartime destructions.18 In terms of the economic theory of migration, this amounts to suggesting that the expected continuation of persecution against the reformed religions, persistence in impartible inheritance traditions, and periodic military hostilities in the Rhine region lowered the expected future net benefits, particularly for Protestants, below that found in the religiously tolerant, land-­plentiful, open-­bordered, and peaceful colony of Pennsylvania, hence a positive flow of emigrants.19 By contrast, economic models typically use quantifiable measures of material well-­being, such as per capita wealth or income, as a substitute for information on the total stream of net benefits (Greenwood 1975: 399–404). Such information for the Rhine region in this period is sparse. Most comparisons of Germany

The flow of immigrants, 1727–1835   25 with the American colonies in terms of material well-­being, therefore, will be done in a roundabout manner by comparing the colonies to England and then assuming that England was superior to other European countries at this time in terms of per capita material well-­being (Deane 1965: 10, 269; Jones 1980: 66; Landes 1969: 124–5). Comparing wages between regions is one way of identifying potential differences in expected future net benefits. The colonial period was marked by frequent comments on the dearness and scarcity of labor in the colonies relative to what was the norm in England and Germany. For example, in 1725 Christopher Sauer in a letter sent to friends and relatives in Germany shortly after his arrival in Philadelphia remarked that in Pennsylvania, “All labor is paid well” (Durnbaugh 1960: 229–30).20 He went on to list the remunerations received by various occupations. In a letter sent to Germany in 1728 another recent German immigrant commented, Very poor people, such as have sound and strong limbs, and are willing to work, may get along here, and in time may even become wealthy. Well to do people and such as cannot perform manual labor, but have all their work done by paying for it, can easily become poor in a few years – yea, even fall into great misery. The reason for this is that the day’s wages are so excessively high (Sachse 1909: 17–18) This letter, like so many others that were sent back to Germany, went on to inform the reader of the exact level of wages and the cost of necessities in Pennsylvania so that the reader could judge the full difference in the standard of living between the two regions. During the early republican period, 1790 to 1820, the margin of U.S. wages in the Philadelphia region over British wages has been calculated to range from 50 to 145 percent for skilled labor and 50 to 100 percent for unskilled labor (Adams 1970: 502). By extension, U.S. wages likely exceeded similar wages in Germany by at least the same margin. In 1819, Ludwig Gall after migrating from Germany to Pennsylvania commented in his journal, “The principal topic of conversation was the present depression, which must strike the newly arrived Europeans as odd, because the physical well-­being of people in this country is incomparably better than in the old country” (Trautmann 1981: 58). It would appear that throughout the eighteenth century the returns to labor in Pennsyl­ vania were generally higher than in Europe and this was communicated in letters sent back to Germany from prior immigrants. The economic rationale for why the colonies were so prosperous, particularly for the laboring classes, relative to Europe was aptly stated in 1776 by Adam Smith, In our North American and West Indian colonies, not only the wages of labour, but the interest of money, and consequently the profits of stock, are

26   German immigration to America, 1709–1835 higher than in England. . . . High wages of labour and high profits of stock, however, are things, perhaps, which scarce ever go together, except in the peculiar circumstances of new colonies. A new colony must always for some time be more under-­stocked in proportion to the extent of its territory, and more under-­peopled in proportion to the extent of its stock, than the greater part of other countries. They have more land than stock to cultivate. . . . Stock employed in the purchase and improvement of such lands must yield a very large profit, and consequently afford to pay a very large interest. Its rapid accumulation in so profitable an employment enables the planter to increase the numbers of his hands faster than he can find them in a new settlement. Those whom he can find, therefore, are very liberally rewarded. (Smith 1937: 92) In other words, the initial difference in resource endowments between America and Europe, namely the relative scarcity of labor and capital and the relative abundance of land in America compared with Europe, determines the relative payments to these inputs. The relatively scarce resource receives relatively more. Under these conditions mobile factors of production will migrate to where they are relatively scarce since this is where they also command a higher return, barring prohibitively high migration costs.21 Wage comparisons do not reveal the entire story. Because eighteenth-­century transatlantic migration was typically a permanent location choice, prospective immigrants would want to compare the difference in the expected sums of material benefits over their lifetime between locations. In an analysis of colonial American probate inventories, Alice Hanson Jones found that per capita lifetime accumulated wealth of American colonists was quite substantial and concluded, “If we had the true figures for both sides of the Atlantic, we might find the colonies in 1774 very nearly on a par in per capita income with England, but somewhat lower in wealth.” Jones also concluded that nonhuman physical wealth in 1650 was only about a third below what was found for 1774, indicating that the relatively high wealth position of the colonies was a progressing phenomenon. In addition, the middle colonies exceeded the southern and New England colonies in terms of per capita nonhuman wealth when slaves and servants were counted in the population. This would be the relevant measure of interest for prospective German immigrants because most arrivals were not capable of buying slave plantations. Per capita wealth in the middle colonies was also more evenly distributed than either in England, New England, or the southern colonies.22 A more even distribution would be attractive to prospective German immigrants who would likely be entering colonial society at the bottom as ser­ vants or laborers. Therefore, in terms of the ability to accumulate wealth, the middle colonies offered a superior expected standard of living to prospective German emigrants relative to other colonial destinations. These comparisons are broadly consistent with an economic explanation for the positive flow of German immigrants. The

The flow of immigrants, 1727–1835   27 presence of these economic incentives may explain why the outcome of other calamities, such as religious persecution and wartime displacement, took the form of migration to Pennsylvania rather than to some other American destination. While detailed information on the level and distribution of colonial wealth would not be known directly by contemporary immigrants, opportunities in the colonies were conveyed in the barrage of competing colonial recruitment pamphlets that flooded Germany in the eighteenth century. Of more importance was the considerable traffic of correspondence between German colonists already in Pennsylvania and relatives and friends in Germany.23 This correspondence established Pennsylvania’s superiority in the minds of prospective German emigrants. Germans took this correspondence seriously and it caused continued German emigration.24 The impact of letters from Christopher Sauer who immigrated to Pennsyl­ vania in 1724 is illustrative of this influence. The effect of his first letters to his home village are told in the diary of Charles Hector of Berleburg, Wittgenstein in 1725, there came letters from Pennsylvania from a man known to us, who long lived at Schwarzenau and was married to a widow of the Pastor Grohr, our friend who had lived with us at Gersbach. These letters contained a description of Pennsylvania that it was like an earthly paradise. The man reported that one could buy a considerable piece of land for 200 or 300 Reichstaler, from which one could comfortably have his living, and which was already cultivated. The country had abundance and was fruitful in all necessary things. There was complete freedom. One could live there as a good Christian in solitude, as one pleased, and if one wanted to work a little, especially craftsmen, among them also clockmakers, then one could earn his livelihood with abundance. Our old widow, the mother of the tailor Sauer, who had written the letter, and his mother-­in-law, Mrs. Gruber, who was at the time with her daughter in Berleburg, resolved to leave for Pennsylvania, with her daughter and yet another daughter who had a family. Other families decided the same, at least one hundred persons from this area, all people whom we knew. (Durnbaugh 1958: 324)25 Not all of Sauer’s reports were enthusiastic. In 1739 a German publication in Frankfurt printed a report from America wherein a group of German-­ Pennsylvanians counseled against people making too hasty or ill-­prepared journeys to the New World. Sauer’s signature was second on the list of fourteen signers (Durnbaugh 1958: 327). Sauer assessed his own contribution to stimulating German immigration in a letter to Governor William Denny on 15 March 1755, when I came to this Province and found everything to the contrary from where I came from, I wrote largely to my friends and acquaintances of the civil and religious liberty, privileges, etc., and of the goodness I have heard

28   German immigration to America, 1709–1835 and seen, and my letters were printed and reprinted, and provoked many thousand people to come to this province, and many thanked the Lord for it, and desired their friends also again to come here. (Brumbaugh 1907: 376–82) The positive flow of German emigrants to Pennsylvania is consistent with the economic theory of migration which suggests that knowledge of substantial wealth differentials between Germany and Pennsylvania would encourage this flow. The problem facing prospective emigrants was in assessing the magnitude of this differential and weighing it against the rather uncertain costs and risks incurred while migrating. It is primarily changes in these costs and risks that determine the variation in immigration over time. The variance in German immigration to Pennsylvania between 1727 and 1835 was quite substantial (see Figure 2.1). The economic theory of migration indicates that temporary fluctuations in economic opportunities should not have a significant impact on changing the migration decision because lifetime opportunities are the key comparisons when making permanent location choices. 9,500 9,000 8,500 8,000 7,500

Number of immigrants

7,000

War of the Austrian Succession

6,500 6,000 5,500 5,000 4,500 4,000 3,500 3,000 2,500

War of the Polish Succession

Seven Years War French Wars and Continental Blockade American Revolution

Year Without a Summer Jefferson Embargo and War of 1812

Panic of 1819

Peace of Luneville

French Wars

2,000 1,500 1,000 500 0 1730 1740 1750 1760 1770 1780 1790 1800 1810 1820 1830 Year

Figure 2.1 The flow of German immigrants through the port of Philadelphia, 1727–1835 (sources: Wokeck (1999: 45–6) for 1727–75; chapter 18, appendix table 18.1 for 1785–1835).

The flow of immigrants, 1727–1835   29 Temporary fluctuations in wage rates, which were considerable, should be of minor importance. Immigration was, in part, the cause of these fluctuations rather than being caused by them.26 Therefore, the explanatory variables of interest in accounting for the variance in immigration are events that significantly changed the long run prospects of individuals and events that significantly shifted the cost of migration. Figure 2.1 indicates that military hostilities on the route of migration are the dominant explanatory variable of the variance in immigration, war being associated with less migration. This pattern is found in the regressions in Table 2.4, where the flow of adult male German immigrants to Philadelphia from 1727 to 1774 is related to war, political dummy variables, and a trend term.27 In these results, peace years had an average of 913 more adult male immigrants arriving per year than war years. Warfare along the route of migration represents a tremendous and sudden rise in the cost of migration. Some of this was due to increased delays along the journey and some to higher freight rates required to cover wartime shipping insurance costs (Jacobs 1897: 39; Bruchey 1966: 182; Chapter 3). But the main rise in migrating costs resulting from war was the increased uncertainty about the safeness of the journey, i.e. the increased probability that one’s property and life would be violently dispatched. Examples of traumatic wartime migrations abound. The ship Argyle, sailing in 1745 at the height of the War of the Austrian Succession, was captured by two Spanish men-­o-war. Pastor Muhlenberg, sailing in 1742, related the fears of wartime transportation in his journal, namely the difficulty and delays of convoying down the English Channel, passenger-­drills to repel boarders, the anxiety occasioned when sighting other sails, and the risks involved when the route necessarily passed too close to the Spanish coast. The Moravian Church lost vessels they had purchased to bring church members across the Atlantic during the War of the Austrian Succession and during the Seven Years War. Even the first edition of Christopher Sauer’s New World Bible shipped to Europe in 1745 was captured by French and Spanish privateers off the cape of St. Malo.28 These fears must have held true on land as well at sea, because many of the wars centered on the Rhine river—the main corridor of emigration. Separating the location of hostilities on the migration route—regression (3) in Table 2.4, indicates that war in North America had a greater impact on stemming the flow of immigration than warfare closer to the emigrants’ homes. Once the commitment to migrate was made it was difficult to disengage, especially once the transatlantic voyage had begun. Wars on the more distant stretches of the migration path should, therefore, generate more uncertainty. They should have a greater impact on stemming immigration both because the information on the true situation might be less reliable and because the situation might change rapidly after the migration commitment had already been made and disengagement was impossible (Greenwood 1975: 404–6; Dunlevy 1980: 77–90). Controlling for periods of war, the year variable in Table 2.4 estimates the trend in immigration between wars. This trend is negative and statistically significant though its magnitude is small. There is a definite peak in immigration just

Regression coefficients (1) 0.654*** (0.110) 53,972.700** (28,186.600) 645.000* (386.400) –31.000** (16.200) 912.900*** (177.600) . . . . . . . . . . . . . . . . . . 0.647 26.300** 2.130 48

Explanatory independent variables

First-order autocorrelation Constant Dynastic change in the Palatinatea Year of arrival Years of peaceb

Individual warsc Polish Succession Austrian Succession Seven Years War

Warfare locationd Upper Rhineland Holland North America

R2 F-statistic Durbin-Watson Observations

0.561 10.470** 2.040 48

. . . . . . . . .

–548.200* (350.100) –1,002.700*** (330.500) –402.600† (328.800)

0.629*** (0.110) 55,152.700** (31,588.900) 681.000* (445.500) –31.300** (18.100) . . .

Regression coefficients (2)

0.538 9.540** 2.150 48

–329.000* (235.200) –73.800 (305.600) –791.400*** (326.400)

. . . . . . . . .

0.544*** (0.120) 36,924.400* (27,806.700) 695.100* (427.900) –20.900* (16.000) . . .

Regression coefficients (3)

Table 2.4 Regression results: the determination of the variance of German immigration to Pennsylvania, 1727–74

Notes The dependent variable is German male immigrants above the age of 15, i.e. signers of the loyalty oaths. Standard errors are in parentheses. The method of estimation was Cochrane-Orcutt iterative technique for correcting autocorrelation in ordinary least squares as performed by the TSP statistical package. *** Indicates that the coefficient is statistically different from zero above the 0.01 significance level. ** Indicates that the coefficient is statistically different from zero above the 0.05 significance level. * Indicates that the coefficient is statistically different from zero above the 0.10 significance level. † Indicates that the coefficient is statistically different from zero above the 0.20 significance level. a This was coded as one if the year was greater than 1742 and zero otherwise. b This was coded as one for the years 1727–32, 1738–41, 1749–54, and 1763–74 and zero otherwise. c This was coded as a series of dummy variables where the War of the Polish Succession was coded as one for the years 1733–7 and zero otherwise, the War of the Austrian Succession was coded as one for the years 1742–8 and zero otherwise, and the Seven Years War was coded as one for the years 1755–62 and zero otherwise. The zero-category was the years of peace. d This was coded as a series of dummy variables where the Upper Rhineland was coded as one for the years 1733–5, 1742–7, and 1757–62 and zero otherwise, Holland was coded as one for the years 1742 and 1744–7 and zero otherwise, and North America was coded as one for the years 1755–60 and zero otherwise. The zero-category was the years of peace.

Sources: The source for German immigration is Wokeck (1981: 260–1). The source for the other explanatory variables are Dorn (1940); Hassall (1907); Roberts (1947); Schlosser (1844: vols 3, 4).

32   German immigration to America, 1709–1835 following the conclusion of a war, and then immigration tapers off at an average rate of 31 adult males a year until the next war (see Figure 2.1). These post-­war peaks represent either accumulated postponed migration decisions during the war or some displacement caused directly by that war (Baines 1991: 22). The German immigration between the Peace of Aix-­la-Chapelle and the outbreak of the Seven Years War, 1749–54, accounts for nearly half of the post-­ 1727 colonial German immigration (see Figure 2.1). It is difficult to attribute this surge solely to the ending of the War of the Austrian Succession.29 Periodic warfare in this region had occurred for centuries, making it difficult to single out one war as the cause of a particularly striking post-­war surge in migration. Second, there was a major change of government in the Palatinate in 1743. Karl Philip, the last of the house of Neuberg, had ruled for 27 years with a relatively steady hand. In 1743 the succession passed to Karl Theodore of the Sulzbach line who was under Jesuit influence and in league with the French. This resulted in rising intolerance toward the Protestant populations (Schlosser 1844: vol. 4, 41–59; Smith 1959: 51). This political change may have altered Protestant assessments of their lifetime opportunities in the Palatinate, causing their net lifetime benefits from emigrating to increase. The results in Table 2.4 indicate that, controlling for war, the years ruled by Charles Theodore are associated with an average increase of 645 adult male immigrants per year above the years ruled by Charles Phillip. The declining influence of the Jesuit order in southern Germany after 1769 may explain the decline in immigration occurring after the Seven Years War (Schlosser 1844: vol. 4, 446–63). Comparing the impact of different wars can only be done meaningfully by controlling for these different political regimes. Consequently, the only relevant comparison for this period is between the Polish and Austrian Wars of Succession. Regression (2) in Table 2.4 indicates that the Polish War had less impact on restricting the flow of immigration. This is understandable given that the Polish War had relatively less fighting along the Rhine corridor and relatively fewer French troops maneuvering or billeting on the Rhine and so obstructing the migration route (Schlosser 1844: vol. 3, 295–313, 328–445). Finally, for the colonial period, 25 percent of the variance in immigration can be accounted for by autocorrelation (see Table 2.4). In other words, the single best predictor of immigration in a given year is immigration from the previous year. This result is consistent with the importance of chain-­migration dominating the migration process—immediately prior immigrants inducing subsequent immigration of fellow countrymen.30 Explaining the variance in German immigration to the Delaware valley between 1776 and 1835 is more difficult owing to the rise of new regions contributing to this migration, namely north-­central Germany, and to the frequency of overlapping shocks affecting the migration decision which cannot be statistically disentangled. The variance in the flow, however, appears consistent with the patterns found for the colonial period. War on the migration route is the dominant

The flow of immigrants, 1727–1835   33 determinant of the yearly variation in immigration—with warfare in America having a stronger impact on dampening immigration than warfare in Europe. During the American revolution and the War of 1812, German immigration fell to near zero (see Figure 2.1). Local peaks in immigration also occurred immediately after the conclusion of a war with immigration flows abating thereafter, similar to the pattern seen in the colonial period. Such peaks are apparent just after the end of the American revolution (1784), after the end of the Second War of the Coalition (1802), and after the end of the War of 1812 and the Napoleonic Wars (1815). This pattern represents either accumulated postponed migration decisions during a given war or some displacement caused directly by those wars. The years 1790–7 appear to be an anomaly to this peace-­versus-war migration pattern. German immigration, while relatively low compared with other years, was on the rise from 1790 through 1796, which coincides with the French revolution (1789) and the First War of the Coalition (1792–7). French armies occupied the west bank of the Rhine producing outbreaks of hostilities all along the Rhine valley. A French refugee army opposed to the French revolution was on the left bank of the Rhine at Worms and Cologne in 1790. The brief truce between the First and Second Wars of the Coalition (1798) was puzzlingly the year with the lowest German immigration between 1791 and 1807 (Schlosser 1844: vol. 6, 296–633; vol. 7, 61–182, 296–325). This apparent anomalous migration pattern is resolved by noting that from 1792 to 1801 the trend of German immigration through Dutch ports was declining. The rise in German immigration from 1792 through 1796 was all due to the rise in immigration through German ports. As discussed above, this appears to be a new source of German emigration to America, perhaps prompted by some anticipation of the French wars spilling across the Rhine into northern Germany. With the Peace of Luneville (1801), German immigration through both Dutch and German ports, from both the upper Rhine valley and the north German plain, accelerated once again producing the 1804 localized peak in immigration seen in Figure 2.1. The variance in German immigration into the Delaware valley between 1815 and 1835 is also difficult to explain. Why German immigration was low between 1819 and 1830 and why it peaked sharply in 1831–2 only to collapse again to near zero is somewhat puzzling. Some of this pattern is due to German immigration shifting from Philadelphia to other U.S. ports after 1819, but that does not explain its low levels across all U.S. ports in the 1820s.31 The sharp peak in 1816–19 and subsequent collapse after 1819, however, corresponds to two freak events. In April of 1815 the volcano Tambora in Indonesian erupted. The eruption ranks as the most explosive over that past 10,000 years—ejecting 100 times as much matter into the atmosphere as the Mount St. Helens eruption in 1980. The ash cloud circled the globe and caused a substantial climate shock that lasted two to four years. This shock was known in northwestern Europe as “the year without a summer.” It devastated the upper Rhine valley where crops failed and frozen waterways prevented food from being

34   German immigration to America, 1709–1835 imported. Substantial outmigration of desperate and starving souls from the upper Rhine occurred, some of whom made their way to Amsterdam and took ships to America. In fact, 1817 witnessed the single largest German immigration into the Delaware valley since 1754 (see Figure 2.1). This shock was over by 1819. At the same time the financial panic of 1819 severely disrupted U.S. industrial production and transatlantic shipping markets. The coincidence of these two events dramatically reversed, at least temporarily, the assessment of the net benefits of migrating from Germany to the U.S. after 1819 compared with what it had been during 1815–19.32 The effect of war on stemming the flow of German immigration may, in part, be attributable to the rigid seasonality of German immigration (see Table 2.5). Choosing to arrive in Pennsylvania almost exclusively in the autumn implies initiating emigration from residences in the upper Rhine valley sometime in the late spring or early summer, namely during the major war-­campaign season. The apparent unwillingness to substitute other seasons for initiating migration implies an inability to take advantage of lulls in the fighting, or the usual fall and winter truce during periods of protracted warfare, to migrate out. The desire not to migrate during the harsh winter months and not to arrive in the colonies during the summer when new immigrants were disproportionately susceptible to the summer cycle of colonial fevers appears to be behind this rigid seasonality pattern.33 William Penn, in his influential 1690 pamphlet, A Further Account of the Province of Pennsylvania and its Improvements. For the Satisfaction of Those that are Adventurers and Inclined to be so., circulated in Germany to encourage immigration to Pennsylvania, stated: Tho Ships go hence at all times of year, it must be acknowledged, that to go as to arrive at Spring or Fall, is best. For Summer may be of the hottest, for Table 2.5 Seasonal distribution of German immigrants arriving in Philadelphia, 1727–1804 Month of arrival

Sailing from Rotterdama 1727–75

Sailing from Amsterdam and Rotterdamb 1785–1804

August September October November All other months

2,837 11,336 5,876 1,732 478

12.8% 50.9 26.4 7.8 2.1

943 1,173 1,302 483 548

21.2% 26.4 29.3 10.8 12.3

681 487 340 587 587

30.4% 21.7 15.2 26.2 26.2

Totals

22,259

100.0%

4,449

100.0%

2,240

100.0%

Sailing from Hamburg and Bremenb 1785–1804

Source: Strassburger (1934: vols 1, 3); the notes to tables 2.2 and 2.3 and figure 2.1. Notes a Only signers of the oath of abjuration and oath of allegiance, i.e., adult male immigrants above age 15, are counted here. b All immigrants are counted here.

The flow of immigrants, 1727–1835   35 fresh Comers, and in the Winter, the wind that prevails, is the NorthWest, and that blows off the Coast, so that sometimes it is difficult to enter the Capes. I must confess I prefer the Fall to come thither, as believing it more healthy to be followed with winter than Summer (Diffenderffer 1899: 29) This pattern of fall arrival also appears among English indentured servants shipped to the Chesapeake, where it has also been ascribed to desires of avoiding the disease effect of summer arrival (Galenson 1981: 88–91). This pattern for British and Irish indentured servants, however, abates during the latter part of the eighteenth century. The arrival of Irish servants in Philadelphia for the years 1745–6 and 1771–3 exhibit strong bi-­seasonality with large numbers arriving in both spring and fall. This abatement is generally attributed to the declining importance of seasoning mortality over the century (immigrant deaths within the first year after arrival).34 However, Chapter 3 shows that substantial seasoning mortality was still present among Germans arriving in Pennsylvania in the middle of the eighteenth century. In addition, the declining importance of autumn arrival only appears among Germans leaving through non-­Dutch ports. Less than 50 percent of German immigrant ships which sailed from London in the colonial period arrived in the fall (Strassburger 1934: vol. 1). This indicates that the long overland migration for southern Germany to coastal ports, and continual worries about high seasoning mortality among late spring and summer arrivals, may have continued to contribute to a rigid seasonal pattern for German immigrant arrivals. The rigid seasonality in German immigrant arrivals is remarkable considering the many factors that could impart variability to the migration journey of Germans to colonial America, compared with that for Irish or British emigrants. Unlike most Irish and British emigrants, Germans had an arduous trek just to get from their homes, typically in the upper Rhine region, to a coastal port. Little firsthand evidence exists describing this first leg in the journey. One often quoted account is by Gottlieb Mittelberger. Writing in 1756 he said, I took the usual route down the Neckar and the Rhine to Rotterdam in Holland. The trip from home to Rotterdam including the sojourn there, took fully seven weeks because of the many delays encountered both in going down the Rhine and in Holland . . . The reason for this is that the Rhine boats must pass by thirty-­six different customs houses between Heilbronn and Holland. At each of these all the ships must be examined, and these examinations take place at the convenience of the customs officials . . . it also means that the trip down the Rhine alone takes from four to six weeks. When the ship with their passengers arrive in Holland they are there held up once again for from five to six weeks. Because everything is very expensive in Holland the poor people must spend nearly all they own during this period. (Mittelberger 1960: 7–11)

36   German immigration to America, 1709–1835 The transatlantic voyage from Holland to America in the colonial period consisted of two stages. The first stage was due to the English navigation acts. Ships leaving Dutch ports bound for the English colonies had to clear customs at an English port. Little evidence exists concerning this first stage of the voyage except for the distribution of English ports used by German passenger ships to clear English customs (see Table 2.6). Cowes, on the Isle of Wight, was by far the most frequently used port to clear English customs. For the miles traveled, this first stage of the voyage was the most time-­ varying as well as time-­consuming due to the prevalence of contrary winds when sailing west through the English Channel. Mittelberger states that the voyage from Holland to Cowes took from less than eight days to four weeks depending on the winds, and once there the ship had to ride at anchor for eight to 14 days while going through customs (Mittelberger 1960: 12). A diary from 1728 describes the journey as taking seven days to sail from Rotterdam out the Rhine delta to Helvoelsluys where the captain paid Dutch customs duties. The length Table 2.6 Port of last clearance for German immigrant vessels arriving in Philadelphia, 1727–75 Port of last clearance

Number of ships

Percentage

Cowes Portsmouth London Deal None listed Plymouth Dover Lisbon Leith, Scotland Gosport Falmouth “England” Orkney Islands, Scotland Boston South Carolina St. Christophers Aberdeen, Scotland Poole Shields Tingmouth [Teignmouth] Berwick-on-Tweed

151 35 32 22 19 16 13 10 5 4 3 3 2 2 1 1 1 1 1 1 1

46.60% 10.80 9.88 6.79 5.86 4.94 4.01 3.08 1.54 1.23 0.93 0.93 0.62 0.62 0.31 0.31 0.31 0.31 0.31 0.31 0.31

Totals

324

100.00%

Source: Derived from Strassburger (1934: vol. 1). Notes Per the Navigation Acts, ships traveling from the northern European continent to British North America had to clear customs in Britain before arrival in America. Clearance ports like Leith, Scotland were used primarily during wartime, such as during the War of the Austrian Succession, particularly in 1747–8.

The flow of immigrants, 1727–1835   37 of this stretch was due to contrary winds and faulty piloting which ran the ship aground several times. It took another eight days to reach Plymouth in England, where the vessel went through customs. The ship stayed at Plymouth for eight days and then sailed three more days before it passed Land’s End and cleared the English Channel (Sachse 1909: 8–9). Information from six separate diaries of German immigrant voyages during the colonial period shows that on average 29 percent of their total voyage time was used just to clear the English Channel.35 Most of this excessive voyage time was due to the required layover in an English port due to wartime convoying and the necessary customs inspection. None of these six voyages were excessively long or suffered above average loss of life among the passengers. The second stage of the transatlantic voyage was from the English clearance port to North America. The overall sailing time from England to America changed little from the late seventeenth to the early nineteenth century. There were no perceptible changes in vessel speed and thus voyage length or variability throughout the late colonial period (North 1968: 935–70; Walton 1967a: 67–78). Significant gains in shipping productivity were made, but they came through declining terminal costs and crew sizes, not increased sailing speed. For German immigrant vessels, the mean voyage time was about two months. William Penn in his 1690 pamphlet recruiting German colonists for Pennsylvania stated, The Passage is not to be set by any man; for Ships will be quicker and slower, some have been four months, and some but one month and as often. Generally between six and nine weeks. One year, of four and twenty Sayl, I think, there was not three above nine, and there was one or two under six weeks in the passage. (Diffenderffer 1899: 29–30) In 1756 Mittelberger wrote, For from there [Cowes] the ships often take eight, nine, ten, or twelve weeks sailing to Philadelphia, if the wind is unfavorable. But even given the most favorable winds, the voyage takes seven weeks. (Mittelberger 1960: 14) Between 1728 and 1754, 18 German immigrant vessels could be found with specific dates for clearance and arrival.36 For these ships the average time at sea was 65 days with a standard deviation of 16 days. The shortest trip was for the Moravian Church ship Irene, which in 1754 sailed from Gravesend to New York in 31 days (Diffenderffer 1899: 260; Jordan 1909: 47). The longest voyage reported for a German immigrant vessel actually arriving in North America was for the ship Love and Unity which wrecked on the coast 24 weeks after departing Rotterdam.37 Decreases in the average time at sea did not occur until sometime after the War of 1812 with the rising use of packet vessels. The average

38   German immigration to America, 1709–1835 westbound passage, Liverpool to New York, for packet vessels between 1818 and 1827 was 38 days (North 1968: 967).38 In the post-­colonial period the necessity of making an English customs stop was eliminated—saving up to a week of time from that experienced on the typical German immigrant voyage during the colonial period. The average total migration time for German immigrants, using Mittel­ berger’s estimate of seven weeks from home to boarding an ocean-­going ship, 65 days sailing time from England to America, and 20 percent of the 65 days as the time needed to sail from Rotterdam through English customs, was a little over one-­third of a year. Total migration time could range from two months to two-­thirds of a year. While this represented a considerable cost in terms of lost work opportunities, perhaps the most striking feature was not the average time which might be planned for, but the large variance and hence non-­predictability of the actual migration time which made the realized cost difficult to know. Given this, it is rather remarkable that such a rigid seasonality of arrival was realized for this German immigration stream. The magnitude, substantial yearly variance, and rigid seasonality of German immigration during the colonial period raise questions about whether German emigrants could expect to find adequate shipping services in Dutch ports when they arrived there looking for passage to America. The lack of adequate shipping services could influence migration patterns. The available cargo capacity on ships carrying German immigrants to America is assessed by matching the 324 German immigrant vessels arriving in Philadelphia between 1727 and 1775 with information in the ship registry for the port of Philadelphia—with 66 out of the 324 ships being so matched. The port registry listed each ship’s registered tonnage. Because registered tonnage is not a consistent measure of shipping capacity, registered tonnage is converted to the 1695 formula for measured tonnage which is then used as the capacity measure of analysis.39 Immigrants per ton on these 66 vessels was 1.04 with a standard deviation of 0.65. By comparison the average slaves per ton in the British transatlantic slave trade to Virginia and Jamaica was over 38 and 86 percent higher, respectively (Klein 1978: 124, 126, 142; Chapter 13). Because the transatlantic slave trade involved only slave cargos on the westbound voyage, slaves per ton will be taken as an upper bound for a ship’s human freight carrying capacity. By comparison, German immigrant shippers were not packing their vessels to capacity with immigrant cargo. The large standard deviation in immigrants per ton relative to the mean also indicates that most German immigrant voyages were not filled to capacity with German immigrants. Only on those few ships in the upper end of the distribution, i.e. one standard deviation above the average immigrant-­per-ton ratio, were German immigrants packed into vessels on a level similar to that experienced by slaves in the British transatlantic slave trade. Therefore, except when experiencing unusually heavy and unanticipated demands for shipping services such as in peak migration years, prospective German immigrants could expect to find shipping services readily available in Dutch ports during the colonial period.

The flow of immigrants, 1727–1835   39 This observation does not mean that German passengers enjoyed ample space on board ships bound for Philadelphia. Transatlantic shippers in Dutch ports did not specialize exclusively in human cargo but carried a mixture of human and nonhuman freight. Immigrant cargo was only a fraction of the overall freight business. The number of ships and shipping tonnage participating in the German immigrant trade was always substantially less than the flow of new ship construction in the Delaware valley, and much smaller than the stock of accumulated shipping tonnage registered at the port of Philadelphia for any given year (Crowthers 1973: 93). Immigrants were typically carried only by ships with established nonhuman freight business between the relevant ports. For example, Bishop Spangenberg, head of the Moravian Church, when questioned by Governor Hamilton of Pennsylvania on why he did not bring his German immigrants to Pennsylvania directly through the port of Philadelphia responded, We wish we could use the port of Philadelphia, but since our captain is a native of New York, and has a large acquaintance with the merchants of the city, he can more readily obtain freight there than in Philadelphia, passengers alone not being sufficient. Another serious objection is, the merchants of Philadelphia own their own vessels. (Jordan 1909: 40–1) As such, prospective German immigrants to America competed for freight space primarily with nonhuman cargo. If shippers chose the human-­to-nonhuman freight mix aboard their vessels, then the variance of crowding on immigrant ships was determined by the conditions currently prevailing in the shipping market. An increase in the demand for immigrant shipping services, namely an increase in total immigration, would be met by an increase in the quantity of immigrant shipping services supplied. This increase in the quantity supplied can be accommodated either by increasing the number of passengers per ton (squeezing out nonhuman cargo on vessels that typically carried immigrants) or by increasing the number of ships participating in the immigrant trade (enticing shippers who only carried nonhuman freight into now carrying some immigrants as well). As such, the variance in German immigrants per ton should be determined by the variance in yearly German immigration coupled with the degree of response in these two margins of supply. The number of ships participating in the immigrant trade should depend on the flow of new ships into the market and the conditions prevailing in the market for nonhuman freight. Thus, new ship construction and the amount of competing freight should affect the German immigrant-­per-ton ratio. Merchants also wanted to minimize the amount of time their vessels were idled in port loading and unloading freight. They tried to collect and store cargo so as to have a full load assembled and ready to go once their ships arrived.40 This was difficult and costly to achieve with human cargo compared with nonhuman freight because humans required additional considerations beyond storage

40   German immigration to America, 1709–1835 and because immigrant arrivals at port were highly variable. Therefore, larger vessels may have found it more difficult and time consuming to wait in port collecting a full load of immigrants than did smaller vessels. As such, smaller ships on average might have higher passenger-­per-ton ratios than larger ships among those participating in the immigrant trade. Table 2.7 reports the results from a regression model used to estimate the determinants of crowding on German immigrant from the sample of 66 voyages between 1727 and 1774 for which ship tonnage and total passengers are known. Over 70 percent of the variance in immigrants per ton is explained by two vari­ ables—total German immigration in that year and the size of the vessel in registered tons. The variance in immigration is largely accommodated by changes in the degree of crowding as measured by changes in the passenger-­per-ton ratio. A 1 percent increase in total immigration raises the passenger-­per-ton ratio by 0.65 percent. Other supply responses are less important. As expected, an increase in the number of ships carrying immigrants reduced the passenger-­per-ton ratio, but the effect was small and only marginally statistically significant. Although the number of ships varies positively with total immigration, this response was not enough to maintain a constant crowding factor, leaving passengers per ton positively related to changes in total immigration. Table 2.7 Regression results: determination of crowding (passengers-per-ton) on vessels carrying German immigrants to Philadelphia, 1727–74 Explanatory variables

Regression coefficients

First-order autocorrelation Year of arrival Adult male illiteracy percentage Registered tonnage of vessel Total German immigration Number of ships arriving that year with German immigrants Total new ship tonnage registered in prior year in Philadelphia Total new tons of Delaware valley ship construction in prior year R2 F-statistic Durbin–Watson statistic Observations

–0.34** (0.192) –45.08** (30.890) –0.27* (0.208) –0.87*** (0.187)   0.65*** (0.242) –0.30* (0.330)   0.90** (0.549) –0.75** (0.530)   0.79   8.62**   2.16 66

Sources: See the text in this chapter, Table 2.4, and Chapters 4–7 and 13 for data source discussions. The evidence on adult male illiteracy was derived from Strassburger (1934: vol. 1). Notes The dependent variable is the natural logarithm of the number of passengers per registered ton. All independent variables are in natural log form. The unit of observation is the per year average of these variables for the years in which observations of passengers per ton were made—25 years in total. Standard errors are in parentheses. The method of estimation was Cochrane–Orcutt iterative technique for correcting autocorrelation in ordinary least squares as performed by the TSP statistical package. *** Indicates that the coefficient is statistically different from zero above the 0.01 significance level. ** Indicates that the coefficient is statistically different from zero above the 0.1 significance level. * Indicates that the coefficient is statistically different from zero above the 0.20 significance level.

The flow of immigrants, 1727–1835   41 The other major explanatory variable in the regression is the size of the ship as measured in registered tons. A 1 percent increase in registered tonnage lowers the passenger-­per-ton ratio by 0.87 percent. In part, this is consistent with the finding that carrying capacity was a declining function of registered tonnage. As registered tonnage per ship increases, the shipping capacity per ton decreases. Therefore, ships with smaller registered tonnage may only appear to carry more passengers per space available because their tons possess more shipping capa­ city. The negative relationship, however, persists even when estimated in 1695 measured tonnage (French 1973: 434–43).41 As such, larger vessels may have run into some difficulty in collecting a full load of immigrants compared with collecting a full load of nonhuman freight. The flow of new registered tonnage into the Philadelphia market, when controlling for the flow of new ship construction in the Delaware valley, represents tonnage which was drawn away from other ports and into the Philadelphia shipping market. Because immigrants represented only a small portion of total shipped freight, this flow represents the short-­run response to changes in the demand for shipping nonhuman freight. Slack periods in the shipping of nonhuman freight in the Philadelphia market relative to other ports would imply increased competition for human cargo and thus a lower passenger-­per-ton ratio.42 A 1 percent fall in total new registered tonnage in the port of Philadelphia in the prior year implies a 0.9 percent fall in the passenger-­per-ton ratio (see Table 2.7). By contrast, the flow of new ship construction from the Delaware valley registered in Philadelphia, when controlling for total new ship registrations, represents the long-­run response to conditions in the nonhuman freight market. This long-­run response has a negative impact on the passenger-­per-ton ratio (see Table 2.7). It was accommodating the pressure for shipping in the nonhuman freight market, providing a more elastic response in the supply of shipping ser­ vices which in turn lowered the passenger-­per-ton ratio, other things equal, when the demand to ship immigrants spiked. These two effects, though statistically significant, account for only a small portion of the overall variance in passengers carried per ton. The long-­run trend in the passenger-­per-ton ratio is negative over the colonial period when controlling for the other factors in the regression (see Table 2.7). This trend is consistent with the increasing economization of port time over the century, and the general decline in the importance of German immigrant cargo in the shipping market across the third quarter of the eighteenth century. The negative first-­order autocorrelation, however, indicates that shipping services were quite responsive in the short-­run. A spike in the passengers-­per-ton ratio in a given year is followed by a sharp decline in the following year, which is consistent with the presence of a quick response by the quantity of immigrant shipping services supplied to unexpected excessive demand for shipping services by emigrants during the colonial era. For 1815 through 1819, 62 ships landing German immigrants at Philadelphia had information on tonnage (“Philadelphia Customs House records”). On these

42   German immigration to America, 1709–1835 ships the average passenger-­per-ton ratio was 0.2, with a standard deviation of 0.3. Even though this period encompassed the second largest peak in German immigration through Philadelphia between 1727 and 1835 (Figure 2.1), the passenger-­perton ratio was on average 80 percent lower than on German immigrant vessels in the colonial era. German immigrants experienced significantly less crowding on their vessels post-­Napoleonic wars than they did pre-­American revolution. The Federal Passenger Act of 2 March 1819 set the maximum passenger-­perton ratio at 0.4 for ships sailing from continental Europe to the U.S. (Hutchinson 1981: 20–2). Violators faced a large fine—$150 for every passenger in excess of the 0.4 ratio. Only 12 of the 62 ships in the 1815–19 sample exceeded that limit. For these 12, the average passenger-­per-ton was 0.8 with a standard deviation of 0.3—still 23 percent below the colonial average. In fact, only three of the 62 ships exceeded the colonial average of 1.04 passengers per ton, the highest being 1.36 passengers per ton on a ship arriving in 1817. Excessive crowding on German passenger vessels before 1819 was not a general problem or the true cause of the 1819 Federal passenger legislation.43 From 1820 to 1827, no ship arriving in Philadelphia with German immigrants exceeded the 0.4 passenger-­per-ton legal maximum—out of a sample of 67 ships. However, these were years of relatively low German immigration. Between 1828 and 1835, when German immigration picked up again, out of a sample of 69 ships landing German immigrants at Philadelphia 12 had over 0.4 passengers per ton, the highest being 0.65 on a ship arriving in 1831. After 1819, evidence that shippers were fined for violating the maximum legal passenger-­per-ton ratio has yet to be found. The fine was not assessed by the government, but was only recoverable by the wronged party (presumable the passengers) through filing suit in a U.S. circuit or district court. Because enforcement of the law was set up as a tort suit that had to be initiated by the wronged party, and with the wronged party being German immigrants, few violators were likely to be punished. As such, the maximum passenger-­per-ton limit set by the 1819 Passenger Act had little influence on the behavior of shippers in the German transatlantic passenger trade. In conclusion, German immigration to the Delaware valley from 1727 through 1835 was substantial. It totaled 108,991 in these years, with almost three-­ quarters of that total coming during the colonial period. It was large compared with other European migrations during the colonial and early republican periods, but declined in both absolute and relative importance after 1808. The typical journey was from the upper Rhine valley through a Dutch port to Philadelphia, though after 1784 emigrants from northern and eastern Germany embarking at German ports rose in importance. Migration was rigidly seasonal with arrival in America almost exclusively in the autumn. The average length of the overall journey, from home in Germany to Philadelphia, was about a third of a year, but could range from two months to two-­thirds of a year. The yearly variance in German immigration was substantial. Warfare on the migration route, particularly on the North American end, depressed migration

The flow of immigrants, 1727–1835   43 flows, with localized migration peaks occurring in years immediately following the end of a war. A quarter of the variance, however, can be explained by the volume of migration in the prior year, which is consistent with a process of chain-­migration. Finally, shipping services in Dutch and German ports were relatively available in most periods. Crowding levels on German immigrant ships were well below that seen in the slave trade. They also declined considerable over time. Passenger-­per-ton ratios on German immigrant ships to Philadelphia after 1815, and through to 1835, were 80 percent lower on average than on such ships during the colonial period. The variance in German immigration was primarily accommodated in that year by changing the passenger-­per-ton ratio on ships in the trade via altering the human to nonhuman cargo mix on those vessels. The quantity of shipping services, however, was relatively responsive thereafter—expanding significantly in the year immediately following unexpected spikes in emigration.

Notes   * Adapted and updated from chapters 2 and 3 of Farley Grubb, “Immigration and servitude in the colony and commonwealth of Pennsylvania: a quantitative and economic analysis,” Chicago: PhD Thesis, University of Chicago, 1984 [previously unpublished].   1 For examples, see Beidelman (1898); Diffenderffer (1899); Fogleman (1996); Fertig (1994: 192–235); Geiser (1901); Herrick (1926); Jacobs (1897); Smith (1947); Strassburger (1934); Wokeck (1981, 1999); Wolf (1976).   2 See Smith (1947: 320–3); Strassburger (1934: vol. 1, xvii–xxix, 3–6); Wokeck (1981: 251–61). Within the Strassburger (1934) collection of 324 ship arrivals only two had absent oath lists.   3 Failure to make this adjustment was the chief criticism of Strassburger’s estimates by Smith (1947: 320–1). For a discussion of the actual adjustments, see Wokeck (1981: 264–5). Freights were passengers paying full fare, namely those over age 12. Children under 12 paid half fare or if under age 4 no fare. These children were thus counted as half or no freight, respectively. As such, total passengers measured in freights undercounts the same total measured in bodies. For a discussion of passenger freight rates, see Chapters 3, 16, and 18.   4 See Wokeck (1981: 260–1, 266; 1999: 45–6); Fertig (1994: 193–203). The non-­ Philadelphia port totals are based on sketchier evidence and more guesswork. There was some indirect immigration into the region, e.g. some 600 Moravians entered through the port of New York between 1735 and 1765, see Jordan (1909: 33–53). German immigration into Pennsylvania via other colonies between 1727 and 1775, however, appears to have been small.   5 See Coldham (1992: 1, 6); Ekirch (1987: 27); Fogleman (1998: 55, 71); Grubb (2000: 94); Morgan (1987: 416).   6 Strassburger (1934: vol. 3, v–xvi) only counts 9,263 immigrants. But this is only a count of names. In many cases wives and children were enumerated but not listed by name. The lists abruptly end in 1808 due to the British continental blockage against the French, which included Dutch ports, and due to the Jeffersonian embargo.   7 See chapters 9 through 15 for more details about German redemptioner servitude and this evidence source.   8 These totals do not include two ships, Nancy and Patsy Rutledge, which were undated but listed by Strassburger with ships arriving in 1800. Since matches with these two ships could not be made they must have arrived after 1804. These ships included 122 immigrants (Strassburger 1934: vol. 3, 103–5).

44   German immigration to America, 1709–1835   9 It is possible that some matches were erroneously made due to the existence of common names. These types of matches, however, had their acceptance substantially discounted once the contract date exceeded the ship arrival date by a month. For more discussion of the procedure see chapters 14 and 18, and Wokeck (1981: 262–3). It is also possible that some of the non-­matches were non-­immigrant contracts, i.e. contract re-­sales recorded to look like sales of new arrivals. 10 A more conservative adjustment would take only those non-­matches that appear clustered in relatively unbroken series as indicating the presence of missing passenger lists—the other non-­matches being due to a failure to identify existing matches. This procedure yields at most 200 non-­matches owing to missing passenger lists. If the rest are simply missed matches, the servant percentage rises to 47.8. Using this servant percentage, these 200 non-­matches imply about 400 additional immigrants above the 7,837 named in Strassburger (1934: vol. 3). This adjustment implies a 5 percent undercount by the Strassburger collection of passenger lists. The comparison of the Strassburger lists with the names in “Book A of redemptioners” can also help correct some erroneous ship arrival dates listed in Strassburger. For example, the ship Holland listed in Strassburger as arriving 19 August 1796 must have arrived in September of 1795. The ship Amiable Matilda, undated, must have arrived in March of 1796. The ship Columbia, undated, must have arrived in April of 1796. The ship Industry, undated, must have arrived in January of 1796. The ship Pennsylvania listed as arriving in November of 1803 must have arrived in August of 1796. 11 “Philadelphia Customs House records;” “Book C of redemptioners;” Bromwell (1969). The Federal passenger lists from 1820 through 1830 substantially undercount immigrants and so need to be supplemented by the additional records mentioned, see Grubb (1990). The alternative sources mentioned include: Virchaux (1818); “Passenger List of the ship Elizabeth;” American Watchman, 7 January 1818; and Nile’s Weekly Register, 31 January 1818. See also Walker (1964: 1–41). 12 “Philadelphia Custom House records;” chapter 18. The term “arrivals” is used when referring to the Custom House records because not all passengers listed were immigrants. Some were merchants, returning citizens, shipping agents, tourists, and so on. 13 Bromwell (1969: 21–88, 176–7); Cohn (2009: 1–69, 127–8, 155–9); Hatton and Williamson (2005: 31–76); chapters 18 and 19. 14 See Beidelman (1898: 15–34); Jacobs (1897: 146–7); Diffenderffer (1897: 266; 1899: 33, 48, 53); Muhlenberg (1942: vol. 1, 28); Wokeck (1999: 50). 15 This impression also comes from work done on selected groups, mostly religious sects, see Fogleman (1996: 15–65); Fertig (1994: 228–35); Jordan (1909: 33–53); Krebs (1957: 32–42). Also diaries of immigrants convey this impression, e.g. see Mittelberger (1960: 26). Germans from other regions appear to be atypical or even accidental immigrants to Philadelphia. For examples, see the personal accounts of John Frederick Whitehead from Pomerania and Johann Carl Büttner from Saxony who immigrated to Philadelphia in 1773 in Klepp, Grubb, and Pfaelzer de Ortiz (2006). 16 The ships are the America from Hamburg with six passengers (1798), the Juno from Hamburg with two passengers (1802), the Margaret from Amsterdam with 269 passengers (1804), the Carolina from Fredericstad with four passengers (1804), the Margaret, port unknown, with 245 passengers (1805), the Verny from Amsterdam with 196 passengers (1805), the Liberty from Amsterdam with 149 passengers (1805), the Three Sisters from Bremen with 57 passengers (1805), the Aeolus from Tonningen with 11 passengers (1805), the Atlantic from Amsterdam with 81 passengers (1806), the Betsey from Tonningen with 67 passengers (1807), and the William P. Johnson from Holland with 153 passengers (1807). 17 For example, see discussions and applications of this theory in Baines (1991: 7, 39); Davis (1972: 156–8); Eltis (1983: 259); Greenwood (1975: 397–433); Sjaastad (1962: 80–93). The economic model treats location as a form of investment in human capital as well as a decision over the wealth-­maximizing choice of residence. The cost of

The flow of immigrants, 1727–1835   45 migration includes everything sacrificed in the act of migrating, not just out-­of-pocket cash expenses or the passage fare paid. 18 For example, see Beidelman (1898: 15–34); Diffenderffer (1897: 276–89; 1899: 14–20); Durnbaugh (1958: 321–2); Fertig (1994: 203–24); Jacobs (1897: 31–3); Knittle (1937: 1–12); Lemon (1972: 5); Sachse (1897: 276–89); Smith (1947: 50–2); Wegge (1999: 30–55); Wokeck (1981: 274–8). 19 In some cases religious persecution or the presence of war was related to more direct economic effects for selected groups that resulted in immigration, e.g. new conscription laws that did away with exemptions for married men in the 1740s (Wokeck 1981: 270) and proscription against Mennonites buying property in the Palatinate after 1733, see Smith (1929: 47). 20 Mittelberger (1960: 51) provides a similar account in 1756. 21 This is the factor price equalization theorem with mobile factors of production. Under   competition, regional resource endowments determine the ratio of marginal products, which in turn determine the ratio of factor payments. The initial difference in resource endowments between regions creates an opposite difference in the ratio of  marginal products and hence factor payments. This encourages the low paid mobile factor to migrate to the area where it is relatively scarce until such migration evens the difference in regional resource endowments to the point where the ratio of marginal products, and thus factor payments, are equalized down to the cost of migration. 22 Jones (1980: 59, 68, 77, 162–3, 265). See also Lemon (1972: 10–12); Lemon and Nash (1968: 1–21); Main (1982: 93–6); Perkins (1980: 44). 23 For a description of recruitment pamphlets, see Sachse (1897: 175–256); Diffenderffer (1899: 14–31). For examples of immigrant correspondence, see Durnbaugh (1960); Meuer (1913: 94–106); Mittelberger (1960); Diffenderffer (1899: 168–98); Trautmann (1981); Westergaard (1932: 9–14). 24 On the influence of immigrant letters sent home, see the opinions of Campbell (1955: 17); Diffenderffer (1899: 257); Eltis (1983: 259); Kalm (1937: vol. 1, 142–3); Wokeck (1981: 274–8); Smith (1947: 55–7). For contemporary studies of the effects of information on the decision and direction of migration, see Greenwood (1975: 405–6). Immigrant shippers in the eighteenth century knew that return letters were critical to stimulating future migration. John Dick advised the British Board of Trade that particular care should be taken to receive the first contingent of Germans kindly in Nova Scotia in 1775 stating, “For all the future Embarkations of these people will depend upon Letters they sent home at their Arrival,” cited in Smith (1947: 56). 25 Durnbaugh took this material from “Das Leben des Herrn Charles Hector Marquis St. George de Marsay,” 294 ff., bound manuscript in the private library of Dr. J.F. Gerhard Goeters, Bonn/Rhine. 26 See Baines (1991: 7, 22, 31, 39); Nash (1979: 178–80, 392–4, 413, 417); Smith (1981: 331). 27 See Fertig (1994: 225); Smith (1947: 336); Wokeck (1981: 262, n. 24). For an alternative regression specification that emphasizes war, along with short-­run relative food prices and mortality proxies for short-­run population pressures in Germany, see Fertig (1994: 225–8). While higher food prices and mortality rates in Germany are associated with greater migration to Philadelphia in the colonial period, the effects are   small compared with the war effect. Fertig’s estimates on these variables may also be simply tracking secular trend effects and uncorrected serial correlation movements. 28 Diffenderffer (1899: 203); Durnbaugh (1958: 334–5); Jordan (1909: 33–53); Muhlenberg (1942: vol. 1, 24–57); Wokeck (1981: 254, n. 10). For other accounts, see Kalm (1937: vol. 1, 1–14); Meurer (1913: 94–106). The fate of the passengers on the Argyle is unknown. 29 See Wokeck (1999: 45–6). For an alternative view, see Wokeck (1981: 267).

46   German immigration to America, 1709–1835 30 Fertig (1994: 225) finds a similar result. For discussions of chain-­migration among German immigrants and the role of the stock of previous migrants on subsequent migrations during the nineteenth century, see Baines (1991: 28); Dunlevy and Gemery (1978); Hatton and Williamson (2005: 51–68); Wegge (1998: 957–86). 31 See Bromwell (1969: 24, 28, 32, 36, 40, 44, 48, 52, 56, 60, 64, 68); Cohn (2009: 127–8). Jones (1992: 63–6) notes that after 1815 German-­language culture and newspapers in the middle states were on the wane, which may have reduced the attraction of these locations to German emigrants and reduced the chain-­migration information and support contacts during the first decades of the nineteenth century. 32 See Fisher, Heiken, and Hulen (1997); Hansen (1940: 103–5); Skeen (1981: 51–67); Walker (1964: 1–41, 250); and chapters 16, 17, and 18 for more details. 33 For a discussion of seasoning mortality, see chapter 3; Galenson (1981: 89–91); Morgan (1975: 184); Smith (1981: 25–73). 34 Derived from Neible (1906–1908); “Records of indentures.” See also Galenson (1981: 88–91); Morgan (1975: 182). 35 Sachse (1909); Muhlenberg (1942: vol. 1, 23–58); Durnbaugh (1960); Kalm (1937: vol. 1, 1–16); Meurer (1913); Smith (1947: 210); Wesley (1902: vol. 1, 106–37). 36 Seven of these ships were derived from Strassburger (1934: vol. 1, ship #6, 7, 8, 9, 10, 11, 91). The date when the oath of abjuration was taken was used as the arrival date. Six ships come from immigrant diaries, see note 35. The last five ships were taken from Jordan (1909: 33–53). 37 Several sources report ships having taken six or so months at sea, but it is hard to find direct confirmation of these cases with exact departure and arrival dates. See Duffy (1951: 27, 32); Jacobs (1897: 141–4); Geiser (1901: 48); Diffenderffer (1899: 61–5, 260–2). 38 Between 1841 and 1855 immigrant ships arriving in Quebec, Canada from Britain averaged around 44 days at sea (Eltis 1983: 271). See also chapter 19. 39 The samples of 66 ships comes from matching the ships listed in Strassburger (1934: vol. 1) with the information in “Ships registered for the port of Philadelphia,” (1899–1904). Only if both the ship’s name and captain were the same was a match considered. Only ships sailing from Dutch ports to Philadelphia were considered. All registered tonnage was converted into 1695 measured tons, see McCusker (1967: 82–91); Walton (1967b: 392–8); French (1973: 434–43). The conversion factor combines French (1973: 443) and McCusker (1967: 89) and are as follows: registered tonnage multiplied by 1.971 for ships registered between 0 and 50 tons, 1.518 for ships between 51 and 100 tons, 1.422 for ships between 101 and 150 tons, and 1.338 for ships registered between 151 and 200 tons equals measured tonnage as expressed in the 1695 formula. This results in a capacity or volume measure as opposed to a burden or weight measure and, therefore, allows for a comparison of cargo composition and capacity. Total passengers per ship follows that in Wokeck (1999: 45–6). 40 North (1968: 953–70); Walton (1967a: 67–8). For example, see the instructions to Thomas and Adrian Hope, merchants of Amsterdam, dated 1742, in Bruchey (1966: 182–3). 41 The correlation coefficient between passengers per ton and ship’s tonnage when measured by the 1695 formula is –0.21 in the sample and it is statistically significant above the 0.05 level. 42 Geiser (1901: 47) suggested that crowding may have been a function of the relative profits on passengers versus merchandise. At the margin, however, competition would equalize the profits earned on these two categories of freight. 43 The 1815–19 sample of German passenger ships does not include the infamous voyage of the German passenger ship April, which only made it as far as New Castle, Delaware, landing there December of 1817. The exaggeration of the deaths and trauma encountered on the voyage, and the misreporting and duplicate reporting of the accounts of the voyage in the newspapers, were the single most important event

The flow of immigrants, 1727–1835   47 propelling the passenger legislation of 1819 (chapter 17; Swierenga and Lammers 1994). Without providing any evidence or analysis, Jones (1992: 57) asserted that “. . . the Passenger Act it [Congress] adopted in March 1819, gave the coup de grâce to the importation of indentured labor.” Such a claim is factually incorrect and logically unsound (chapters 17 and 18).

References “Account of vessels arrived at the port of Philadelphia kept by Captain Nathaniel Falconer, Health Officer of said port, viz., September 30, 1789 to May 3, 1794,” Harrisburg, PA: Unpublished manuscript held at the State Archives of Pennsylvania, microfilm RG-­41, roll 1. Adams, D. R. Jr. (1970) “Some evidence on English and American wage rates, 1790–1830,” Journal of Economic History, 30: 499–520. Bailyn, B. (1986) Voyagers to the west, New York: Alfred A. Knopf. Baines, D. (1991) Emigration from Europe, 1815–1930, New York: Cambridge University Press. Beidelman, W. (1898) The story of Pennsylvania Germans, Easton, PA: Express Book Prints. “Book A of redemptioners, 1785–1804,” Philadelphia: Unpublished manuscript held at the Historical Society of Pennsylvania. “Book C of redemptioners, 1817–1831,” Philadelphia: Unpublished manuscript held at the Historical Society of Pennsylvania. Bromwell, W. J. (1969) History of immigration to the United States, New York: Arno Press. Bruchey, S., ed. (1966) The colonial merchant, New York: Harcourt, Brace, and World. Brumbaugh, M. (1907) A history of the German Baptist Brethren in Europe and America, Elgin, IL: Brethren Publishing House. Campbell, M. (1955) “English immigration on the eve of the American revolution,” American Historical Review, 61: 1–20. Cohn, R. L. (2009) Mass migration under sail: European immigration to the antebellum United States, New York: Cambridge University Press. Coldham, P. W. (1992) Emigrants in chains, Baltimore: Genealogical Publishing Company. Crowthers, S. J. (1973) “The shipbuilding output of the Delaware Valley, 1722–1776,” Proceedings of the American Philosophical Society, 117: 90–104. Davis, L. E., Easterlin, R. A., Parker, W. N., Brady, D. S., Fishlow, A., Gallman, R. E., Lebergott, S., Lipsey, R. E., North, D. C., Rosenberg, N., Smolensky, E., and Temin, P. (1972) American economic growth: an economist’s history of the United States, New York: Harper and Row. Deane, P. (1965) The first industrial nation, New York: Cambridge University Press. Diffenderffer, F. R. (1897) “The German exodus to England in 1709,” part 1 of “The German influence in its settlement and development,” Pennsylvania German Society, 7: 255–414. Diffenderffer, F. R. (1899) “The German immigration into Pennsylvania through the port of Philadelphia, and ‘the redemptioners’,” part 7 of “The German influence in its settlement and development,” Pennsylvania German Society, 10: 1–328. Dorn, W. (1940) Competition for empire, 1740–1763, New York: Harper and Row. Duffy, J. (1951) “The passage to the colonies,” The Mississippi Valley Historical Review, 28: 21–38.

48   German immigration to America, 1709–1835 Dunlevy, J. A. (1980) “Nineteenth-­century European immigration to the United States: intended versus lifetime settlement patterns,” Economic Development and Cultural Change, 29: 77–90. Dunlevy, J. A., and Gemery, H. A. (1978) “Economic opportunity and the responses of ‘old’ and ‘new’ migrants to the United States,” Journal of Economic History, 38: 901–17. Durnbaugh, D. F. (1958) “Christopher Sauer Pennsylvania-­German printer,” Pennsyl­ vania Magazine of History and Biography, 83: 316–40. Durnbaugh, D. F. (1960) “Two early letters from Germantown,” Pennsylvania Magazine of History and Biography, 84: 219–33. Ekirch, A. R. (1987) Bound for America, New York: Oxford University Press. Eltis, E. (1983) “Free and coerced transatlantic migration: some comparisons,” American Historical Review, 88: 251–80. Fertig, G. (1994) “Transatlantic migration from the German-­speaking parts of central Europe, 1600–1800: proportions, structures, and explanations,” in N. Canny (ed.) Europeans on the move: studies on European migration, 1500–1800, New York: Oxford University Press. Fisher, R. V., Heiken, G., and Hulen, J. B. (1997) Volcanoes, crucibles of change, Princeton, NJ: Princeton University Press. Fogleman, A. S. (1995) “Immigration, German immigration and 18th-century America,” in E.Reichman, L. J. Rippley, and J. Nagler (eds.) Emigration and settlement patterns of GermancCommunities in North America, Nashville, IN: Max Kade German-­ American Center. Fogleman, A. S. (1996) Hopeful journeys: German immigration, settlement, and political culture in colonial America, 1717–1775, Philadelphia: University of Pennsylvania Press. Fogleman, A. S. (1998) “From slaves, convicts, and servants to free passengers: the transformation of immigration in the era of the American revolution,” Journal of American History, 85: 43–76. French, C. J. (1973) “Eighteenth-­century shipping tonnage measurements,” Journal of Economic History, 33: 434–43. Galenson, D. W. (1981) White servitude in colonial America, New York: Cambridge University Press. Geiser, K. F. (1901) Redemptioners and indentured servants in the colony and commonwealth of Pennsylvania, New Haven, CT: Tuttle, Morehouse, and Taylor Company. Grabbe, H. G. (1989) “European immigration to the United States in the early national period, 1783–1820,” Proceedings of the American Philosophical Society, 133: 190–214. Greenwood, M. J. (1975) “Research on internal migration in the United States: a survey,” Journal of Economic Literature, 13: 397–433. Grubb, F. (1990) “The reliability of U.S. immigration statistics: the case of Philadelphia, 1815–1830,” International Journal of Maritime History, 2: 29–54. Grubb, F. (2000) “The transatlantic market for British convict labor,” Journal of Economic History, 60: 94–122. Hansen, M. L. (1940) The Atlantic migration 1607–1860, Cambridge, MA: Harvard University Press. Hassall, A. (1907) The balance of power 1715–1789, New York: Macmillian Company. Hatton, T. J., and Williamson, J. G. (2005) Global migration and the world economy, Cambridge, MA: MIT Press.

The flow of immigrants, 1727–1835   49 Herrick, C. A. (1926) White servitude in Pennsylvania, Philadelphia: John Joseph McVey. Hutchinson, E. P. (1981) Legislative history of American immigration policy, 1798–1965, Philadelphia: University of Pennsylvania Press. Jacobs, H. E. (1897) “The German emigration to America, 1709–1740,” Pennsylvania German Society, 8: 29–150. Jones, A. H. (1980) Wealth of a nation to be, New York: Columbia University Press. Jones, M. A. (1992) American immigration, Chicago: University of Chicago Press, second edition. Jordan, J. W. (1909) “Moravian immigration to Pennsylvania, 1734–1765,” Pennsylvania Magazine of History and Biography, 33: 33–53. Kalm, P. (1937) Travels in North America, 2 vols, New York: Wilson-­Erickson, Adolph B. Benson trans. and ed. Klein, H. (1978) The middle passage, Princeton, NJ: Princeton University Press. Klepp, S. E., Grubb, F., and Pfaelzer de Ortiz, A., eds. (2006) Souls for sale: two German redemptioners come to revolutionary America: the life stories of John Frederick Whitehead and Johann Carl Büttner, University Park, PA: Pennsylvania State University Press. Knittle, W. A. (1937) Early eighteenth century Palatine emigration, Philadelphia: Dorrance and Company. Krebs, F. (1957) “Emigrants from Baden-­Durlach to Pennsylvania, 1749–1755,” National Genealogical Society Quarterly, 45: 32–42. Landes, D. S. (1969) The unbound Prometheus, London: Cambridge University Press. Lemon, J. T. (1972) The best poor man’s country, Baltimore: The John’s Hopkins Press. Lemon, J. T., and Nash, G. B. (1968) “The distribution of wealth in eighteenth century America: a century of change in Chester County, Pennsylvania, 1693–1802,” Journal of Social History, 2: 1–21. Main, G. L. (1982) Tobacco colony life in early Maryland 1650–1720, Princeton, NJ: Princeton University Press. McCusker, J. J. (1967) “Colonial tonnage measurement: five Philadelphia merchant ships as a sample,” Journal of Economic History, 27: 82–91. Mittelberger, G. (1960) Journey to Pennsylvania in the year 1750 and return to Germany in the year 1754, Cambridge: Harvard University Press. Morgan, E. S. (1975) American slavery, American freedom, New York: W. W. Norton. Morgan, K. (1987) “English attitudes towards convict transportation, 1718–1775,” History, 72: 416–31. Muhlenberg, H. M. (1942) The journals of Henry Melchior Muhlenberg, 3 vols, Philadelphia: The Muhlenberg Press, T.G. Tappert and J. W. Doberstein trans. Meuer, J. P. (1913) “From London to Philadelphia, 1742,” Philadelphia Magazine of History and Biography, 37: 94–106. Nash, G. B. (1979) The urban crucible, Cambridge, MA: Harvard University Press. Neible, G. W., ed. (1906–1908) “Servants and apprentices bound and assigned before James Hamilton Mayor of Philadelphia, 1745,” Pennsylvania Magazine of History and Biography, 30: 348–52, 427–36; 31: 83–102, 195–206, 351–67, 461–73; 32: 88–103, 237–49, 351–70. North, D. C. (1968) “Sources of productivity in ocean shipping, 1600–1850,” Journal of Political Economy, 76: 935–70. “Passenger list of the ship ‘Elizabeth,’ which arrived at Philadelphia in 1819,” Pennsyl­ vania Magazine of History and Biography, 25: 257–60.

50   German immigration to America, 1709–1835 Perkins, E. J. (1980) The economy of colonial America, New York: Columbia University Press. “Philadelphia Customs House records, 1800–1883,” Washington, DC: Unpublished manuscript held at the National Archives, microfilm #425, rolls 16–50. “Pilot’s report for the port of Philadelphia, April 26, 1793 to September 5, 1808,” Harrisburg, PA: Unpublished manuscript held at the State Archives of Pennsylvania, microfilm RG-­41, roll 1. “Record of indentures of individuals bound out as apprentices, servants, etc. and of German and other redemptioners in the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773,” Philadelphia: Unpublished manuscript, Philadelphia City Archives. Roberts, P. (1947) The quest for security 1715–1740, New York: Harper and Row. Roeber, A. G. (1993) Palatines, liberty, and property, Baltimore: Johns Hopkins Press. Sachse, J. F. (1897) “The Fatherland,” the first part of part 1 of “The German influence in its settlement and development,” Pennsylvania German Society, 7: 175–289. Sachse, J. F. (1909) “A missive from Pennsylvania in the year of grace 1728,” Pennsylvania German Society, 28: 5–25. Schlosser, F. C. (1844) History of the eighteenth century, 10 vols, London: Chapman and Hall. “Ships registered for the port of Philadelphia, 1727–1775,” (1899–1904) Pennsylvania Magazine of History and Biography, 23: 254–64, 370–85, 498–515; 24: 108–15, 212–23, 348–66, 500–19; 25: 118–31, 266–81, 400–16, 560–67; 26: 126–43, 280–4, 390–400, 470–5; 27: 94–107, 238–47, 346–70, 482–98; 28: 84–100, 218–35, 346–74, 470–504. Sjaastad, L. A. (1962) “The cost and return of human migration,” Journal of Political Economy, 70: 80–93. Skeen, E. C. (1981) “The year without a summer: a historical view,” Journal of the Early Republic, 1: 51–67. Smith, A. (1937) An inquiry into the nature and causes of the wealth of nations, New York: Modern Library. Smith, A. E. (1947) Colonists in bondage, New York: W. W. Norton. Smith, B. G. (1981) “Struggles of the ‘lower sort’: the lives of Philadelphia’s laboring people, 1750–1800,” Los Angeles: PhD Dissertation, UCLA. Smith, C. H. (1929) The Mennonite immigration to Pennsylvania, Norristown, PA: Norristown Press. Strassburger, R. B. (1934) Pennsylvania German pioneers, 3 vols., Norristown, PA: Pennsylvania German Society. Swierenga, R. P. and Lammers, H. (1994) “ ‘Odyssey of woe’: the journey of the immigrant ship April from Amsterdam to New Castle, 1817–1818,” Pennsylvania Magazine of Biography and History, 118: 303–23. Trautmann, F. (1981) “Pennsylvania through a German’s eyes: the travels of Ludwig Gall, 1819–1820,” Pennsylvania Magazine of History and Biography, 105: 35–65. Virchaux, H. T. (1818) “Report [22 January 1818] to the German Society of the ship April,” Philadelphia: Unpublished manuscript held by the German Society of Pennsylvania. Walker, M. (1964) Germany and the emigration, 1816–1885, Cambridge, MA: Harvard University Press. Walton, G. M. (1967a) “Sources of productivity change in American colonial shipping, 1675–1775,” Economic History Review, 22: 67–78.

The flow of immigrants, 1727–1835   51 Walton, G. M. (1967b) “Colonial tonnage measurement: a comment,” Journal of Economic History, 27: 392–8. Wegge, S. A. (1998) “Chain migration and information networks: evidence from nineteenth-­century Hesse-­Cassel,” Journal of Economic History, 58: 957–86. Wegge, S. A. (1999) “To part or not to part: emigration and inheritance institutions in nineteenth-­century Hesse-­Cassel,” Explorations in Economic History, 36: 30–55. Wesley, J. (1902) The journal of the Reverend John Wesley, A.M., 4 vols, London: Robert Cully, N. Curnock, ed. Westergaard, W., trans. (1932) “Two Germantown letters of 1738,” Pennsylvania Magazine of History and Biography, 56: 9–14. Wokeck, M. S. (1981) “The flow and composition of German immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, 105: 249–78. Wokeck, M. S. (1999) The trade in strangers: the beginnings of mass migration to North America, University Park, PA: Pennsylvania State University Press. Wolf, S. G. (1976) Urban village: population, community, and family structure in Germantown, Pennsylvania, 1683–1800, Princeton, NJ: Princeton University Press. Yoder, D., ed. (1980) Pennsylvania German immigrants, 1709–1786: lists consolidated from yearbooks of the Pennsylvania German Folklore Society, Baltimore: Genealogical Publishing Company. Yoder, D., ed. (1981) Rhineland emigrants: lists of German settlers in colonial America, Baltimore: Genealogical Publishing Company.

3 The transatlantic shipping market*

By the end of the seventeenth century the power of royally chartered trading companies to monopolize transatlantic routes had waned. Within the confines of the Navigation Acts, eighteenth-­century British shipping was openly competitive.1 However, the transportation of German immigrants from Rotterdam to America in the middle of the eighteenth century might have been the exception. Several authors have suggested that a few shippers monopolized this trade prior to the Seven Years War.2 Monopolization of the German passenger trade would have dire consequences for the emigrants. A monopolist’s goal of higher profits could only be achieved by artificially restricting the supply of services, thus causing the price to rise far above the cost of supplying the service. Noncompetitive behavior has been used to explain why the 55,000 Germans migrating to Philadelphia between 1730 and 1756 experienced conditions worse than British emigrants.3 With no alternative shipping readily available, Germans might have had little choice but to accept relatively higher transportation prices and less tolerable voyage conditions. High passage fares could have discouraged German emigration or led to more Germans relying on servitude to finance their voyage compared with British emigrants. Poor voyage conditions could have led to more discomfort, sickness, and death among German emigrants compared with British emigrants. In addition, Germans who paid for passage through servitude could have had little choice but to accept the less desirable redemptioner contracts instead of the standard British indentured contracts that prevailed in competitive markets. Thus, German servants may have labored under less advantageous conditions in America compared with British servants.4 Recently, the conclusion that German immigrants experienced the worst voyage conditions of any immigrant group on the North Atlantic passage has been challenged. In addition, it has been suggested that redemptioner contracts were not necessarily less desirable than indentured contracts because redemptioner financing offered a more flexible contract.5 Although these findings call into question some of the suggested consequences of monopolization in the German immigrant trade, the conclusion that the trade was monopolized prior to the Seven Years War has remained unchallenged. Therefore, the existing evidence for and against monopolization will be evaluated to better establish the

The transatlantic shipping market   53 market structure of German immigrant shipping in the middle of the eighteenth century. The relevant evidence for judging effective monopolization should come from the market characteristics of monopoly. An effective monopolist should be observed making profits above what is normal for competitive merchants. The monopolist would acquire these profits by artificially restricting the supply of services, a restriction that would lead to higher prices. Higher prices and excess profits would attract competitors who would increase the supply of services, drive prices down, and eliminate the monopolist’s profits. Therefore, the monopolist should be observed driving existing competitors out and keeping potential competitors away through legal or extra-­legal means. Because maintaining a monopoly is easier for one or at best a few merchants, a high degree of concentration in the trade should be observed, and easy entry and exit from the market should not be observed. If these conditions are not met, then effective monopolization of the German immigrant trade would be unlikely.

Evaluation of the literary evidence suggesting monopolization There are three separate allegations of monopolization, each supported by quotations from different contemporary observers of the trade. The first allegation, and the only one from within the Pennsylvania market, presented by Frank Diffenderffer, came from the Pennsylvania-­German printer Christopher Sauer. In 1755 Sauer wrote to Governor Morris about the wrongs done to German immigrants in the late 1730s. Steadman at the time bought a license in Holland that no captain or merchant could load any [German emigrants] as long as he had not two thousand loaded. I wrote to the magistrate at Rotterdam, and immediately the “Monopolium” was taken from John Steadman.6 John and Charles Stedman commanded eight German immigrant voyages between 1731 and 1740, one each in 1737 and 1738 (see Table 3.1). They may also have been partial owners in three other German immigrant voyages in the 1730s. If the Stedmans had a license to load at Rotterdam the first 2,000 Germans bound for Philadelphia each year during the 1730s, they would have had substantial monopoly power because only in 1738 did emigration definitely exceed 2,000.7 Table 3.1 outlines the activities of the Stedmans in the German passenger market between 1731 and 1754. They only handled 18.7 percent of the voyages and 20 percent of the adult males arriving in Philadelphia, although they were probably the largest single shipper. The Stedmans’ market share in the 1730s was not much different than in the 1740s and 1750s. They acquired the capacity to carry 2,000 immigrants yearly around 1738, but never came close to carrying 2,000 in any year.8 They probably possessed the carrying capacity to monopolize

Shipments

  1 (25.0%)b   1 (9.1)   1 (14.3)   1 (50.0)  0  0   2 (28.6)   4 (25.0)   1 (12.5)   1 (16.7)   2 (20.0)  0   3 (30.0)  0  0  0   1 (20.0)   1 (12.5)   3 (12.5)   4 (26.7)   4 (25.0)   4 (18.2)   4 (15.8)   5 (26.3)

43 (18.7)

Year

1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754

Totals

57 (24.3%)b 73 (9.0) 71 (16.6) 89 (63.6) 0 0 259 (41.7) 321 (26.5) 24 (4.5) 68 (16.8) 174 (27.2) 0 171 (27.2) 0 0 0 14 (4.1) 124 (21.3) 397 (15.0) 405 (23.7) 336 (17.9) 423 (19.7) 420 (23.0) 507 (24.3)

4,781 (20.0)



Adult males carried a

Partial owners (Lydia) Partial owners (Patience) Partial owners (St. Andrew, Patience, and Lydia) Partial owners (St. Andrew, Patience, Brothers, and Nancy) Partial owners (St. Andrew, Patience, Brothers, and Nancy)f Partial owners (St. Andrew, Brothers, Nancy, and Halifax) Partial owners (Patience, Brothers, Peggy, and Halifax)f Partial owners (Nancy, Barclay, Brothers, Peggy, and Halifax)f

Partial owners (Lydia, St. Andrew, and Endeavor)

Captain (St. Andrew and Charming Nancy) Captain (St. Andrew and Charming Nancy) and partial owners of two ships named Thistled Partial owners (Lydia)e Partial owners (Lydia) Captain (St. Andrew) and partial owners (Lydia)

Captain (Pennsylvania Merchant)c Captain (Pennsylvania Merchant) Captain (Pennsylvania Merchant) Captain (St. Andrew)

Methods and ships used

Table 3.1 The activities of the Stedman family in shipping German immigrants to Philadelphia, 1731–54

Notes The activities include those of John, Charles, and Alexander Stedman as captains or partial owners of vessels that transported German immigrants from Holland or London to Philadelphia. a Includes those who were required to sign the loyalty oaths which provides the most direct, consistent, and accurate measure of the relative number of passengers across shippers. b The percentage of the respective totals from German immigrant shipping to Philadelphia are in parentheses. c John Stedman captained voyages in 1731–34 and in 1737–38. Charles Stedman captain voyages in 1737, 1738, and 1741. d 1738 includes deliveries by two different vessels named Thistle commanded by captains not otherwise employed by the Stedmans before or after in the trade. The Stedmans registered two vessels named Thistle in Philadelphia, in July 1750 and May 1752. That these are the same vessels and were under the control of the Stedmans in 1738 is not known, possibly making the estimated numbers high. e The Stedmans registered 14 vessels at the port of Philadelphia between 1731 and 1754: Charming Nancy (1736), Lydia (1738), Argyle (1743), St. Andrew (1749), Nancy (1749), Brothers (1750), Patience (1750), Thistle (1750), Endeavor (1750), Halifax (1751), Minerva (1752), Thistle (1752), Barclay (1752), and Peggy (1754). Since eighteenth century ship names were commonly duplicated, consistency across the captains employed, consignments of cargo, and the port register were used to confirm that the German immigrant ships in question were probably owned by the Stedmans. f If the shipments of redemptioners consigned to the Stedmans by other merchants for sale in Philadelphia were included, it would raise the Stedmans’ total to 10 ships and 1,003 men in 1751 (62.5 and 53.3 percent of the totals, respectively), four ships and 497 men in 1753 (21.1 and 27.2 percent of the totals, respectively), and cause no change in 1754. This consignment information was only recorded in 1751 and 1753–74, see Table 3.3 and text.

Sources: Derived from Ralph B. Strassburger, Pennsylvania German Pioneers Vol. 1 (Norristown, Pa., 1934); “Ship Registers for the Port of Philadelphia, 1726–1775,” Pennsylvania Magazine of History and Biography Vols. 23–8 (1899–1904); and Wokeck, “Flow and Composition,” 260–1.

56   German immigration to America, 1709–1835 the trade between 1734 and 1748, having the potential to transport all the passengers who actually emigrated, except in 1738 when emigration exceeded 3,000. However, even with a monopoly license, they could not prevent other shippers from taking a substantial portion of the trade. Without strict government enforcement of a legal monopoly, which according to Sauer did not long continue, the Stedmans were not powerful enough to exclude other British merchants or even other Philadelphia merchants from the German immigrant trade. Charles Stedman was only the seventeenth largest shipowner in Philadelphia.9 Therefore, it seems unlikely that the Stedmans exercised effective monopoly power in the market. The Stedmans may have tried to acquire a monopoly license because of their changing role in the trade. Between 1737 and 1739 they made the transition from being active ship captains to just partial owners of the vessels involved in the trade. At the same time they expanded their yearly carrying capacity by investing in more ships. They increased their involvement in the German passenger trade from one ship a year to at least two in 1737 and four in 1738. Committing four vessels to the trade in 1738 may have been risky because the yearly flow of German emigrants was very volatile. From 1733 to 1737 German passengers averaged only 465 per year and the number of voyages averaged less than three per year. Since some of the same ships in the 1750s averaged 300 passengers per voyage, underutilized freight space and the cost of searching for alternative cargo may have been a problem. For example, in 1737 Hope & Co. would not finance a separate immigrant voyage to Georgia unless the Georgia trustees recruited at least 140 to 150 immigrants rather than the 60 proposed. The costs of underutilized freight space or excessive port delays could be avoided by finding some way of guaranteeing an emigrant cargo. Therefore, given the commitment in 1738 of four vessels to the trade, it would have been sensible for the Stedmans to seek the rights to load the first 2,000 emigrants, an effort that proved to be unsuccessful.10 The cause of Sauer’s complaint against John Stedman may have been, rather than true monopolization, the loss of life and illness aboard the St. Andrew, commanded by John Stedman, by the time it arrived in Philadelphia in 1738. John Stedman had been praised for other voyages. Christopher Schultz, who arrived on the St. Andrew in 1734, wrote of Captain Stedman, “We had a very good captain, who kept strictly to his contract, and very able sailors, who had very much patience with us.”11 Unsuccessful voyages were commonly attributed to the cruelty, incompetence, or monopoly power of the shipper instead of bad luck or changes in voyage plans caused by the weather. The second and most convincing allegation of monopolization was uncovered by Abbot E. Smith. He quoted a letter written in 1739 by the British secretary of legation at the Hague, “Till of late there was but one Merchant (Zachary Hope) at Rotterdam, with his Associates, who was allowed to answer for, and transport these [German] emigrants.”12 However, Smith did not tell the full story behind the quotation. In February, 1739, the trustees for the colony of Georgia wrote to Robert Trevor, British minister at the Hague, and to D. Wolters, English royal

The transatlantic shipping market   57 agent at Rotterdam, informing them that seven Salzburgers had set out for Rotterdam, by way of Augsburg and Frankfort, to take passage to England and from there to Georgia. The trustees further requested that Wolters and Trevor assist these Georgia colonists if any difficulties arose on their passage through Holland.13 The full content of Trevor’s reply was to inform the Georgia trustees that sometime before 1739 the Dutch had instituted a passport policy for German emigrants passing through Holland in response to complaints by local inhabitants. The passports had to be acquired by a “subject of credit and substance” who would guarantee the conduct, maintenance, and speedy transportation of the emigrants through Holland. Trevor also claimed that prior to 1739 only one merchant (unnamed) was allowed to acquire these passports, but by 1739 the government ended this legal monopoly to prevent the potential abuses associated with monopoly behavior. After 1738 any subject of credit and substance could acquire the passports. Trevor recommended that the Georgia trustees place an agent in Rotterdam to acquire the passports whenever prospective Georgia colonists were to pass through Holland. Apparently the Salzburgers in question had little difficulty embarking for Georgia in the spring of 1739.14 The history of the Dutch passport regulation and its possible creation of a monopoly in the German passenger trade prior to 1739 can be traced in the correspondence of emigrant recruiters for the colony of Georgia. The cause of the regulation was the trouble created by 50 Swiss families from Zurich who arrived in Rotterdam in the fall of 1734. They had resolved to migrate to England and then to British America but had no money for passage. They evidently refused to use redemptioner servitude to finance their migration and expected colonial recruiters to pay their expenses. However, no colony had invited, encouraged, or promised to pay for their migration. They were supported by the charity of the Rotterdam Magistrates and shipped to England from a collection taken by the inhabitants of the city. Passport regulations instituted in May, 1735 were meant to prevent the re-­occurrence of such behavior.15 The only recorded enforcement of the regulation occurred in the spring of 1736 when 300 Palatines were arrested “contrary to all expectation” at Schenkenschauz on the Dutch frontier “under the pretense that they first settle with a merchant of Rotterdam, named Zachary Hope.”16 This was also the first instance where Zachary Hope was named as the monopolizing merchant; surprisingly throughout the Georgia correspondence the Stedmans were never mentioned. Therefore, Hope & Co. may have had a legal monopoly over the German emigrant trade through Holland for the years 1735 through 1738, perhaps maybe only for part of 1736. However, from 1735 to 1738 there were many shipments of Germans from Rotterdam to the colonies that did not pass through the hands of Hope & Co. The colony of Georgia, employing the London merchant Peter & J.C. Simonds, had few problems shipping German colonists through Holland to Georgia. The Pennsylvania Brethren relied on the Rotterdam merchant Peter De Koker to aid their members. Two hundred Swiss were able to contract directly with a ship

58   German immigration to America, 1709–1835 sailing from Rotterdam to Carolina in 1736. Hans Trachsler, in the account of his migration to Carolina in 1736, claimed that those with money could engage passage by themselves with the many ships at Rotterdam, but that those without money would “be taken charge of and sent over by Mr. Hope, a prominent merchant.”17 Therefore, the larger German passenger market was competitive. Any legal monopoly Hope & Co. may have had was confined to “leftover” redemptioner servants. If emigrants reached the Dutch border without money or sponsorship for passage or without having agreed to purchase passage via servitude with some other merchant, then by default they were directed to Hope & Co. These emigrants would only be allowed to enter Holland if they employed the Hopes to finance their migration using redemptioner servitude. The ability of the Hopes to use this passport regulation to restrict competition for German redemptioners was very limited because the recruiting of German emigrants beyond the Dutch border was competitive. Hope & Co. also failed to control the market for redemptioners in Holland. In 1735, 1736, and 1737, recruiters from Georgia could have shipped German redemptioner servants if they had been willing to put up the financing. In 1738 Captain Thomson shipped German redemptioners to Georgia “at his own risk.” In the fall of 1737 agents for Georgia intercepted a group of German servants at Cowes who were under contract by Hope & Co. to go from Rotterdam to Philadelphia, gained their release from the Hopes contract, and diverted them to Georgia under a new contract.18 It seems unlikely, then, that Hope & Co. could have excluded competitors from the market and exercised effective monopoly power. The third allegation of monopolization, also presented by Abbot E. Smith, was by John Dick in a report sent to the British Board of Trade in 1750. Dick complained that the Hopes and John Stedman had monopolized the German immigrant trade in Rotterdam. In 1750 Dick had recruited German colonists for Nova Scotia. As these emigrants arrived at the Dutch border they changed their destination to Pennsylvania after recruiters from competing colonies described Nova Scotia in harsh terms. Frustrated in his recruitment effort, Dick charged that the Hopes and John Stedman had monopolized the German immigrant trade. Smith interpreted this evidence as indicating that the Hopes and the Stedmans were able to maintain their monopoly, even after losing their legal right in 1739, through extralegal recruitment methods in the 1740s and 1750s. But Dick was able to ship 860 Germans to Nova Scotia in 1752 raising questions about the validity of his claims about the Hopes and John Stedman. The difficulty John Dick experienced in recruiting German colonists was probably the result of intense competition rather than restrictions to competition. Several colonies— Georgia, Massachusetts, Nova Scotia, New York, Pennsylvania, Virginia—sent recruiters to the Rhine-­land to compete for immigrants. Dick combated Joseph Crellius, agent for Massachusetts, in the press of German towns, praising destinations he offered while denouncing rival colonies.19 The competitive success of such shippers as the Hopes and the Stedmans in the Pennsylvania market should not be confused with monopoly power over the

The transatlantic shipping market   59 immigrant business. Other recruiters were not excluded from competition. But it was difficult to counter the amenities offered by Pennsylvania, often the preferred New World destination of German emigrants after 1709. For example, in 1737 John Kramer tried to recruit 60 German servants for Georgia, but he found that all the servants in Rotterdam wanted to go to Pennsylvania. Kramer concluded that only after all the Pennsylvania ships had departed might some ser­ vants be convinced to take passage to Georgia.20 Germans preferred Pennsylvania because it possessed an established and prosperous German community and was hospitable to German immigrants in terms of ethnic and religious tolerance. A large flow of independent and reliable information had been sent to Germany by Pennsylvania Germans which corroborated the information presented by Pennsylvania recruiters.21 Another reason for Pennsylvania’s competitive success in recruiting German immigrants was the willingness of shippers serving the Philadelphia market to transport immigrants without requiring advance payment, either in cash or by signature on indentures. Passengers were allowed to arrange payment after their arrival in Philadelphia. This redemptioner system allowed poor immigrants a chance to find alternative ways of paying their passage, such as borrowing from American relatives, before being forced to resort to servant contracting. The redemptioner system was invented in the Rotterdam to Philadelphia market in the 1720s and quickly became universal in that market. Recruiters serving other markets were reluctant to adopt the innovation, clinging to the older indentured contract or preferring to avoid the difficulty of dealing in servants by recruiting emigrants who could pay for their own passage.22 Although various colonies competed for German emigrants, Pennsylvania was the German preference. The supposed monopolization of the passenger or redemptioner trade from Rotterdam to Philadelphia may have been enough to cause the noncompetitive behavior attributed to shippers. Hope & Co. provided financing for other merchants who carried emigrants under the redemptioner system. Only once did Hope & Co. take a role in actually shipping emigrants to Philadelphia. Although the proportion of German redemptioners who used Hope & Co. could not be ascertained, the frequency with which the Hopes were mentioned as a redemptioner merchant suggests that their share was substantial. They may have specialized in the trade, more than other Rotterdam merchants, because of their British and North American family connections.23 Without a legally enforced monopoly, however, the Hopes were probably not powerful enough to prevent competition from spoiling efforts to reap monopoly profits. They were only one of a number of prominent merchant houses in Rotterdam and certainly not the largest. In 1742 John Hancock pointed out in a letter to Thomas and Adrian Hope that he had been advised to employ a different merchant house in Holland but would not do so, as long as the Hopes treated him fairly. In 1749 the London firm of John Hunt & Issac Greenleafe decided to enter the Rotterdam to Philadelphia market in German servants and was able to negotiate both with John Stedman and Hope & Co. Hunt & Greenleafe rejected Stedman’s conditions but accepted Hope’s, suggesting that competition existed

60   German immigration to America, 1709–1835 between merchants providing redemptioner financing. This also illustrates the ease with which new shippers could enter the market.24 Once at their port of embarkation, German emigrants may have been able to negotiate passage with any merchant or captain.25 Under the redemptioner system the initial recruiter, like Hope & Co., would lose the money spent on delivering the emigrants to port (providing food, shelter, toll fees, passports, etc.) if the emigrants were able to sail with other merchants. Recruiting may therefore have been a risky business. The reluctance of recruiters, like the Hopes, to allow emigrants to break their agreements and negotiate passage with other shippers should not be interpreted as monopoly power over recruiting but rather as an effort to keep from losing the investment already spent on delivering their recruits to port. Hope & Co. experienced competitive recruitment problems. In 1756, for example, a group of 26 Germans contracted with the Hopes to be shipped from Rotterdam to Philadelphia for 7.5 doblons each. The contract also contained the following contingency: But if anyone agrees to take these Germans for less than the above-­ mentioned sum, Messrs. Isaac & Zacharias Hope promise to do the same, except where it is plainly done as spite work against Messrs. Isaac & Zacharias Hope, in which case they release the people from the contract, however in such case those who offer cheaper transportation are to pay Messrs. Isaac & Zacharias Hope for the expenses which they incurred before the people arrived in port.26 This contract suggests that competition was quite spirited and that bidding emig­ rants away from other recruiters was common. The Hopes contracted not to be undersold on the passage fare, which suggests that they had no monopoly power to raise prices. Second, they realized that other shippers could underbid them on the passage fare because these shippers would not have to pay recruitment costs. The Hopes thus tried to protect their investment by contractually requiring compensation for recruiting expenses if their emigrants deserted them for other merchants. This behavior is more characteristic of competitive than monopolized markets. Finally, the redemptioner trade must have represented only a small part of the Hopes’ business. Hope & Co.’s annual turnover at the Amsterdam Exchange Bank was over 900,000 pounds sterling by mid-­century. Even if they shipped all the Germans going to Philadelphia between 1730 and 1756, the potential gross revenue would represent less than 2.5 percent of the annual turnover. Their annual turnover at the Amsterdam Exchange Bank between 1730 and 1756 was also very stable, especially compared with other large Dutch merchant houses and with the volatility of German emigration. This suggests that changing fortunes in the German redemptioner trade, like the possible acquisition and loss of monopoly rights or interruptions in the flow of emigrants during upheavals like the War of the Austrian Succession, had little effect on Hope & Co. Therefore, it seems unlikely that Hope & Co. would have invested the considerable resources

The transatlantic shipping market   61 needed to acquire a monopoly position and prevent competition in the German redemptioner market.27

Quantitative evidence suggesting competition in the Pennsylvania market Pennsylvania was the overwhelming choice of German immigrants during the middle half of the eighteenth century. The available quantitative evidence on the market structure of the German passenger trade to Philadelphia is probably more extensive and complete than for any other large colonial migration.28 The evidence exhibits none of the characteristics associated with a monopolized market. High market concentration, restriction of services, above-­normal prices and profits, and difficulty in entering or exiting the market are not observed. Between 1727 and 1775 there were 317 German immigrant voyages employing 190 different captains (see Table 3.2). Seventy-­five percent of the captains Table 3.2 Role of ship captains in carrying German passengers to Pennsylvania Voyages undertaken

Number of captains

Total adult males carried

1727–75   13 (4.1%)   11 (3.5)    8 (2.5)    7 (6.6)    6 (7.6)    5 (3.2)    4 (2.5)    3 (12.3)    2 (12.6)    1 (45.1)

   1 (0.5%)    1 (0.5)    1 (0.5)    3 (1.6)    4 (2.1)    2 (1.1)    2 (1.1)   13 (6.8)   20 (10.5) 143 (75.3)

  1,277 (5.3%)    812 (3.4)   1,077 (4.5)   2,140 (8.9)   2,140 (8.9)   1,216 (5.1)    673 (2.8)   2,740 (11.4)   2,775 (11.6)   9,182 (38.2)

317 (100%)

190 (100%)

24,015 (100%)

1730–56    8 (7.3%)    7 (6.4)    6 (10.9)    5 (2.3)    4 (5.5)    3 (6.8)    2 (12.7)    1 (48.2)

   2 (1.5%)    2 (1.5)    4 (2.9)    1 (0.7)    3 (2.2)    5 (3.7)   14 (10.2) 106 (77.4)

  1,884 (10.0%)   1,568 (8.3)   2,123 (11.3)    405 (2.2)   1,393 (7.4)   1,303 (6.9)   2,352 (12.5)   7,792 (41.4)

220 (100%)

137 (100%)

18,820 (100%)

Source: Derived from Strassburger, German Pioneers Vol. 1. Only ships arriving from London or Holland were included. Notes Percentages of the respective totals are in parentheses. Only passengers who signed the loyalty oath in Philadelphia, required of adult male German immigrants, were used as the most direct and accurate measure of the relative number of passengers carried across shippers.

62   German immigration to America, 1709–1835 carried German immigrants only once. These one-­time participants in the trade accounted for 45 percent of the voyages and 38 percent of the immigrants. Only ten captains, 5 percent, made six or more voyages. They accounted for 24 percent of the voyages and 31 percent of the immigrants. The subsample from the alleged period of monopolization, 1730–56, is similar to the overall sample except that captains who participated only once are more important during the 1730–54 period. This pattern of small numbers of specialists along with large numbers of occasional shippers may have been common for colonial trade routes in general.29 Seventy-­eight ships in the German passenger trade could be traced in the Philadelphia port register, which provides information on vessel ownership. Only half of the 78 were owned by the 20 largest shippers in Philadelphia, those with interest in 11 or more vessels. Only half of the captains who made over five German immigrant voyages were employed by these large shippers. The pattern changed little between 1727 and 1775. Quite a diverse group of large and small, Philadelphia and non-­Philadelphia merchants, were engaged in shipping German immigrants to Philadelphia. In particular, small independent shippers entered and exited the trade at will. This low level of concentration in the shipping of German immigrants suggests that little effective monopoly power could have been exercised in the market during the mid-­eighteenth century.30 When the ships docked at Philadelphia the redemptioner servants were consigned to local merchants who collected the amounts due from their sale. Prepaying passengers did not have to be consigned. The consignment of German immigrant ships was recorded for 1751, 1753–6, and 1763–74 (see Table 3.3). The market structure of ship consignments was similar to that of ship captains. A large number of merchants took occasional consignments. A few merchants specialized in the trade. As a group Stedman, Shoemaker, Willing & Morris, Howell, and Beneset gained around 60 percent of the consignments. The degree of concentration in consignments was similar between the two periods 1751–6 and 1763–74, but with substantial changes in individual merchant shares. Therefore, compared with what happened after the War, it seems unlikely that monopoly power was exercised in the sale of German redemptioners in Philadelphia prior to the Seven Years War. Occasionally an individual merchant or two attained a substantial market share of consignments in a given year. The Stedmans and Shoemaker, for example, had 71 percent in 1751, Shoemaker had 95 percent in 1763, and Willing & Morris had 86 percent in 1771. There was, however, no consistent yearly pattern in market shares across merchants. Merchants presumably took consignment of the vessels they partially owned.31 But there was competition for consignments from other shippers. George Parish, for example, consigned the redemptioners aboard the Queen of Denmark to the Stedmans in 1751, but to Keplee in 1753. Charles Smith consigned the redemptioners aboard the Chance to Robert Ruecastle in 1765, but to Willing & Morris in 1766.32 It seems unlikely, therefore, that effective monopoly power could have been consistently exercised over the consignment and sale of German redemptioners in Philadelphia.

The transatlantic shipping market   63 Table 3.3 Consignment of German redemptioner cargos in Philadelphia, 1751–74 Philadelphia merchant

1751 and 1753–6

1763–74

Number of ships

Total adult males carried

Number of ships

Total adult males carried

Benjamin Shoemaker Charles Stedman Charles Willing Willing & Morris Samuel Howell Daniel Beneset Robert Ruecastle Keplee Keplee & Steinmetz Robert Ritchie Joshua Fisher John Pole Greenway & Rundel Rundel Able James James & Drinker James Pemberton James Searle James Christie Hillegas O’Kill John Ross Richard Neave Jr. Warder & Parker Warder & Sons Barclay Cunningham & Nesbit Gibs & West Mease & Callents John Jones Shervel & Salter Unconsigned

12 (21.1%) 16 (28.1)   1 (1.8)  0  0   7 (12.3)  0   6 (10.5)  0  0  0   1 (1.8)   1 (1.8)  0   2 (3.5)  0   1 (1.8)  0  0   1 (1.8)   1 (1.8)  0  0  0  0  0  0  0  0  0  0   8 (14.0)

1,535 (26.7%) 1,667 (29.0)    49 (0.9)     0     0   712 (12.4)     0   504 (8.8)     0     0     0   139 (2.4)    53 (0.9)     0    60 (1.0)     0    84 (1.5)     0     0    82 (1.4)    64 (1.1)     0     0     0     0     0     0     0     0     0     0   807 (14.0)

  8 (10.5%)  0  0 20 (26.3) 12 (15.8)  0   6 (7.9)  0   3 (4.0)   2 (2.6)   4 (5.3)  0  0   1 (1.3)  0   1 (1.3)   1 (1.3)   2 (2.6)   2 (2.6)  0  0   2 (2.6)   1 (1.3)   1 (1.3)   1 (1.3)   1 (1.3)   1 (1.3)   1 (1.3)   1 (1.3)   1 (1.3)   1 (1.3)   2 (2.6)

  680 (16.1%)     0     0 1,270 (30.0)   810 (19.2)     0   507 (12.0)     0    52 (1.2)   228 (5.4)   175 (4.1)     0     0    68 (1.6)     0    52 (1.2)     9 (0.2)    92 (2.2)    82 (1.9)     0     0    42 (1.0)    37 (0.9)    15 (0.4)     8 (0.2)    13 (0.3)    12 (0.3)    12 (0.3)     8 (0.2)     7 (0.2)     7 (0.2)    35 (0.8)

Totals

57 (100%)

5,754 (100%)

76 (100%)

4,227 (100%)

Source: Strassburger, German Pioneers Vol. 1. Notes Percentages of the respective totals are in parentheses. Only ships arriving from Holland or London were considered. The adult males carried are those whose names appear on the loyalty oath lists for each ship which represents the most direct and consistent measure of the relative number of passengers across shippers. The consignment was for collection of passage fares from passengers under redemptioner contract. The proportion of redemptioners to prepaying passengers was not available for these ships.

64   German immigration to America, 1709–1835 If the market for shipping Germans from Rotterdam to Philadelphia was monopolized prior to the Seven Years War, then the passage fare and shipping profits should have been higher before the War than after. There is little evidence on shipping profits in general. However, the rate of return to shipping redemptioners was estimated to be 11 to 15 percent above the return to prepaid passengers on voyages between 1802 and 1819.33 After subtracting compen­ sation for expected defaults and a return to uncertainty, this return was approximately equal to the return on other competitive, safe investments. In 1750 John Dick claimed it was customary to add 15 percent to the passage fare of redemptioners “as an Indemnity for the Charges and laying out of the money.”34 The similarity in the return to shipping redemptioners between these two periods suggests that above normal profits were not realized prior to the Seven Years War. Evidence on German passage fares exhibits a pattern opposite from that suggested by the monopolization argument. Fares were higher after the Seven Years War when shipping was less risky and supposedly more competitive (see Table 3.4). Passage fares from 1708 to 1756 were consistently between five and six pounds sterling during peace and seven to 10 pounds sterling during war.35 After the Seven Years War passage fares were consistently above ten pounds sterling. Effective monopolization of the market before 1755 seems unlikely, given evidence on passage fares and profits. Finally, if the German passenger market was effectively monopolized, then the quantity of services would have been restricted to force passage fares up, lower the cost of underutilized freight space, and thus increase profits. The restriction of the number of vessels available to transport immigrants would lead to a higher number of passengers per ship. Transatlantic shipping was characterized by underutilized cargo space on the return voyage from Europe to America. Immigrants represented important freight on the westward voyage throughout the eighteenth and early nineteenth centuries.36 A merchant could eliminate some of his underutilized freight space by gaining a monopoly over the passenger trade. Table 3.5 presents the average number of immigrants per ship in the German trade for five year intervals. From 1727 to 1749 this average stayed fairly constant at about 185. The surge in German emigration between the War of the Austrian Succession and the Seven Years War raised the average number of passengers per ship to about 300. The revival of emigration after the Seven Years War brought an average number of passengers per ship only slightly less than in the 1727–49 period.37 This suggests that there was substantial excess capacity in German immigrant shipping before 1750 and after 1755. The increase in passengers per ship in the early 1750s was probably the result of unexpected increases in emigration and not monopolization. This unexpected demand attracted more merchants and ships into the German passenger market. The evidence suggests that there was no systematic restriction in the quantity of German passenger services, especially prior to 1749. Therefore, it seems unlikely that the market was effectively monopolized.

The transatlantic shipping market   65 Table 3.4 Transatlantic passage fares for adult male Germans, Holland to Philadelphia, 1708–1819 Source

Year

Passage fare in pounds sterling

Kocherthala

1708

Samuel Guldinb Unknown Immigrantc Andrew Bonid Hans Trachslere Muhlenbergf

1710 1728 1736 1736 1742

Mittelbergerg Ship Nancyh Hope & Co. contracti Ship King of Prussiah Muhlenbergj

1750 1750 1756 1764 pre-1769 after-1769 1773 1802 1803 1803 1819

  5–6   7–8  7  5  5   5.1 10   8.4 10   5.9   6.2   8.6   6.1–10.2 14.3–17.3   9.9 14.4 13.65 16.45 18

Ship Britanniak Ship Belviderel Ship Commercel Ship Pennsylvanial Ship Elizabethm

(in peace time) (in war time) (war time) (servant rate) (Holland to Carolina) (normally, war time) (on his ship, London to Georgia) (servant rate) (servant rate, from London) (servant rate) (servant rate)

Sources a Joshua Kocherthal, Full and Circumstantial Report Concerning the Renowned District of Carolina in English America, quoted in Henry E. Jacobs, “The German Emigration to America 1709–1740,” Pennsylvania German Society 8 (1897): 39. b William J. Hinke, “Diary of the Rev. Samuel Guldin, Relating to his Journey to Pennsylvania, June to September, 1710,” Journal of the Presbyterian Historical Society 14 (1930): 71. c Julius F. Sachse, “A Missive From Pennsylvania in the Year of Grace 1728,” Pennsylvania German Society 28 (1909): 18. d Durnbaugh, Brethren in Colonial America: 39. e Voigt, “Swiss Notes,” 98. f Henry M. Muhlenberg, The Journals of Henry Melchior Muhlenberg, Theodore G. Tappert and John W. Doberstein, trans., (Philadelphia, 1942) 1: 23. g Gottlieb Mittelberger, Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754, Oscar Handlin and John Clive, trans. and ed., (Cambridge, MA, 1960): 17. h “Redemptioners, Philadelphia, 1750–1830,” (unpublished manuscript held at the Historical Society of Pennsylvania, Miscellaneous Collection, Box 7a, Folder 7). i Langguth, “Pioneers from the County of Wertheim,” 260–1. j Henry M. Muhlenger, Hallesche Nachrichten, new ed., 2: 460–1, quoted in Strassburger, German Pioneers 1: xxxvii. k “Passenger Account of Ship Britannia (mustering book), Capt. James Peter (which arrived in Philadelphia, 18 September 1773),” (unpublished manuscript held at the Historical Society of Pennsylvania). l Strassburger, German Pioneers 3: 112–14, 131–4, and 137–8. m “Passenger List of the Ship “Elizabeth,” Which Arrived at Philadelphia in 1819,” Pennsylvania Magazine of History and Biography 25 (1901): 255–8. Notes All fares are for free, prepaying immigrants in peace time, except where indicated. Appropriate currency conversions were made from McCusker, Money and Exchange.

66   German immigration to America, 1709–1835 Table 3.5 German immigrants per ship, Holland to Philadelphia, 1727–74 Year

Average immigrants per ship

Average ships per year

1727–9 1730–4 1735–9 1740–4 1745–9 1750–4

173.76 (57.75) 189.45 (24.19) 188.59 (59.85) 186.44 (12.26) 180.48 (106.29) 300.80 (17.72)

  3.3 (1.53)   5.4 (3.65)   7.4 (5.32)   7.0 (2.35)   8.2 (9.04) 18.0 (2.54)

Seven Years War 1763–9 1770–4

170.90 (63.83)   97.67 (18.34)

  6.0 (2.45)   9.0 (3.54)

Sources: Strassburger, German Pioneers, Vol. 1; and Wokeck, “Flow and Composition,” 260–1. Note Standard deviations on a yearly basis are in parentheses. Where possible the evidence is presented in bi-decade intervals.

The market structure of German immigrant shipping from Rotterdam to Philadelphia during the middle half of the eighteenth century was more consistent with competitive behavior than monopolization. It was similar to other transatlantic routes. A few merchants specialized in various aspects of the trade— recruiting, shipping, financing, and redemptioner selling—but only for comparatively short periods. There was always a large number of merchants on the fringe who entered and exited the trade at will. Evidence on profits, passage fares, and shipping services was more consistent with competition than monopolization. Recruiters from many colonies competed for German emigrants in Europe, and contract evidence from Hope & Co. suggests that there was competition among recruiters from Pennsylvania. The evidence in favor of monopolization—limited essentially to three quotations from different contemporary observers—is not consistent as to time period, monopolizing circumstances, or merchant monopolist. This literary evidence could just as easily be interpreted as complaints about competitive rivalry and unlucky voyage conditions than true monopolization. At best, Hope & Co. may have achieved some temporary and limited monopoly over financing the emigration of unsponsored, impoverished Germans as they entered Holland around 1736. The passport restriction which legally created the monopoly was temporary and did not prevent competition by other merchants. It may have influenced the sharp downturn in German immigration between 1734 and 1737 more than restricted competition so that Hope & Co. could charge excessive fees.38 Finally, the diaries of German immigrants to Pennsylvania never mention or complain about shipping or recruiting monopolies; nor do the diaries report difficulties in obtaining passports to enter Holland. The diaries do complain about the deceptive practices of competing recruiters.39 The consistency of the evidence with competition in all aspects of the trade shifts the burden of proof to those who want to maintain that the market was monopolized.

The transatlantic shipping market   67

Notes   * Originally published as Farley Grubb, “The market structure of shipping German immigrants to colonial America,” Pennsylvania Magazine of History and Biography, vol. 111, no. 3 (January 1987): 27–48. Copyright 1987. © 1987 by the Historical Society of Pennsylvania. Reprinted with permission. The author wishes to thank David Galenson and Jimmie Lee for helpful comments on an earlier draft.   1 See Simeon J. Crowther, “The Shipbuilding Output of the Delaware Valley, 1722–1776,” Proceedings of the American Philosophical Society 117 (1973): 90–104; Ralph Davis, The Rise of the English Shipping Industry (London, 1962); Thomas M. Doerflinger, “Commercial Specialization in Philadelphia’s Merchant Community, 1750–1791,” Business History Review 57 (1983): 20–49; Marc Egnal, “The Changing Structure of Philadelphia’s Trade with the British West Indies, 1750–1775,” Pennsylvania Magazine of History and Biography 99 (1975): 156–79; Farley Grubb, “The Market for Indentured Immigrants: Evidence on the Efficiency of Forward-­Labor Contracting in Philadelphia, 1745–1773,” Journal of Economic History 45 (1985): 855–68; W. E. Minchinton, ed., The Growth of English Overseas Trade in the Seventeenth and Eighteenth Centuries (London, 1969); James F. Shepherd and Gary M. Walton, Shipping, Maritime Trade, and the Economic Development of Colonial North America (Cambridge, 1972); Gary M. Walton, “New Evidence on Colonial Commerce,” Journal of Economic History 28 (1968): 363–89.   2 For example, see Frank R. Diffenderffer, “The German Immigration into Pennsyl­ vania Through the Port of Philadelphia, and ‘The Redemptioners’,” pt. 7 of “The German Influence in its Settlement and Development,” Pennsylvania German Society 10 (1899), 239–42; Karl F. Geiser, Redemptioners and Indentured Servants in the Colony and Commonwealth of Pennsylvania (New Haven, 1901), 47–8; Abbot E. Smith, “Some New Facts About Eighteenth-­Century German Immigration,” Pennsylvania History 10 (1943): 105–17; Abbot E. Smith, Colonists in Bondage (New York, 1947), 59, 208–18; Marianne S. Wokeck, “A Tide of Alien Tongues: The Flow and Ebb of German Immigration to Pennsylvania, 1683–1776,” Ph.D. dissertation, Temple University, 1983, 137–202.   3 The vast majority of Germans who migrated to the New World in the eighteenth century came between 1730 and the Seven Years War. See Marianne S. Wokeck, “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography 105 (1981), 260–1.   4 See Diffenderffer, “German Immigration into Pennsylvania,” 55–68, 147, 256–62; John Duffy, “The Passage to the Colonies,” Mississippi Valley Historical Review 28 (1951): 21–38; David W. Galenson, White Servitude in Colonial America (Cambridge, MA, 1981), 13–15; Geiser, Redemptioners and Indentured Servants, 26–7, 43–58; Richard B. Morris, Government and Labor in Early America (New York, 1946), 319–22; Billy G. Smith, “Death and Life in a Colonial City: A Demographic Analysis of Philadelphia,” Journal of Economic History 36 (1977), 872; Smith, “Eighteenth-­Century German Immigration,” 105–17; Smith, Bondage, 20–2, 35–41; Colonial Records of the State of Georgia 29, 86, 96; 21, 418–19. For the incidence of European immigrant servitude in Philadelphia see Farley Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History 22 (1985): 316–39.   5 Farley Grubb, “Morbidity and Mortality on the North Atlantic Passage: Eighteenth-­ Century German Immigration to Pennsylvania,” Journal of Interdisciplinary History 17 (1987), forthcoming; Farley Grubb, “Redemptioner Immigration to Pennsylvania, Evidence on Contract Choice and Profitability,” Journal of Economic History 46 (1986): 407–18. See also Colonial Records of Georgia 21, 418–19; Donald F. Durnbaugh, “Two Letters from Germantown,” Pennsylvania Magazine of History and  Biography 84 (1959), 231–3; Galenson, White Servitude, 14; Farley Grubb, “Immigrant Servant

68   German immigration to America, 1709–1835 Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­Atlantic Economy,” Social Science History 9 (1985): 249–75.   6 “Christopher Sauer’s First Letter to Governor Morris on the Trials and Wrongs of the Early German Immigrants,” March 15, 1755, quoted in Diffenderffer, “German Immigration into Pennsylvania,” 239–45.   7 Wokeck, “Flow and Composition,” 260.   8 The Stedmans registered the Charming Nancy in 1736, the Lydia in 1738, and may have controlled the St. Andrew and two ships named Thistle though they did not register them in Philadelphia until a decade later. With at best five ships at their command they would have had to carry 400 passengers per ship to attain 2,000 immigrants per year. The implied passengers per ship ratio, although not impossible, would have been at the extreme upper end of the observed distribution for any period. See Table 3.5 and Strassburger, German Pioneers vol. 1. Wokeck has estimated total passengers to be 2.5 to 2.8 times the number of adult men. Wokeck, “Flow and Composition,” 260. Therefore, the Stedmans’ best year was 1754 when they carried around 1,400 emigrants or roughly 280 per ship. Their best year in the 1730s was 1738 when they carried around 800 emigrants or roughly 200 per ship.   9 Charles Stedman was part owner of 11 vessels registered in Pennsylvania, a total investment of 316 tons with a mean tonnage per investment of 29. See Crowther, “Ship-­building Output,” 103. 10 Colonial Records of Georgia 21, 418–23, 437–8; 29, 200–1. The Stedmans’ vessels were not the first to arrive in Philadelphia during the 1738 fall migration season. See Strassburger, German Pioneers vol. 1. Their ships must have loaded the emigrants who arrived late in the season rather than the first arrivals in port, or they had to stay in Rotterdam for a longer period than other shippers to complete their emigrant load. 11 Donald F. Durnbaugh, ed., The Brethren in Colonial America (Elgin, IL, 1967), 42–53: Strassburger, German Pioneers 1, xxxiii; Waldmar Westergaard, “Two Germantown Letters of 1738,” Pennsylvania Magazine of History and Biography 56 (1932): 9–14. 12 Smith, “Eighteenth-­Century German Immigration,” 106; Smith, Bondage, 208. 13 Colonial Records of Georgia 30, 22–3. 14 Colonial Records of Georgia 22, pt. 2, 106–7; 30, 42–3. Efforts to regulate German emigration through Holland to lessen the burden on Dutch charity were not unprecedented, see the discussion of the 1709–11 migration in Walter A. Knittle, Early Eighteenth Century Palatine Emigration (Philadelphia, 1937), 47–65; Frank R. Diffenderffer, “German Exodus to England in 1709,” pt. 1 of “The German Influence in its Settlement and Development,” Pennsylvania German Society 7 (1897): 264–75. In later periods the Dutch resorted to the same procedures when similar problems reappeared, see the discussion of the 1816–17 German migration in Marcus L. Hansen, The Atlantic Migration, 1607–1860 (Cambridge, MA, 1940), 87–9. 15 Colonial Records of Georgia 20, 114–15, 328–34, 345; 21, 100–2, 406–13, 473–8; 29, 35–6, 41–2, 136. The colony of Georgia may have created some of these unrealistic expectations by transporting small groups of German immigrants, mostly Salzburgers, at the colony’s expense, and by allowing some unauthorized recruiting to occur on its behalf. Groups of Swiss and German emigrants continually petitioned the Georgia trustees to be transported under the same conditions as the Salzburgers. 16 Colonial Records of Georgia 21, 100–2. Over 200 Swiss were detained by the Magistrates of Rotterdam in September, 1736, but the detention involved an unsuccessful effort to have them use a Dutch vessel rather than an English vessel. This was not a passport violation. Gilbert P. Voigt, “Swiss Notes on South Carolina,” South Carolina Historical and Genealogical Magazine 21 (1920), 93–4. 17 Colonial Records of Georgia 4, 54–5; 20, 287; 21, 433–4; 22, pt. 1, 86, 107–8, 250, 297, 321, 342–50; 29, 74, 77, 89, 95–104, 136, 200–1, 229–31, 242–5, 293–5; Durnbaugh, Brethren in Colonial America, 39–41; Voigt, “Swiss Notes,” 93–4, 100.

The transatlantic shipping market   69 18 Colonial Records of Georgia 21, 418–23, 437–8; 22, pt. 1, 107–8, 297, 321; 29, 95–104, 135, 200–1, 229–31, 242–5, 293–5. 19 Smith, “Eighteenth Century German Immigration,” 107; Smith, Bondage, 55–9, 208–9. See also Bernard Bailyn, The Peopling of British North America (New York, 1986), 70–2; Knittle, Palatine Emigration, 217; Morris, Government and Labor, 401. 20 Colonial Records of Georgia 21, 418–23, 437–8. 21 See Wokeck, “Flow and Composition,” 249–78; Historical Statistics of the United States From Colonial Times to 1970 (Washington, DC, 1975), 1, 168; Diffenderffer, “German Immigration into Pennsylvania,” 97–106; Knittle, Palatine Emigration; Peter Kalm, Travels in North America, Adolph B. Benson, trans. and ed. (New York, 1937), 1, 142–3; fn. 5; fn. 6. 22 For discussions of redemptioner servitude and servant recuriting, see Durnbaugh, “Two Letters,” 231–3; Diffenderffer, “German Immigration into Pennsylvania,” 141–315; Geiser, Redemptioners and Indentured Servants; Galenson, White Servitude, 13–15; Grubb, “Redemptioner Immigration,” 407–18; Smith, Bondage, 20–66; Colonial Records of Georgia 21, 418–23, 437–8; 22, pt. 1, 321; 29, 89, 96, 229–31, 242–5. 23 Alexander Hope commanded the Queen Elizabeth to Philadelphia with 324 German emigrants in the fall of 1738. Strassburger, German Pioneers 1, 216–21. For the history of the Hopes see Marten G. Buist, At Spes Non Fracta, Hope & Co. 1770–1815 (The Hague, 1974). 24 Stuart Bruchy, The Colonial Merchant (New York, 1966), 182–3; Buist, Hope & Co., 6–22; Wokeck, “Tide of Alien Tongues,” 141–4. 25 There was little enforcement of Dutch regulations on loading and shipping passengers. Geiser, Redemptioners and Indentured Servants, 47–8; Voigt, “Swiss Notes,” 100; Colonial Records of Georgia 21, 418. 26 Otto Langguth, “Pennsylvania German Pioneers from the County of Wertheim,” Pennsylvania German Folklore Society 12 (1947), 260–1. 27 Buist, Hope & Co., 6–8, 476–85. The turnover represented the transfer of sums, credits and debits, to and from other bank accounts and so represented a low estimate of total business activity. Appropriate currency conversions were taken from John J. McCusker, Money and Exchange in Europe and America, 1600–1775 (Chapel Hill, 1978). Expected annual German migration was taken from Wokeck, “Flow and Composition,” 260. Ten pounds sterling was used as a high estimate of the gross revenue from shipping each emigrant. Thus, the 2.5 percent estimate is probably high. Hope & Co. did not undergo dramatic growth until after 1756 when they also began to phase out their emigrant business. 28 See the assessment in Smith, Bondage, 320. 29 See Doerflinger, “Commercial Specialization in Philadelphia’s Merchant Community,” 20–49; Egnal, “Changing Structure of Philadelphia’s Trade,” 156–79; Walton, “Colonial Commerce,” 363–89. 30 Derived from Strassburger, German Pioneers, vol. 1; “Ship Registers for the Port of Philadelphia, 1726–1775.” A match between a listing in the port register and a German immigrant vessel listed in Strassburger was considered only when both the name of the ship and the captain or owner’s names were the same in both sources. Duplicate ship names were common in the register. The concentration of ship owning was taken from Crowther, “Shipbuilding Output,” 103. 31 There were exceptions. James Abercrombie commanded the Peggy for the Stedmans and consigned his redemptioners to the Stedmans in 1753, but changed to the Stedmans’ competitor, Daniel Beneset, in 1754. Strassburger, German Pioneers 1, 545–50, 636–42. 32 Strassburger, German Pioneers 1, 472, 516–21, 705, 708–9. 33 Grubb, “Redemptioner Immigration,” 411–17. 34 Smith, Bondage, 40.

70   German immigration to America, 1709–1835 35 A notable exception was the ten-­pounds sterling fare reported by Mittelberger during peace in 1750. However, Mittelberger’s journal was an open attempt to discourage emigration and so may have been prone to exaggeration. See Grubb, “Morbidity and Mortality.” 36 See Davis, Rise of English Shipping; Hansen, Atlantic Migration; Shepherd and Walton, Shipping, Maritime Trade, and the Economic Development. 37 There is very little evidence on passengers per ton, which would be better evidence with which to compare changing capacity utilization. Grubb, “Redemptioner Immigration,” 416–17. Because the tonnage of ships in the trade changed little and because the number of passengers on repeating ships in the trade fluctuated considerably, the aggregate number of passengers per ship over five-­year intervals should provide a reasonable approximation of the true trend. 38 See Wokeck, “Flow and Composition,” 260. Emigrants may have postponed their journey until they had made guaranteed arrangements or until the passport restriction was clarified. 39 For denunciations of the infamous recruiters called ‘Newlanders,’ see Diffenderffer, “German Immigration into Pennsylvania,” 188–93; Geiser, Redemptioners and Indentured Servants, 18–21; Mittelberger, Journey, 26–30; Sachse, “A Missive From Pennsylvania,” 23–4; Westergaard, “Germantown Letters,” 12.

4 Morbidity and mortality on the North Atlantic Passage*

The voyage from Europe to America in the eighteenth century has been portrayed as a harrowing experience. The historical literature has been full of  examples of traumatic crossings where starvation, thirst, sickness from a host  of diseases, cramped quarters, heat, cold, dampness, contrary winds, raging  seas, unpalatable food, and foul water led to much suffering and death. German migrants apparently suffered the worst of these conditions. In some  cases the causes of this misery were attributed to the cruelty or indifference of ship owners and their captains. Overall, European voyage mortality has  been considered comparable to what prevailed in the slave trade. However,  little direct quantitative evidence has been offered of mortality on representative North Atlantic voyages to support this view. In the case of voyage morbidity and post-­voyage mortality, no estimates based on direct, quantitative evidence exist.1 In order to provide a better picture of migration conditions, estimates of voyage mortality, debarkation morbidity, and immediate post-­voyage mortality for eighteenth-­century German immigrants arriving in Philadelphia are presented in this study. These estimates were derived from quantitative evidence in surviving passenger manifests and from the bills of mortality for the city of Philadelphia. Derived from relatively representative and direct evidence, these estimates provide an improvement over previous estimates. In addition, since Germans suffered some of the worst voyage conditions, these estimates may provide an upper bound for what was experienced by other immigrants.2 Passage conditions, particularly morbidity and mortality, were important costs to prospective emigrants. Many colonists sent return letters to Europe describing their journey. Reports of unexpected hardship on the voyage discouraged further migration. For example, Charles Hector and his neighbors changed their plans to migrate from Germany to America when letters arrived from friends who had left the previous year. He wrote in 1726: As we were ready to leave, there came letters from Pennsylvania, which reported that all of the poor people, whom we knew, had suffered such misery on their trip, as hunger, thirst, and scarcity of all help had cost the lives of the majority of them on the ship.3

72   German immigration to America, 1709–1835 These costs posed a formidable restraint on European emigration to America. As a consequence, this threat may have altered the growth path of world output, and the growth in the personal wealth of prospective emigrants, by postponing the movement of labor from low productivity employment in Europe to high productivity employment in America.4 The traumas of the crossing also altered the anticipated opportunities that immigrants experienced upon arrival. A difficult voyage left them in a weak condition with lingering voyage illnesses and susceptible to colonial diseases. These effects may have lowered the anticipated value of their labor. Within the servant trade this was a particular problem because servants were sold soon after suffering the hardships of the voyage. As would be expected, the ill were difficult to sell. In 1750, Mittelberger related the fate of German servants who arrived ill in Philadelphia: In the whole process the sick are the worst off for the healthy are preferred and are more readily paid for. The miserable people who are ill must often remain at sea and in sight of the city for another two to three weeks—which in many cases means death.5

Voyage mortality The single most important concern to immigrants and the focus of historical research on voyage conditions has been voyage mortality. However, direct evidence on mortality from passenger ship records for eighteenth-­century North Atlantic voyages has been extremely scarce. The historical literature has relied on three general sources: a few immigrant diaries, isolated voyage disasters, and three sample migrations—2,814 Palatines transported by the British Crown to New York in 1709, English convicts transported to Maryland between 1718 and 1775, and 8,400 British troops sent to the West Indies between 1775 and 1782. These three samples yielded a voyage mortality of 10 to 16 percent, which has been used to assert a 10 percent mortality rate as the lower bound for the eighteenth-­century North Atlantic passage. The voyage descriptions of Mittleberger, who claimed that children seldom survived, and the voyages of ships such as the Love and Unity, shipwrecked in 1732 with the loss of 70 percent of its 156 passengers, have been used to assert that voyage conditions for German immigrants were more severe than for Irish or English immigrants.6 Evidence from large samples has suggested that passage mortality among slaves averaged somewhere between 9 and 15 percent in the eighteenth century. Since average voyage time was roughly similar, about two months, some researchers have concluded that white passage mortality on the North Atlantic was comparable if not worse than in the slave trade. For example, Smith concluded: German redemptioners, indentured servants, and poor immigrants, who collectively comprised the bulk of Philadelphia’s new arrivals, frequently

Morbidity and mortality   73 experienced appalling voyages across the Atlantic, at times equalling even the horrors of the African middle passage.7 It has been suggested that white passengers received worse treatment than slaves because shippers had an economic interest in the survival of slaves. They were saleable property whereas free immigrants, having paid their passage, presented no such economic interest.8 These claims may not be justified, however, because the evidence cited was biased in favor of high mortality and was not representative of the typical immigrant experience. Convicts and troops received harsher treatment both before and during the voyage and so may have experienced a relatively higher mortality. The samples of British troops and Palatines did not involve the private shipping sector which transported most of the immigrants. Due to competitive pressures, the private sector may have been more sensitive to mortality-­ producing conditions than ventures run by the government. The sample of British troops also involved migration into the West Indies, a relatively virulent disease environment for Europeans, which may have contributed to mortality above what was normal for North Atlantic destinations. In addition, the Palatine sample was not based on direct evidence but on broad and conflicting reports by contemporary observers who included pre- and post-­voyage mortality from outbreaks of pestilence in their estimates. Finally, with regard to the evidence from the German trade, the frequency of non-­arriving voyage disasters was small: 1.5 percent between 1727 and 1775. Mittelberger’s voyage sailed under atypically harsh circumstances, which colored his views.9 In an effort to derive a less biased estimate of voyage mortality, 14 German immigrant vessels, which enumerated passenger deaths directly in the ship records, were taken from the Strassburger collection of German ship lists for the port of Philadelphia. This sample had over 1,566 passengers, covered the years 1727 to 1805, and appeared to be relatively representative of the typical immig­ rant voyage: voluntary, white, civilian immigrants transported by the private shipping market on the North Atlantic route. All 14 reported on adult men. But only six had mortality enumerated for the separate categories of adult men, adult women, and children. Three ships listed only total immigrants and adult men. The overall passage mortality for these 1,566 Germans was 3.8 percent, an average of about 70 percent below slave passage mortality (see Table 4.1). The voyage mortality for the 1,153 adult men was slightly above that for the 237 adult women, 3.5 versus 2.5 percent, respectively, although this difference was not statistically significant. The 382 children fared far worse with a passage mortality of over 9 percent, or almost three times the adult rate. Only the sample of adult men had enough observations in the two periods of evidence, 1727 to 1754 and 1785 to 1805, to allow for a trend estimate. The estimates were not statistically different between the two periods, suggesting no trend in voyage mortality over the second half of the century.10 The point in the voyage where these deaths occurred and most of the causes of death were unknown. A few passengers were listed as drowned and three

74   German immigration to America, 1709–1835 Table 4.1 Passage mortality: German immigration to Pennsylvania, 1727–1805

Total passengers Adult mena Adult womenb Children

Ships

Observations

Passage mortality

 9 14  6  6

1,566 1,153 237 382

3.83% 3.56 2.53 9.16c

(3.79) (4.14) (4.08) (7.17)

Source: Derived from Strassburger, Pioneers, I, III. Notes Standard errors by ship are in parentheses in the final column. a Adult men are those who signed the loyalty oath, i.e., those above age 15. b The age for specifying adult status for females was not given except in a few cases where those above age 15 were listed as adults. c This mean is statistically different from all other means at the 0.001 level, based on a one-tailed t-test. All other means are not statistically different from one another.

children died of smallpox. The time spent at sea for these ships was also unknown. The average voyage from Holland to Philadelphia in this period was between two and two and a half months, which would suggest a monthly mortality of at least 15 per 1,000 boarded, or an annualized crude death rate of at least 184 per 1,000 for all immigrants and 440 per 1,000 for children. Although these estimates were far lower than previous estimates, they still indicated that trans­ atlantic travel was far more deadly than living ashore and over five times more deadly than living in Philadelphia (see Table 4.3). There were still tragic outcomes for many immigrants. For example, on the ship Fortune, arriving in Phil­ adelphia in 1804, death affected 38 percent of the family groups on board.11 By the 1830s to 1860s, North Atlantic passage mortality had fallen to between 2.4 and 1.0 percent, or as low as 10 per 1,000 per month. Since these voyages lasted around one to one and a half months, the annualized crude death rate was as low as 80 to 120 per 1,000. Thus late eighteenth-­century passage mortality was only about twice as high as early nineteenth-­century passage mortality.12 There were sound economic reasons why European voyage mortality was as low as suggested in Table 4.1 and lower than during the slave trade. First, a substantial portion of eighteenth-­century European immigrants were servants. The death of a servant, like the death of a slave, meant the total loss of the investment. Shippers realized this problem and took precautions like those observed by Pastor Henry Muhlenberg in 1764: The more human freight the ship captain can crowd into a ship the more profitable it is for them, if they do not die on the way, otherwise they may lose by it. For that reason the ships are kept clean and all kinds of precautions are taken to keep the passengers in good health, and to bring them to market in good condition. Second, the death of a free migrant, or servant, imposed a higher cost on that person than a slave’s death would on the slave’s owner, ceteris paribus, since the

Morbidity and mortality   75 Table 4.2 Debarkation morbidity for men above age 15, German immigration to Pennsylvania, 1727–54 Year

Ships

Observations

Sick

Debarkation morbidity rate

1727 1728 1729 1730 1732 1733 1734 1735 1736 1737 1738 1739 1741 1743 1748 1749 1750 1751 1752 1753 1754

4 2 1 1 6 7 1 1 1 3 3 3 1 1 2 11 2 4 9 5 11

273 72 64 77 491 435 89 59 117 335 202 192 15 82 198 1,208 114 486 1,091 498 1,094

24 2 4 1 36 13 6 5 5 10 3 5 1 9 4 38 2 12 28 16 27

8.79% 2.78 6.25 1.30 7.33 2.99 6.74 8.47 4.27 2.99 1.49 2.60 6.67 10.98 2.02 3.15 1.75 2.47 2.57 3.21 2.47

Totals

79

7,192

251

3.49

Source: Derived from Strassburger, Pioneers, I.

free migrant would lose the non-­pecuniary value of his life apart from his earnings. Since free migrants, even servants, voluntarily contracted for shipping services, this higher cost would have made them more discriminating in choosing their voyage conditions relative to how slave owners determined voyage conditions for their slaves.13

Debarkation morbidity Morbidity on the eighteenth-­century North Atlantic passage has been regarded as pervasive. It has been virtually impossible to find an immigrant diary where illness during the voyage was not a constant topic of discussion. Diseases brought by immigrants have been blamed for epidemics and high mortality in the colonial cities which received them. By far the most prevalent problem was seasickness, although several serious diseases, commonly called “fever” and “bloody flux,” also plagued the immigrants. These diseases were probably either typhoid, typhus, or dysentery, but, since their observed symptoms were relatively indistinguishable, they were commonly confused with one another in the eighteenth century. Their importance among immigrants has been attributed to their endemic status in Europe and to the conducive conditions produced by

1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752

Year

512 326 270 692 380 421 381 391 638 681 625 704 651 843 626

269 97 80 300 98 150 100 – 165 172 135 – 250 319 286

(1) (2) Estimated total Burials in the burials in the city city’s “strangers’ ground” 145 18 15 132 6 48 8 – 10 4 0 – 92 115 134

(3) Estimated first year German immigrant burials in the city 3,314 1,463 1,103 1,742 891 1,715 1,027 54 439 934 1,585 7,580 4,693 5,103 6,129

(4) Estimated total German immigration through the city 105.1 29.5 32.6 181.9 16.1 67.2 18.7 – 54.7 10.3 0.0 – 47.0 54.0 52.6

(5) Estimated first year immigrant annual crude death rate per 1,000

Table 4.3 Estimates of first year German immigrant versus resident mortality: Philadelphia, 1738–62

  9,631   9,874 10,117 10,360 10,755 11,150 11,545 11,940 12,336 12,731 13,126 13,521 13,926 14,330 14,812

(6) Estimated city population

38.1 31.2 25.2 54.1 34.8 33.5 32.3  – 50.9 53.2 47.6  – 40.1 50.8 33.2

(7) Estimated established resident annual crude death rate per 1,000

579 796 467 984 664 712 1,299 890 725 1,073

224 446 110 237 199 193 311 234 164 190

84 253 0 0 0 0 0 0 0 0

5,220 5,806 324 138 0 0 0 0 76 0 61.4

38.6 104.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

15,295 15,901 16,508 17,114 17,485 17,856 18,227 18,598 19,425 20,251

37.0

32.4 34.1 28.3 57.5 38.0 39.9 71.3 47.9 37.3 53.0

Sources: Column (1) was taken from B. G. Smith, “Death and Life,” 871; Column (2) was taken from Labaree, Benjamin Franklin, III, 439, and from the bills of mortality on microcard as listed in Charles Evans, American Bibliography: A Chronological Dictionary of All Books, Pamphlets, and Periodical Publications Printed in the United States, 1639–1830 (Chicago, 1903–1959); Column (3) was derived as follows: ((2)/(1)–0.2425)*(1). The number 0.2425 is the average ratio of burials in the Strangers’ Ground to all burials for the years 1757 to 1762 when German immigration did not occur. This number (0.2425) represents the proportion of burials in the Strangers’ Ground which was due to resident or non-German immigrant burials. Where the proportion of burials in the Strangers’ Ground was below 0.2425 for a given year, a zero was recorded in column (3); Column (4) was taken from Wokeck, “German Immigration,” 260–1; Column (5) was derived as follows: ((3)/(4) * 1,000) * 2.4. Since German immigration occurred between August and November exclusively, arriving German immigrants were only at risk for a maximum of five months. Thus to derive a lower bounded annual crude death rate entailed multiplication by a factor of 2.4; Column (7) was derived as follows: ((1)–(3))/(6) * 1,000.

1738–56 Subtotal

1753 1754 1755 1756 1757 1758 1759 1760 1761 1762

78   German immigration to America, 1709–1835 crowded, poorly sanitized ships. Some smallpox was also present, although it generally caused death only among small children.14 The overall frequency and extent of illness among eighteenth-­century immigrants has never been estimated from direct quantitative evidence, and even a reasonable guess has seldom been offered. However, for German immigrants, quantitative evidence was gathered on morbidity at the end of the voyage for 79 ships registered in Philadelphia between 1727 and 1754. Ship captains recorded “sick” next to the names of adult men who were incapable of travelling from the ship to the courthouse for the legally required signing of the loyalty oath immediately after landing. Of the 7,192 adult men in the sample, 251 were listed as sick for a debarkation morbidity of 3.5 percent (see Table 4.2). The morbidity was measured as a discrete variable and so was difficult to interpret. It may be best characterized as the rate of adult males bedridden when the ship arrived in port.15 What illnesses caused this morbidity remained unknown, but the afflictions must have been serious enough to have caused these men to be bedridden. This evidence captures most of the serious diseases potentially contracted on the voyage. Typhus, typhoid, and smallpox had short incubation periods of seven to 14 days. In most cases symptoms appeared in only a few days. Therefore, immigrants who were infected by these diseases before the last leg of the voyage would have been identifiable by debarkation. And the severity of the fever and pains would most likely have kept those infected in bed. Dysentery, however, had a highly variable incubation period, not invading the bowel and causing ulcerations until one to six months after infection. Therefore, dysentery experienced during the voyage was probably contracted prior to embarkation, whereas dysentery contracted during the voyage was probably not observed until after debarkation.16 Other variables on a per ship basis that were related to debarkation morbidity were the year of arrival, the number of adult men on board, and the illiteracy rate as measured by the ability to sign a required oath. A regression relating these variables to the debarkation morbidity rate was performed for the seventy-­nine ships in the sample: = 181.765* (93.61) – 0.028 (Number of Men per Ship)* (0.015) –0.100 (Year of Arrival)* (0.054) – 0.037 (Illiteracy Rate) (0.041) F = 4.14*, R2 = 0.14, N = 79, * = significant above 0.1. (Standard errors are in parentheses. Method of estimation was ordinary least squares.) debarkation morbidity rate per ship

Evaluated at the means, this estimate yielded an average debarkation morbidity of 5.23 percent. The illiteracy rate was insignificant. The trend in morbidity over the 27 years indicated a significant fall, over 38 percent. This trend can also be observed in Table 4.2. Serious voyage illness was on the decline.17

Morbidity and mortality   79 Controlling for the year of arrival, the number of men per ship had a significantly negative impact on debarkation morbidity. For ships one standard devi­ ation above the average number of men, compared to ships one standard deviation below, the fall in debarkation morbidity was 38 percent. This relationship is surprising since crowding has generally been regarded as one of the important causes of shipboard illness, particularly with respect to infectious diseases. Although the number of men per ship was not an exact measure of space per immigrant, it was positively associated with it. After 1749, the surge in German immigration brought a clear increase in crowding, from 180 to around 300 passengers per ship on average. Yet it was in these more crowded vessels that debarkation morbidity was the lowest. One explanation for this effect is that shippers who specialized in immigrant cargo may have taken more care in provisioning and sanitizing their vessels. Another explanation is that infectious diseases, which thrived in crowded conditions, were not as endemic on ocean voyages as has been previously thought.18 Additional evidence on debarkation morbidity was gathered for the peak migration years of 1753 and 1754, when German immigrant ships averaged over 300 passengers per vessel. Pennsylvania regulations required that all German immigrant ships be inspected by a physician for the presence of contagious diseases before being allowed to land. Although several reports mentioned that passeners were in a weakened condition, a total of only 70 individuals were enumerated as sick out of approximately 8,898 immigrants for a debarkation morbidity of under 1 percent. Only on the Henrietta were any immigrants (16) suspected of having contagious fevers. This evidence supports that in Table 4.2 in suggesting that serious voyage illness was not as prevalent as has been commonly believed. This rather low level of debarkation morbidity was also consistent with the rather low level of voyage mortality found in Table 4.1, although, because the mortality and morbidity samples do not overlap, no direct connection can be made.19

Post-­voyage mortality Above-­normal mortality and morbidity did not end with the voyage. Immigrants had to pass through a seasoning process during which they experienced a higher rate of sickness and death than did established residents. This process, usually thought to have lasted a year, was apparent to colonial contemporaries, who advised prospective immigrants to avoid arriving in the more deadly summer months.20 Seasoning mortality is an important and widely discussed topic. However, its magnitude has remained largely unmeasured. Walsh and Menard concluded: “Unfortunately, no method of measuring the proportion who died during seasoning has yet been discovered, but the number was substantial.” Gemery, Galenson, and Morgan assumed that seasoning mortality fell from substantial levels to near zero by 1660 in the middle colonies. Studies of eighteenth-­century migration to the middle colonies seldom mention the seasoning process. Yet, the high

80   German immigration to America, 1709–1835 mortality in eighteenth- and nineteenth-­century immigrant receiving cities, such as Philadelphia, Boston, and New York, has been partly attributed to immigrants dying at a higher rate than established residents.21 Due to this measurement problem, the source of the mortality differential between first year immigrants and established residents has remained obscure. Some have attributed seasoning mortality to the shock of entering a new epidemiological environment. Immigrants were exposed to diseases unfamiliar in Europe, such as mosquito-­borne malaria and yellow fever, to which they had relatively low immunities. For those who survived the infection in the first year, some immunity was conferred against infection in subsequent years, but first year immigrants did experience a relatively higher death rate. Others have attributed seasoning mortality to the weakened health conditions and lingering diseases contracted on the ocean voyage. They have suggested that first year immigrants died at a relatively higher rate because living aboard ship in the preceding months was more trying than living ashore.22 Evidence was taken from German immigration records and the bills of mortality for Philadelphia to estimate seasoning mortality for German immigrants to that city for the years 1738 to 1756 (see Table 4.3). This evidence allowed for a crude separation of the burials of new arrivals from that of established residents. Since the evidence was confined to Philadelphia the problem of controlling for different environmental zones was also minimized. However, the evidence relied on estimates of the city’s population, German immigration, and total burials which were derived through interpolation from incomplete records. Therefore, conclusions based on this evidence must remain speculative.23 The burials of new arrivals were estimated as a percentage of the burials in the city’s “Strangers’ Ground.” This ground accommodated those who died without belonging to a parish or without colonial relatives who took responsibility for their burial. From 1738 to 1756 these burials consisted largely of newly arrived German immigrants. In order to separate German immigrants from other burials in the Strangers’ Ground, the proportion buried in that ground relative to total burials in the city during the years when German immigrants did not arrive (1757 to 1762) was taken as representative of the proportion of non-­ German immigrant burials in that ground for the year 1738 to 1756.24 The bills of mortality counted burials for the entire calendar year. But German immigrants, who arrived only between August and November, were not at risk for the entire calendar year. Not all German immigrants remained long in Philadelphia. Many went into the countryside as servants or as farmers. Therefore, in the calendar year of arrival, a German immigrant was at risk in Philadelphia from anywhere between five months to only a few days. In terms of an annualized death rate, a downward biased estimate of seasoning mortality may be generated by assuming that all German immigrants were at risk for at least the full five months. This was the estimate reported in Table 4.3.25 The estimated annualized death rate, between 1738 and 1756, was 61.4 per 1,000 for first year immigrants and 37 per 1,000 for established residents. Therefore, even with this conservative estimate, seasoning mortality in the middle of

Morbidity and mortality   81 the eighteenth century was still substantial. New immigrants died at a rate at least 66 percent higher than established residents, and possibly much higher. Finally, immigrants were concentrated in the late teens and twenties relative to the age distribution of the established population. This age group experienced lower mortality relative to the very young and very old, ceteris paribus. Therefore, the age specific mortality differential may have been even greater. Although post-­voyage mortality was above the normal mortality for Philadelphians, it was still below voyage mortality. The evidence in Tables 4.1 and 4.3 suggest that the average annualized death rate was around 67 percent lower after arrival than during the voyage. There were a few years, however, when post-­ voyage mortality exceeded voyage mortality, most notably 1738, 1741, and 1754.26 What was the relationship between post-­voyage mortality and debarkation morbidity? For the years 1738 to 1754 debarkation morbidity in Table 4.2 was 2.42 percent. The estimated proportion of German immigrant burials to total German immigration for the same period was 2.58 percent. Thus it is possible that diseases contracted by the end of the voyage fully explained the estimated post-­voyage mortality, assuming that adult male debarkation morbidity was representative of total debarkation morbidity. A closer look, however, mitigated against this possibility. For the ten years of common evidence in Tables 4.2 and 4.3, the correlation coefficient between debarkation morbidity and the annualized death rate for first year immigrants was positive at 0.33, but its magnitude was small and only marginally statistically significant. Also, it was unlikely that all those who were sick when the ship docked died soon after. For example, in 1784, out of 276 sick immigrants admitted to Philadelphia’s marine hospital after landing, only 16 died; 260 were listed as “cured and discharged.”27 The yearly fluctuation in the mortality rate of resident Philadelphians was primarily due to periodic epidemics. These fluctuations may have affected the post-­ voyage mortality of immigrants. The correlation coefficient between the annualized death rates of new immigrants and established residents for the years 1738 to 1754 was positive, but at 0.24 its magnitude was small and only marginally statistically significant. The variance in immigrant mortality per year was also higher than the variance in resident mortality per year. The standard devi­ ation of the annualized death rate per 1,000 between 1738 and 1756 was 47 for first year immigrants but only 9 for established residents. Thus, fluctuations in mortality among Philadelphians had some influence, though probably minor, on the variance in immigrant post-­voyage mortality.28 Epidemics of various diseases in Philadelphia may have affected new immigrants and colonial residents differently. Since smallpox was more endemic to Europe than America, immigrants would have been more immune to smallpox than colonial residents. And since yellow fever and malaria were more endemic to Philadelphia than Europe, immigrants were less immune to yellow fever than colonial residents. Smith indicated that smallpox epidemics were chiefly responsible for mortality fluctuations in Philadelphia during the middle of the

82   German immigration to America, 1709–1835 eighteenth century. The evidence from Christ-­Church Parish in Philadelphia between 1747 and 1775 also indicated that smallpox was the single largest killer among parish residents and had the highest yearly variance among the identified causes (see Table 4.4). Yellow fever was a minor killer among parish residents, but it was more epidemic in nature than any other identified cause, with a standard deviation almost twice the mean number of deaths attributed to it. Except for smallpox, yellow fever, and the flux, the other major causes of death had a relatively constant impact on mortality in the parish over time, with relatively small standard deviations in relation to their means.29 Major outbreaks of smallpox were associated with above average mortality among Philadelphians. However, these epidemics were also associated with below average mortality among first year immigrants, most notably for the years 1751, 1752, and 1756. Yellow fever had the opposite impact. Duffy identified a yellow fever epidemic in Philadelphia in 1741, and Table 4.4 indicates that yellow fever was present in 1754. Both these years experienced dramatic peaks in immigrant post-­voyage mortality. The mortality of resident Philadelphians was above average in 1741 and below average in 1754. Thus, even as late as the eighteenth century, epidemiological events affected immigrants and established residents differently and so explained part of the differential fluctuations in mortality between these two groups.30 The eighteenth-­century North Atlantic voyage for European immigrants was difficult and hazardous; grim accounts are frequently cited in the historical literature. However, direct quantitative evidence for eighteenth-­century German immigration to Philadelphia indicates that the journey was not as deadly or prone to promoting serious diseases as has been commonly suggested. The voyage mortality for adults was under 4 percent, around 70 percent below that in the slave trade on average, and only about twice as high as early nineteenth-­century passage mortality. Children, however, suffered over twice the voyage mortality of adults. Although sickness was prevalent, the complaints by immig­rants were most likely caused by temporary seasickness rather than infectious disease. Debarkation morbidity was under 4 percent, declining over time, and lower on vessels which carried more passengers. Crowding, by itself, was not a contributor to voyage morbidity. Post-­voyage or seasoning mortality was more severe than has been commonly believed for the middle colonies in the eighteenth century. In the first few months after landing, immigrants died at a rate at least 66 percent higher than resident Philadelphians. This post-­voyage mortality was still below voyage mortality, by some 67 percent. The magnitude and variance of this post-­voyage mortality was related to several different causes. The magnitude of first year immigrant burials was similar to the magnitude of those arriving sick but a direct connection could not be established. To a minor extent, the annual variance in post-­voyage mortality was positively associated with the annual variance in debarkation morbidity and with the annual variance in the mortality of resident Philadelphians. Finally, immigrant post-­voyage mortality appeared to be relatively higher when yellow fever was prevalent, whereas resident Philadelphian

169 184    –    – 160 140 107 97 106 213 102 125    – 178 142 200 188 143 182 107 75 121 188    –    – 136 179 166 169 145 25

1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 Mean Standard deviation

1.18% 0.00  –  – 29.38 9.29 6.54 0.00 0.00 52.58 7.84 0.00  – 18.54 2.11 15.00 29.79 4.20 31.32 18.69 0.00 3.31 20.74  –  – 6.62 24.58 6.63 6.51 12.28 13.48

0.00% 0.00 – – 0.00 0.00 0.00 7.22 0.00 0.00 0.00 0.00 – 0.00 0.00 8.00 1.60 0.70 0.00 0.00 0.00 0.00 0.00 – – 0.00 0.00 0.00 0.00 0.73 2.15

Small pox Yellow fever 5.92% 5.43  –  – 2.50 12.86 2.80 0.00 1.89 0.94 3.92 9.60  – 0.00 2.82 0.50 4.26 4.20 5.49 4.67 0.00 5.79 3.72  –  – 2.21 0.56 0.00 9.47 3.73 3.36

Flux 24.85% 17.93  –  – 4.38 4.29 13.08 7.22 6.60 2.82 6.86 7.20  – 7.30 14.79 8.00 4.26 4.20 4.40 1.87 16.00 4.96 4.26  –  – 4.41 10.61 9.64 7.10 8.21 5.41

Fever 0.00% 0.00  –  – 0.00 0.00 5.61 7.22 5.66 3.76 4.90 5.60  – 12.36 4.93 4.50 4.79 7.69 3.85 3.74 5.33 2.48 2.13  –  – 3.68 4.47 8.43 4.14 4.39 2.89

Nervous fever 10.06% 2.17  –  – 0.00 0.71 0.00 13.40 7.55 7.04 9.80 11.20  – 8.99 13.38 11.00 9.57 9.09 13.74 9.35 8.00 17.36 9.57  –  – 17.65 8.38 7.23 14.20 9.14 4.81

12.43% 11.41  –  – 0.63 0.00 0.00 3.09 5.66 0.94 0.98 4.80  – 1.12 7.04 10.00 5.32 9.79 8.24 7.48 12.00 11.57 8.51  –  – 8.09 6.70 12.05 9.47 6.56 4.25

Purging and Fits and vomiting convulsions

Source: The records of the bills of mortality for Christ-Church Parish, Philadelphia on Mirocard as listed in Evans, American Bibliography.

Total buried

Year 0.00% 20.11  –  – 16.25 16.43 13.08 14.43 16.98 6.57 15.69 11.20  – 10.67 16.90 6.50 7.98 8.39 6.59 13.08 16.00 12.40 13.83  –  – 11.76 6.70 3.01 13.61 11.59 4.96

54.44% 57.05  –  – 53.14 43.58 41.11 52.58 44.34 74.65 49.99 49.60  – 58.98 61.97 63.50 67.57 48.26 73.63 58.88 57.33 57.87 62.76  –  – 54.42 62.00 46.99 64.50 56.63

Consumption Row total

Table 4.4 Distribution of major descriptive causes of mortality due to infectious disease in Christ-Church Parish, Philadelphia, 1747–75

84   German immigration to America, 1709–1835 mortality appeared to be relatively higher when smallpox was prevalent. Thus differential immunity between immigrants and established residents was a contributor to seasoning mortality. The identification of a single cause of seasoning mortality, however, is unlikely.

Notes   * Originally published as Farley Grubb, “Morbidity and mortality on the north Atlantic passage: eighteenth-­century German immigration to Pennsylvania,” Journal of Interdisciplinary History, vol. 17, no. 3 (Winter 1987): 565–85. Copyright 1987. © 1987 by the Massachusetts Institute of Technology and the editors of The Journal of Interdisciplinary History. Reprinted with permission. Farley Grubb is Assistant Professor of Economics at the University of Delaware. The author would like to thank Raymond Cohn, David Eltis, Stanley Engerman Robert Fogel, David Galenson, Henry Gemery, Saul Hoffman, Jon Moen, Randy Nelson, and Richard Steckel for helpful comments on earlier versions of this article.   1 See the discussions of general passage conditions in John Duffy, “Passage to the Colonies,” Mississippi Valley Historical Review, XXXVII (1951), 21–38; Herbert S. Klein, The Middle Passage (Princeton, 1978), 68–72; Stanley Lebergott, The Americans: An Economic Record (New York, 1984), 182–6; Abbot E. Smith, Colonists in Bondage (New York, 1947), 207–55. For discussions of voyage conditions particular to the German trade, see Frank R. Diffenderffer, “The German Immigration into Pennsylvania through the Port of Philadelphia, and ‘The Redemptioners’,” Pennsylvania German Society, X (1899), 55–68, 147, 256–62; Karl F. Geiser, Redemptioners and Indentured Servants in the Colony and Commonwealth of Pennsylvania (New Haven, 1901), 43–58; Henry E. Jacobs, “The German Emigration to America, 1709–1740,” Pennsylvania German Society, VIII (1897), 80–109; Walter A. Knittle, Early Eighteenth Century Palatine Emigration (Philadelphia, 1937), 135–59.   2 Concentration on the German trade was chosen because of the availability of evidence. Germans were the major non-­British colonists in North America. They arrived primarily in the middle of the eighteenth century, most entering through Philadelphia. By mid-­ century they were thought to comprise a majority of Pennsylvanians and by 1790 they represented about one-­third of the state’s population. The threat of German influence overwhelming English culture led colonial authorities in Pennsylvania to keep more comprehensive records of German immigrants than of any group. See the discussion in Diffenderffer, “Immigration into Pennsylvania,” 34–147; Jacobs, “Emigration To America,” 80–149; Thomas L. Purvis, “The European Ancestry of the United States Population, 1790,” William and Mary Quarterly, LXI (1984), 85–135; Marianne S. Wokeck, “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, CV (1981), 249–78.   3 Donald F. Durnbaugh, “Christopher Sauer, Pennsylvania-­German Printer,” Pennsylvania Magazine of History and Biography, LXXXIII (1958), 324–5. There were many favorable accounts of the passage, for the most part ignored in the historical literature, which were also sent to Germany. For example, Sauer, although constantly seasick, wrote of his voyage in 1724: “The ship voyage is as one takes it. For my part, I maintain that it is a comfortable trip if one carries victuals to which one is accustomed and controls his imagination.” Durnbaugh, “Two Letters from Germantown,” ibid., LXXXIV (1959), 228.   4 For discussions of the relative income and wealth differentials between Europe and the colonies, see Donald R. Adams, Jr., “Some Evidence on English and American Wage Rates, 1790–1830,” Journal of Economic History, XXX (1970), 499–520; Alice H. Jones, Wealth of a Nation to Be (New York, 1980). For examples of these income differences based on German immigrant letters, see Durnbaugh, “Two

Morbidity and mortality   85 Letters,” 219–33; Julius F. Sachse, “A Missive From Pennsylvania in the Year of Grace 1728,” Pennsylvania German Society, XXVIII (1909), 5–25. See also the discussions in Raymond L. Cohn, “Mortality on Immigrant Voyages to New York, 1836–1853,” Journal of Economic History, XLIV (1984), 289–90; Lebergott, Americans, 178–82; Adam Smith, The Wealth of Nations (New York, 1937), 92.   5 Gottlieb Mittelberger (trans. and ed. Oscar Handlin and John Clive), Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754 (Cambridge, Mass., 1960), 16–17. See also Diffenderffer, “Immigration into Pennsylvania,” 189–92. For a discussion of the magnitude of servant immigration see Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, XXII (1985), 316–39. Roughly 70 percent of German servants were sold within 15 days of arrival, a figure derived from “The Book of Redemptioners, 1785–1804” (ms. held at the Pennsylvania Historical Society, Philadelphia).   6 The samples of British troops, convicts, and Palatines had a voyage mortality of 11, 10–15, and 15.8 percent, respectively. See Klein, Middle Passage, 71; A. E. Smith, Bondage, 118; Knittle, Palatine Emigration, 147–8. However, recent studies have lowered the estimate of convict voyage mortality to around 4 percent, or similar to that found in table 4.1. See A. Roger Ekirch, “Bound for America: A Profile of British Convicts Transported to the Colonies, 1718–1775,” William and Mary Quarterly, XLII (1985), 190; Kenneth Morgan, “The Organization of the Convict Trade to Maryland: Stevenson, Randolph & Cheston, 1768–1775,” ibid., 213–15. For conclusions about the minimum passage mortality drawn from this evidence, see Henry A. Gemery, “Emigration from the British Isles to the New World, 1630–1700: Inferences from Colonial Populations,” Research in Economic History, V (1980), 187; Lebergott, Americans, 182. For discussions of the relative severity of the German trade, see Duffy, “Passage,” 23, 27; Richard Hofstadter, America at 1750 (New York, 1973), 37–42; Mittelberger, Journey, 14–15; Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, 1934), I, 58. By contrast, two other large samples, seldom cited, yielded radically lower estimates, i.e., only one death in over 600 Moravians passengers arriving between 1736 and 1762 and only six deaths out of 1,500 immigrants transported to Georgia in 1741. See John W. Jordan, “Moravian Immigration to Pennsylvania, 1734–1765,” in Michael Tepper (ed.), Emigrants to Pennsylvania 1641–1819 (Baltimore, 1978), 53; A. E. Smith, Bondage, 217.   7 Billy G. Smith, “Death and Life in a Colonial City: A Demographic Analysis of Philadelphia,” Journal of Economic History, XXXVI (1977), 872. See also the opinion of Carl Wittke, We Who Built America (New York, 1936), 113–14; Lebergott, Americans, 182. The range of 9–15 percent was taken from the following studies: Roger Anstey, “The Volume and Profitability of the British Slave Trade, 1761–1807,” in Stanley L. Engerman and Eugene D. Genovese (eds.), Race and Slavery in the Western Hemisphere: Quantitative Studies (Princeton, 1975), 25–9; David Eltis, “The Direction and Fluctuation of the Transatlantic Slave Trade, 1821–1843: A Revision of the 1845 Parliamentary Paper,” in Gemery and Jan S. Hogendorn (eds.), The Uncommon Market (New York, 1979), 294; David W. Galenson, Traders, Planters, and Slaves: Market Behavior in Early English America (New York, 1986), 38–9; Klein, “The Trade in African Slaves to Rio de Janeiro, 1785–1811: Estimates of Mortality and Patterns of Voyage,” Journal of African History, X (1969), 538; idem, Middle Passage, 187, 195; Johannes Postma, “Mortality in the Dutch Slave Trade, 1675–1795,” in Gemery and Hogendorn (eds.), Uncommon Market, 225. The bottom of the distribution in these slave studies were samples with 3.5 to 5 percent passage mortality.   8 Lebergott, Americans, 182; Klein, Middle Passage, 70–1, 90–3; Wittke, We Who Built America, 113–14.   9 Mittelberger, Journey, indicated that he took over 100 days at sea, whereas the average crossing was only about two months. Strassburger, Pioneers, I, 445, indicated that Mittelberger’s ship carried 486 passengers. The average in the German trade was

86   German immigration to America, 1709–1835 300 between 1750 and 1754, and almost half that at other times. Wokeck, “German Immigration,” 260–1. Thus Mittelberger experienced a relatively longer, more crowded passage. The disaster rate was derived from Strassburger, Pioneers, I; Wokeck, “German Immigration,” 254–5, and covered only the port of Philadelphia. On mortality in the tropics versus that in the temperate zone for eighteenth-­century Europeans, see Klein, Middle Passage, 69–70; Phillip D. Curtin, “Epidemiology and the Slave Trade,” Political Science Quarterly, LXXXIII (1968), 190–216; Richard S. Dunn, Sugar and Slaves (New York, 1972), 300–34. 10 The sample of 1,566 Germans excluded port deaths and ships which experienced no fatalities, because it could not be determined if a ship with no fatalities really had no deaths or was simply not reporting on mortality. See Cohn, “Mortality on Immigrant Voyages,” 293, on the use of this method. This construction produced an upward bias in the mortality estimated from this sample, since evidence from German diaries indicated that there were voyages that suffered no fatalities, although the overall frequency of these voyages is unknown. See Jacobs, “Emigration to America,” 98; John P. Meurer, “From London to Philadelphia, 1742,” Pennsylvania Magazine of History and Biography, XXXVII (1913), 94–106; Sachse, “Missive From Pennsylvania,” 24–5. 11 This sample excluded shipwrecks. For 1727–75, adding an estimate of deaths from shipwrecked German immigrant vessels to the estimated 3.8 percent mortality per voyage raised the overall passage mortality to around 5.5 percent, or an annualized crude death rate of at least 264 per 1,000. There were roughly 70,000 German immigrants and five shipwrecks. Wokeck, “German Immigration,” 254–61. If each shipwreck had 400 additional immigrants and suffered 70 percent mortality, this would add 2,000 to the immigrant count and 1,400 to the number dead. The relative hazards of transatlantic travel were even greater than suggested, since immigrants were concentrated in the low mortality ages of the late teens and early twenties relative to non-­ immigrant populations. Thus age-­specific differences in crude death rates would be even higher. In addition, eighteenth-­century Philadelphia was found to have higher crude death rates than Andover, Boston, or Nottingham, England. See B. G. Smith, “Death and Life,” 887–8. 12 Cohn, “Mortality on Immigrant Voyages,” 229; Eltis, “Free and Coerced Transatlantic Migration: Some Comparisons,” American Historical Review, LXXXVIII (1983), 273; Friedrich Kapp, Immigration and the Commissioners of Emigration of the State of New York (New York, 1870), 188–95. 13 Diffenderffer, “Immigration into Pennsylvania,” 190. Many European immigrant passage contracts were very detailed in terms of provisions, accommodations, and liabilities, as in the contract for the voyage of the Commerce, which carried German passengers and servants from Holland to Philadelphia in 1803. Strassburger, Pioneers, III, 131–4. 14 Seasickness might have been more troublesome than was commonly believed. Constant seasickness over a two month voyage with highly salted provisions and limited fresh water might have led to temporary malnutrition, a greater susceptibility to serious disease, and possibly even death. Durnbaugh, “Two Letters,” 220; A. E. Smith, Bondage, 215. Typhus, thought to be the most common problem, was an acutely infectious bacteria transmitted by lice feces being crushed into bite sites and other skin breaks, or by being inhaled. Both typhoid and dysentery were spread by the human feces to oral ingestion route. Thus poor sanitation, close contact, and the poor personal hygiene, prevalent on long ocean voyages, were conducive to the spread of these diseases. Smallpox was endemic to Europe, but enough immunity had been built up by the eighteenth century that it was regarded almost as a childhood disease by Europeans. Finally, dietetic disorders were also thought to be a minor additional problem for immigrants. Duffy, Epidemics in Colonial America (Baton Rouge, 1953); B. G. Smith, “Death and Life,” 870–9; Dunn, Sugar and Slaves, 302–11; Curtin, “Epidemiology,” 190–216; Paul B. Baeson, Walsh McDermott, and James B. Wyngaarden

Morbidity and mortality   87 (eds.), Cecil Textbook of Medicine (Philadelphia, 1979), 255–9, 318–28, 446–9, 589–94; Richard A. Meckel, “Immigration, Mortality, and Population Growth in Boston, 1840–1880,” Journal of Interdisciplinary History, XV (1985), 395–403; Robert Higgs, “Cycles and Trends of Mortality in Eighteen Large American Cities, 1871–1900,” Explorations in Economic History, XVI (1979), 390. 15 See Strassburger, Pioneers, I, for a description of immigrant processing and the evidence used. This sample comprised 34 percent of all German immigrant ships and adult males, over age 15, listed by Strassburger as arriving in Philadelphia in these years. Ships where no one was listed as sick were excluded from the sample, because it was not known whether these ships really had no morbidity or if they simply were not reporting any. This exclusion biased the morbidity estimate upwards. The number listed as “sick” and “on board” fully accounted for the discrepancy between adult males listed on the captain’s passenger list and those who signed the loyalty oath. Only those explicitly listed as sick were included in the morbidity calculation. 16 Baeson, McDermott, and Wyngaarden (eds.), Cecil, 255–9, 318–28, 446–9. 17 The average was estimated at the mean year of the sample, 1742. Since more of the sample was concentrated in the low morbidity years at the end of the sample, the average reported in table 4.2 was lower. 18 Grubb, “Was Shipping of German Immigrants to Colonial America Monopolized?” Pennsylvania Magazine of History and Biography, forthcoming; Wokeck, “German Immigration,” 260–1; Strassburger, Pioneers, I; “Ship Register for the Port of Philadelphia, 1726–1775,” Pennsylvania Magazine of History and Biography, XXIII– XXVIII (1899–1904). Ships that made repeat voyages between 1727 and 1754 showed a clear increase in the number of passengers carried. Although the latter sample years saw an increase in the ratio of women and children to adult men, this shift was small, thus leaving the number of adult men as a good proxy for changes in the number of passengers per ship. Evidence from the slave trade also suggested that crowding did not appreciably affect voyage mortality, except perhaps during loading. Eltis, “Mortality and Voyage Length in the Middle Passage: New Evidence from the Nineteenth Century,” Journal of Economic History, XLIV (1984), 307; Galenson, Traders, Planters, and Slaves, 42–5; Charles Garland and Klein, “The Allotment of Space for Slaves aboard Eighteenth-­Century British Slave Ships,” William and Mary Quarterly, XLII (1985), 238–48; Richard H. Steckel and Richard A. Jensen, “New Evidence on Causes of Slave and Crew Mortality in the Atlantic Slave Trade,” Journal of Economic History, XLVI (1986), 64–70. 19 These reports were listed with their corresponding ships in Strassburger, Pioneers, I, and covered 31 of the 40 German passenger ships arriving in these years. The number of passengers on these ships was derived from Wokeck, “German Immigration,” 260–1, adjusted for the nine ships without surviving reports. The health report estimate was lower than those in table 4.2 because the health reports were concerned with contagious diseases and focused on all passengers rather than just adult males. See also Thomas Graeme and Thomas Bond, “A Colonial Health Report of Philadelphia, 1754,” Pennsylvania Magazine of History and Biography, XXXVI (1912), 476–9. 20 David P. de Vries (trans. Henry C. Murphy), Voyages from Holland to America (New York, 1857), III, 75; Diffenderffer, “Immigration into Pennsylvania,” 29; David W. Galenson, White Servitude in Colonial America (New York, 1981), 87–91: Gemery, “Emigration from the British Isles,” 187–91; Edmund S. Morgan, American Slavery-­ American Freedom (New York, 1975), 158–62, 175, 184. Over 97 percent of the German immigrants leaving Rotterdam between 1727 and 1775 arrived in Philadelphia during the months of late August through November, derived from Strassburger, Pioneers, I. 21 Galenson, White Servitude, 89–91, 216, 266; Gemery, “Emigration from the British Isles,” 185–90; Higgs and David Booth, “Mortality Differentials within Large Amer­ ican Cities,” Human Ecology, VII (1979), 363–4; Morgan, American Slavery, 158–62,

88   German immigration to America, 1709–1835 175, 181–4; B. G. Smith, “Death and Life,” 870–4; Lorena S. Walsh and Russel R. Menard, “Death in the Chesapeake: Two Life Tables for Men in Early Colonial Maryland,” Maryland Historical Magazine, LXIX (1974), 211–27. 22 Curtin, “Epidemiology,” 190–216; Leonard W. Labaree et al., The Papers of Benjamin Franklin (New Haven, 1959), III, 439; Darret B. Rutman and Anita H. Rutman, “Of Agues and Fevers: Malaria in the Early Chesapeake,” William and Mary Quarterly, XXXIII (1976), 31–60; B. G. Smith, “Death and Life,” 872–3. 23 The major problem in studies of seasoning mortality has been the inability to separate immigrant deaths from resident deaths within the same disease environments: e.g., see the analysis of the first years of the Virginia colony in Carville V. Earle, “Environment, Disease, and Mortality in Early Virginia,” in Thad W. Tate and David L. Ammerman (eds.), The Chesapeake in the Seventeenth Century (New York, 1979), 96–125; E. S. Morgan, American Slavery, 73, 101, 158–85. These problems were also present in the discussions of colonial South Carolina by Peter H. Wood, Black Majority (New York, 1974), 63–91, and in the studies of nineteenth-­century immigrant effects in Gretchen A. Condran and Rose A. Cheney, “Mortality Trends in Philadelphia: Age- and Cause-­ Specific Death Rates, 1870–1930,” Demography, XIX (1982), 117–19; Higgs, “Cycles and Trends of Mortality,” 381–408; idem and Booth, “Mortality Differentials,” 356–69; Meckel, “Immigration, Mortality, and Population Growth,” 395–403. 24 See the assessment of Franklin in Labaree, Benjamin Franklin, III, 439. The bills of mortality directly stated that “Dutch and Other White People” were buried in this ground up through 1756 when the “Dutch” heading was dropped. The 1754 colonial council minutes also stated that newly arrived German immigrants were buried in the Strangers’ Ground. Pennsylvania Colonial Records, VI, 173, quoted in Jacobs, “Emigration to America,” 98. The procedure used to separate German immigrant from other burials may be very accurate, since the number of German immigrant burials estimated in Table 4.3 for 1754 exactly matched the number reported buried there in 1754 by the colonial council. See Jacobs, “Emigration to America,” 98. The proportion of arrivals who had colonial relatives to tend to their burial is unknown. Given some positive number, this figure would bias the estimated seasoning mortality presented in Table 4.3 to below the true seasoning mortality. 25 Over half the German immigrants were probably servants. Grubb, “Incidence of Servitude,” 320–32. On average these servants were sold within half a month of arrival and only 33 and 45 percent were purchased by Philadelphians from 1771–3 and from 1787–1804, respectively. The rest were sold into the countryside. See idem, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­Atlantic Economy,” Social Science History, IX (1985), 249–75; “Book of Redemptioners, 1785–1804.” 26 These were also the years when contemporaries remarked on the appalling conditions in the German trade. The use of literary evidence from only these high mortality years would overstate the average mortality in the trade for other years. E.g., see Waldemar Westergaard, “Two Germantown Letters of 1738,” Pennsylvania Magazine of History and Biography, LVI (1932), 9–14; Jacobs, “Emigration to America,” 98; Diffenderffer, “Immigration into Pennsylvania,” 239–49; B. G. Smith, “Death and Life,” 875. 27 The t-­statistic was 0.988, which was statistically significant only at the 0.2 level. Graeme and Bond, “Colonial Health Report,” 478; Luther R. Keller, “List of Patients Admitted to and Discharged from the Marine Hospital, Philadelphia, 1784,” Pennsylvania Magazine of History and Biography, XXV (1901), 92–100. 28 See B. G. Smith, “Death and Life,” 874–9; Duffy, Epidemics. The t-­statistic was 0.891, which was only statistically significant at the 0.2 level. There were 15 years of common evidence between the two samples. 29 Curtin, “Epidemiology,” 190–216; Duffy, Epidemics, 16–112, 138–63; Rutman and Rutman, “Agues and Fevers,” 31–60; B. G. Smith, “Death and Life,” 876–9. 30 Duffy, Epidemics, 153–5.

5 Age, occupation, and family composition*

The largest group of non-­British Europeans arriving in North America during the eighteenth and nineteenth centuries were Germans. Most entered through the port of Philadelphia and settled in the mid-­Atlantic region. From 1730 to 1760, German immigrants represented 20 to 30 percent of the population growth in the middle colonies. Pennsylvanians of German ancestry accounted for 50 to 60 percent of Pennsylvania’s population in 1760 and 33 percent in 1790. They were a potent force in shaping the social, economic, and political life of the mid-­ Atlantic region.1 The social composition of German immigrants has been used to explain why Germans emigrated and how German colonists influenced, and were influenced by, American society. However, the exact nature of this composition has remained unresolved. Were Germans “pushed” to emigrate because economic conditions and inheritance customs made farm holdings or craft trades marginal? Or were they “pulled” to America by recruiters? Were German colonists mostly uneducated peasant farmers who arrived in family groups, clung to traditional kinship or communal ways, and only slowly assimilated Anglo-­American culture? Or did they have an individualistic spirit, an independent migratory propensity, an extensive interest in education, books, and newspapers, many craft skills, and an ability quickly to assimilate Anglo-­American culture?2 The literary and quantitative evidence on the social composition of German immigrants has been fragmentary and less than representative. Literary evidence has many biases. For example, British colonists may have held prejudicial views of German immigrants. Henry Melchior Muhlenberg, the head of the German Lutheran Church in Pennsylvania during the mid-­eighteenth century, concluded that “The ‘superior’ nations look upon the Germans as nothing but wig-­blocks, which are thick and hard to be sure, but wanting brains.” The quantitative evidence has been limited and has typically measured the character of German immigrants either before they left their home village or after they had acclimatized to the New World, rather than at the point of debarkation in America. These measures may be biased because German emigration records are fragmentary, unrepresentative, and seldom indicate whether emigrants reached their destinations and because the social characteristics of immigrants changed while they lived in  America. Finally, generalizations about the social composition of German

90   German immigration to America, 1709–1835 immigrants have seldom indicated magnitudes or trends—for example, the number of farmers or families who emigrated, their level of education, and how these patterns changed over time.3 Measuring the social composition of German immigrants at the point of debarkation in America with quantitative evidence yields better estimates. Ship manifests for German immigrants arriving in Pennsylvania have survived for 1727 through 1820. A total of 89,544 passengers were enumerated. Although the purpose and information in these manifests varied over time, the coverage is relatively complete and unbiased. Compared with records of other immigrant groups, the German passenger manifests represent the most extensive and comprehensive body of direct information collected at the point of immigration for a single migration stream.4 The passenger lists frequently recorded information on occupation, literacy, age, and family composition. This information was extracted and forms the basis here for estimating the secular trends in the social composition of German immigrants. Evidence from the first large German migration to America in 1709 was added to expand the coverage of the study. In addition, evidence concerning English immigration to Pennsylvania and the Pennsylvania population as a whole was included to aid ethnic and geographical comparisons. The magnitude, duration, and continuity of this evidence provides valuable insights into the long­term trends in immigration to America, the causes of German emigration, and the assimilation of Germans into American society.5

Family composition The measure of family composition which uses the largest amount of evidence is the dependency ratio, which is the number of women and children (dependents) per adult male (see Table 5.1). The number of dependents per adult male declined continuously from three in 1709 to one by the 1790s, a 68 percent reduction. Major breaks in immigration during the Queen Anne’s War, the Seven Years’ War, and the American Revolution were followed by renewed migrations which were progressively less family oriented. Although the trend reversed and the dependency ratio rose to 1.3 in 1815–20, it remained well below pre-­1760 levels. The dependency ratio among English immigrants in the 1770s was 5.4 times lower than the German ratio. Relative differences in dependency between German and English immigrants and among Germans over time could be caused either by differences in family size or by differences in the proportion of married to single adult migrants.6 For a smaller sample of German passenger manifests, family composition can be decomposed into single adults, married persons, single household heads, and dependent children (see Table 5.2). Between 1709 and the 1790s, the proportion of single adult males rose by 33 percentage points whereas the proportion of married persons and dependent children fell by 15 and 25 percentage points respectively. The proportion of single adult females and single household heads was small and changed little over the period. The trend reversed after 1804, but

Age, occupation, and family composition   91 Table 5.1 Ratio of dependent migrants to adult male migrants Immigrant group

Adult men

Women and children

Ratio of women and children to adult men

Germans, 1709 Germans, 1727–56 Germans, 1761–75 Germans, 1785–1804 Germans, 1815–20 English, 1773–6 (all destinations) English, 1774–6 (Pennsylvania)

1,662 21,107 5,081 4,051 3,205 4,800 732

4,858 37,330 6,882 3,786 4,170 1,200 176

2.92 1.77 1.35 0.93 1.30 0.25 0.24

Sources: Derived from “Philadelphia Custom House Records, 1800–1883” (ms. held at the National Archives, Washington, D.C., microfilm # 425); Mildred Campbell, “English Emigration on the Eve of the American Revolution,” American History Review, LXI (1955), 4–5; Frank R. Diffenderffer, “The German Exodus to England in 1709,” Pennsylvania German Society, VII (1897), 321–2; Gerald Fothergill, “Emigrants from England, December 1773 to April 1776,” New England Historical and Genealogical Register, LXII–LXV (1908–11); Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, Pa., 1934), I, III; Marianne S. Wokeck, “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, CV (1981), 260–1.

Table 5.2 Social composition of German and English immigrants to Pennsylvania, 1709–1820 Social group

Germans 1709

Single adult males Single adult females Single male household heads Single female household heads Married persons Dependent children

English 1727–38

1785–1804 1815–20

N = 6,520 N = 4,568 N = 7,837 5.9% 18.1% 39.2% 1.6 8.0 7.7 1.2 0.6 2.0 0.7 0.9 38.0 38.9 23.7 52.5 33.1 27.9 100.0% 100.0% 100.0%

1774–6

N = 7,375 N = 908 30.5% 77.2% 5.2 12.0 0.4 0.2 0.3 0.3 25.0 6.4 38.6 3.9 100.0% 100.0%

Sources: See Table 5.1. Notes The observations per group are from different sample sizes and only the percentage distribution should be compared. The English sample includes emigrants bound for Pennsylvania only. The 1709 German sample consists of emigrants, most of whom embarked for New York.

remained far from early eighteenth-­century levels. The proportion of single adults rose from 8 percent in 1709 to 47 percent by the 1790s and then fell to 36 percent by 1815–20. Although family orientation declined over the century, single adult migrants never became a majority.7 German immigration was significantly more family oriented than English immigration. Single adults comprised 89 percent of English immigrants in the 1770s, twice the German percentage, whereas the percentage of English married

92   German immigration to America, 1709–1835 persons and children was over 3.7 and 7.1 times lower than among German immigrants. Although single adult females were proportionally twice as prevalent among English immigrants, the female proportion was higher among German immigrants when all females were considered (single and married women, and children). The number of dependent females was over three times as large as independent females among German immigrants, and the number of independent females was over twice as large as dependent females among English immigrants. Changes in female immigration help to explain the transformation from bound to free labor in the mid-­Atlantic region.8 For a smaller sample of German passenger lists, family size can be reconstructed (see Table 5.3). The average number of children per married couple was 2.76 in 1709; it dipped to 1.50 in the 1730s, rose to 2.12 by the 1790s, and finally to 2.66 by 1815–20. Although the dependency ratio decreased over the Table 5.3 Family composition of German and English immigrants to Pennsylvania, 1727–1820 Family composition

Germans 1727–38

Germans Germans 1785–1804 1815–20

English 1774–6

Married couples with: No children 1 child 2 children 3 children 4 children 5 children 6 children 7+ children

N = 888   38.1%   21.0   16.4   11.6    7.5    2.9    1.4    1.1 100.0%

N = 927   27.5%   19.4   17.6   11.8   10.1    7.1    3.3    3.2 100.0%

N = 490   22.3%   11.6   17.4   16.5   10.6   10.0    7.1    4.5 100.0%

N = 29   55.2%   17.2   13.8   10.3    3.5    0.0    0.0    0.0 100.0%

Children in families with: 1 child 2 children 3 children 4 children 5 children 6 children 7+ children

N = 1,329   14.0%   22.0   23.2   20.2    9.8    5.4    5.5 100.1%

N = 1,961    9.2%   16.6   16.7   19.0   16.8    9.5   12.2 100.0%

N = 1,302    4.4%   13.0   18.7   16.0   18.8   16.1   13.0 100.0%

N = 26   19.2%   30.8   34.6   15.4    0.0    0.0    0.0 100.0%

Average number of children per married couple Average number of children per married couple with children

1.50

2.12

2.66

0.90

2.42

2.92

3.42

2.00

Sources: See Table 5.1. notes See the notes for Table 5.2. All samples include immigrants bound for Pennsylvania only and exclude families with single household heads. The 1709 evidence and half of the married couples in the 1815–20 evidence could not be disaggregated by family size.

Age, occupation, and family composition   93 eighteenth century and then rose after 1804, family size initially fell and then rose after the 1730s. How shifts in family size changed the social composition of German immigrants can be revealed with counterfactual estimates. If family size in 1709, the highest number found, had remained constant through 1820 and the numbers in other social categories had remained unchanged, then the dependency ratio would have declined by 40 percent less for the 1730s, and 8 percent less for the 1790s, with no change for 1815–20 compared with 1709. The percentage of single adult males would have been reduced by about 4 percentage points in the 1730s and 1790s, with no decrease in 1815–20. Alternatively, if family size in the 1730s, the lowest number found, had persisted for other periods, then the dependency ratio would have been reduced by 28 percent in 1709, 14 percent in the 1790s, and 32 percent in 1815–20. Single adult males would have increased to 7, 42, and 37 percent of the 1709, 1785–1804, and 1815–20 migrations. Finally, if the 1790s family size has persisted to 1820, then the dependency ratio for 1815–20 would have been reduced by 18 percent, and single adult males would have increased by 4 percentage points.9 These counterfactual estimates indicate that changes in the social composition of German immigrants had different causes over the century. Shifts in family size and in the relative proportion of single adults sometimes moved in the same direction and sometimes worked in opposite directions in altering the dependency ratio. Forty percent of the decline in the dependency ratio from 1709 to the 1730s was the result of decreases in family size and 60 percent the result of increases in the number of single adults relative to married persons. The decline in the dependency ratio from the 1730s to the 1790s was caused 115 percent by increases in single adults relative to married persons. The relative increase in single adult migrants swamped increases in family size and created a net reduction in the dependency ratio. Finally, the rise in the dependency ratio from the 1790s to 1815–20 was caused 35 percent by increases in family size and 65 percent by decreases in single adults relative to married persons. The difference in family size between German and English immigrants was substantial. German family size was 67 percent larger in the 1730s and 136 percent larger in the 1790s than English family size in the 1770s. The difference between the social composition of English immigrants in the 1770s and German immigrants in the 1790s was caused 6 percent by relatively smaller English family size and 94 percent by relatively more single adults compared with married persons among the English. The distribution of children per married couple can be used to explain the secular increase in family size among German immigrants. The most notable change was the decline in the proportion of married couples with no children, from 38 percent of all couples in the 1730s to 22 percent by 1815–20. The increase in family size from the 1730s to the 1790s was caused 74 percent by decreases in the proportion of couples without children and 26 percent by increases in the number of children per couple with children. The increase in family size from the 1790s to 1815–20 was caused 59 percent by decreases in the proportion of couples without children and 41 percent by increases in the

94   German immigration to America, 1709–1835 number of children per couple with children. Finally, the smaller family size of English immigrants in the 1770s was caused 74 percent by relatively more couples without children and 26 percent by relatively lower numbers of children per couple with children compared with German immigrants in the 1790s.10 For German immigrants, the increase in family size among couples with children occurred through an expansion and flattening of the distribution of children per family. Most children were in families with two to four children in the 1730s, in families with two to five children by the 1790s, and in families with two to seven children by 1815–20. The proportion of one-­child families also declined, from 14 to 4 percent from the 1730s to 1820. The larger family size in 1815–20 was caused in part by a relative increase in the number of older children accompanying their parents. By comparison, family size among English immigrants in the 1770s was much smaller, with all families having fewer than five children. The changing proportion of childless couples and single adults among German immigrants to America may have been influenced by the marriage restrictions and inheritance customs that affected population growth and family formation in eighteenth-­century Germany, and by competing colonization schemes in Eastern Europe. Because marriage regulations ceased to apply when citizens left the jurisdiction, many requests for permits to emigrate were accompanied by marriage applications. The relatively high incidence of childless couples among German immigrants early in the eighteenth century may have represented newlyweds making a joint migration–marriage decision. By late in that century, marriage restrictions were less important. The rising proportion of single adults reflected a shift in German emigration from regions of partible inheritance to regions of impartible inheritance. Partible inheritance encouraged the formation of nuclear families on increasingly smaller farms. The prospects of further fragmentation might have caused these families to emigrate. Impartible inheritance both reduced the wealth of nonheirs and restricted their ability to marry. Emigration was one way in which these single adult nonheirs could improve their circumstances. Therefore, a relative increase in emigration from those regions which practiced impartible inheritance would be consistent with the rising proportion of single adults among German immigrants. Finally, declines in the proportion of childless couples and married adults was also caused by the rising German colonization of Eastern Europe after 1760. Promoters of Eastern European colonization schemes preferred married emig­ rants because they were considered more likely to become stable agrarian settlers. This preference siphoned off married immigrants, especially newlywed childless couples, from the pool of emigrants leaving for America.11 The resurgence of family-­oriented migration among German immigrants after 1804 continued through the nineteenth century. By the 1840s, the proportion migrating as families reached 70 percent and subsequently stabilized at about 80 percent. The proportion of single adult migrants did not return to the high levels of the 1790s until the very end of the nineteenth century. The proportion migrating with other family members also increased among British immigrants, reaching 77 percent of English, 66 percent of Scottish, and 61 percent of Irish immigrants by the

Age, occupation, and family composition   95 1830s. Compared with other immigrants, the family composition of German immigrants was less atypical in the nineteenth century than in the eighteenth century.12

Age structure German passenger lists occasionally recorded the ages of individuals. Table 5.4 presents the age distributions for adult men and women by five-­year intervals. Over the century, the age distribution for males did not change and the distribution for females became only slightly more concentrated in the 21–5 age range. The female distribution is roughly similar to the male distribution, which would be expected given that adult females were largely dependent migrants. The secular stability of the age structure indicates that the increasing proportion of single adult migrants, the rising family size, and the declining number of childless couples were caused more by relative shifts in economic prospects and legal restrictions in Germany and America for various family groups per age category than by changes in the age structure of the immigrants. Younger adults have a higher propensity to migrate because the returns to investing in a new permanent location are realized over a longer period. Although adult migration was dominated by those in their twenties, a fair number of older migrants were also present. Over 30 percent were above age 30, and over 14 percent were above age 40. A few individuals were above age 60, but they appear to be widowed dependents of younger family groups. Within the 16–20 age category there was a sharp decline in those reported as being age 16 or 17. The age distribution of immigrants between ages 1 and 15 was relatively uniform, which would be expected for dependent migrants such as children. Table 5.4 Age distribution of adult German and English immigrants to Pennsylvania, 1730–1820 Age

16–20 21–25 26–30 31–35 36–40 41–45 46–50 51+

Germans 1730–54

Germans 1787–1807

Germans 1816–20

English 1774–6

Men

Women Men

Women Men

Women Men

Women

N = 4,513   19.0%   24.6   20.6   10.3   10.4    5.8    5.4    3.9 100.0%

N = 794   20.3%   17.3   20.8   11.2   12.6    5.3    6.8    5.8 100.1%

N = 333 16.2% 20.1 21.6 12.0 13.5   7.5   6.3   2.7 99.9%

N = 241   20.3%   25.7   17.0   13.3    9.5    7.1    4.6    2.5 100.0%

N = 92   20.6%   47.8   15.2    4.4    5.4    2.2    3.3    1.1 100.0%

N = 837   18.8%   24.6   20.7   10.4   11.0    7.5    3.7    3.4 100.1%

N = 605   20.0%   26.3   22.6    9.3    7.8    5.8    3.6    4.6 100.0%

N = 639   25.5%   42.3   16.9    5.3    6.1    2.0    1.3    0.6 100.0%

Sources: See Table 5.1. Notes See the notes for Table 5.2. The female sample in the 1730–54 category covers only the years 1730–38.

96   German immigration to America, 1709–1835 This  age distribution indicates that independent migration decisions were not generally made until about age 18. The adult age distribution of English immigrants was different, with over 40 percent being concentrated in the 21–25 age category. Only 15 percent were above age 30. German and English differences in age and family composition were related. Adult German immigrants were generally older than English immigrants, because more of the German immigrants were married and married persons were generally older than single adults. However, even within the single and married categories, German immigrants were generally older than the English. This difference is also consistent with the relatively smaller family size among English immigrants.13

Literacy Immigrants possessed knowledge and skills which they had acquired in their homeland. Measuring the complete spectrum of educational, occupational, and cultural attainment with quantitative evidence for any population is difficult. Typically, dichotomous variables such as reported occupation and demonstrated ability to sign one’s name are used to approximate relative knowledge and skill levels. Considerable quantitative evidence on the literacy and occupation of adult male German immigrants can be gathered from passenger manifests. As Bailyn has shown, “Literacy is a basic index of cultural attainment.” In many cases, it was a prerequisite for subsequent occupational, religious, and political education and advancement. An accepted measure of literacy is the ability to affix a signature, which is the measure used here.14 From 1727 to 1775, male German immigrants above age 15 were required to sign loyalty oaths upon landing in Philadelphia. The resulting 24,025 signatures are the single largest and least biased body of literacy evidence for colonial America. A three-­year moving average of the literacy percentage is presented in Figure 5.1. In total, 71 percent could sign their name. In relation to other European states, this literacy rate placed immigrants from Germany second only to the population of Calvinist Scotland and on par with the population of the Netherlands. There can be little doubt that German immigrants contributed to the general superiority of colonial literacy compared with literacy levels in Europe. In addition, German immigrants were at least as literate as the established rural colonial population, but slightly less literate than the urban colonial population.15 The literacy rate of German immigrants rose from 60 percent in the 1730s, to 70 percent in the early 1750s, and finally to 80 percent after 1760. By the time of the American Revolution, German immigrants were approaching universal literacy. Because the age structure of adult German immigrants did not change, the rise in literacy represented secular gains in education. Rising literacy was coincident with the rising proportion of single adults among the immigrants, but a direct causal relationship between these two trends cannot be determined. Given the exceptional level of literacy among German immigrants, why did many contemporary observers and subsequent scholars claim that German

Percentage literate

Age, occupation, and family composition   97 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Three-year moving average weighted by the number of observations Men age 16 and older Number of observations � 24,025

1730 1735 1740 1745 1750 1755 1760 1765 1770 1775 Year of arrival

Figure 5.1 Adult male literacy among German immigrants to Pennsylvania (source: Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, Pa., 1934), I).

colonists were ignorant, or stupid, or had little interest in education? Although some observations were biased by prejudice, mistrust, or a misunderstanding of an alien culture, other observations accurately reflected the difficulties Germans experienced after settling in the New World. For instance, the high cost of providing education on the frontier may have restrained the ability of German immigrants to transfer their high level of literacy to their children. The eighteenth-­century Pennsylvania-­German community, for example, was keenly aware of this problem.16

Occupational composition The occupational composition of German immigrants has been difficult to determine. Professions followed in America may be poor indicators of trades practiced before leaving Europe. For example, Christopher Sauer left Wittgenstein in 1724 as a tailor. After arriving in America he was, at one time or another, a  farmer, clockmaker, cabinetmaker, apothecary, and joiner before finally becoming a printer. Some German passenger lists recorded the occupations of adult males (see Table 5.5). This evidence should reflect the occupa-

Farmer Tailor Butcher Baker Unlisted Merchant Weaver Shoemaker Cooper Carpenter Clerk Blacksmith Miller Accountant Smith Gardener Distiller

The most frequently listed occupations

N = 2,928 62.77% 3.38 0.99 2.66 0.00 0.00 0.24 2.32 0.00 4.23 0.00 1.64 1.54 0.00 1.20 0.00 0.00

N = 707 12.02% 2.97 0.99 2.26 0.00 1.98 4.53 3.68 1.41 3.54 1.98 1.27 0.28 0.00 0.71 2.12 0.42

N = 579 27.34% 6.74 5.62 4.49 10.49 0.94 5.81 4.12 1.69 3.37 2.43 1.69 2.06 0.00 1.12 1.12 0.94

N = 647 21.33% 9.58 7.88 7.26 5.72 4.49 4.33 3.55 2.16 2.01 1.85 1.70 1.39 1.24 1.24 1.24 1.24

N = 61 39.34% 1.64 1.64 0.00 16.39 0.00 13.11 4.92 1.64 4.92 0.00 0.00 1.64 0.00 3.28 0.00 0.00

1774–6

1709

1793–1807

1815–20

1733

English immigrants

German immigrants

N = 47,073 43.22% 4.25 >0.77 >0.77 >0.77 3.10 3.82 1.43 1.63 4.58 >0.77 3.51 2.49 >0.77 1.12 >0.77 0.95

Rural census 1800

N = 3,350 0.81% 4.69 1.61 3.10 0.00 10.09 1.37 4.42 2.96 6.03 0.96 2.63 0.03 0.00 – – 1.10

Philadelphia tax lists 1773

Resident Pennsylvanians

Table 5.5 Occupational distribution of adult male German and English immigrants arriving in Pennsylvania, 1709–1820

14.37 100.01% 77

1.24 1.08 1.08 0.93 0.93 0.93 0.93 0.31 14.04 100.00% 67

1.12 0.75 0.75 0.00 1.31 0.00 0.37 1.69 4.92 100.00% 14

0.00 0.00 1.64 1.64 0.00 0.00 0.00 3.28 1.53 99.98% 17

0.20 0.07 0.44 0.00 16.29 0.00 0.48 0.00 31.42 100.00% 106

0.28 0.42 0.42 0.00 0.71 25.46 0.42 0.71

>0.77 0.84 >0.77 0.77 1.50 9.14 >0.77 >0.77 9.95 100.00% 970

47.48 100.00% 143

0.09 1.01 0.72 0.00 0.96 7.55 0.39 2.00

Notes See the notes for Table 5.2. In their sample of rural Pennsylvanian occupations, Soltow and Keller did not report the number of observations for occupations with less than 362 individuals. Therefore, occupations not reported by Soltow and Keller must have been greater than zero but less than 0.77 percent of all occupations. The occupations of smith and gardener were not reported separately in the Philadelphia tax evidence.

sources: See Table 5.1; Lee Soltow and Kenneth W. Keller, “Rural Pennsylvania in 1800: A Portrait From the Septennial Census,” Pennsylvania History, XLIX (1982), 36; Gary B. Nash, The Urban Crucible (Cambridge, Mass., 1979), 387–91.

Number of occupations reported

Others

Turner Doctor Saddler Wagonmaker Mason Laborer Tanner Joiner

100   German immigration to America, 1709–1835 tions practiced prior to emigration. However, because the data are concentrated in the 1709 and post-­1793 samples, the projected trend in occupational composition across the middle of the eighteenth century should be regarded as tentative.17 The proportion of farmers declined significantly over the century, falling from 63 percent in 1709 to 21 percent in 1820. German immigrants may not have been predominantly farmers as many scholars have claimed. Farming may have been pursued after immigration because of favorable opportunities in America and not because of past experience. The prospect of German colonization in Eastern Europe after 1760 may have proven especially attractive to farmers, thus leaving relatively fewer farmers among immigrants to America. Because farmers were traditionally less literate and more likely to be married, a declining proportion of farmers is consistent with rising literacy and the increasing proportion of single adults among German immigrants to America after 1760.18 Other skilled occupations replaced farmers, most notably the three retail trades of tailor, butcher, and baker. By 1815–20, these three trades represented 25 percent of the occupations of German immigrants, more than the number of farmers. Significant gains were also made in professional occupations. The proportion of merchants, clerks, accountants, and doctors rose from 0 to 9 percent between 1709 and 1815–20. However, the proportion in building crafts, such occupations as masons, carpenters and joiners, declined from about 21 percent of the occupations in 1709 to under 7 percent after the Revolution. Finally, except for weavers, the proportion in industrial crafts increased slightly, especially among coopers, turners, distillers, and the numerous but infrequent occupations grouped in the “other” category.19 Table 5.5 also provides the occupational composition of English immigrants in 1774–6, Philadelphia residents in 1773, and rural residents of Pennsylvania in 1800. The English sample has a significantly lower proportion of farmers, over twice as low as the Germans. However, if the categories of farmer, laborer, and unlisted are combined into an unskilled category, then the proportions of English and German unskilled are similar. Although German colonists had a similar proportion of agricultural workers, they represented a higher proportion of established farmers relative to agricultural laborers compared with English colonists. Proportionally, more laborers were among English immigrants in the 1770s (25 percent), Philadelphians in 1773 (8 percent), and rural Pennsylvanians in 1800 (9 percent) than among German immigrants in any individual sample between 1709 and 1820 reported here (less than 1 percent).20 Compared with German immigrants arriving between 1793 and 1820, English immigrants in the 1770s had approximately half the proportion of retail craftsmen (such as tailors, bakers, and butchers) and twice the proportion of the “others” category (which, for the English, was largely composed of low-­skilled occupations such as footman, coachman, groom, and so forth). Finally, the proportion of industrial and building craftsmen was roughly similar between the two immigrant groups, most notably among weavers, carpenters, blacksmiths, shoemakers, and coopers. The lower proportion of laborers and workers in low-­skilled trades

Age, occupation, and family composition   101 among German immigrants indicates that, outside the farmer category, Germans were more skilled than English immigrants. The higher proportion of farmers and members of highly skilled occupations is consistent with the older age distribution of adult German immigrants relative to English immigrants. Compared with the occupations of rural Pennsylvanians in 1800, German immigrants had a lower proportion of farmers and laborers, 16 and 9 percentage points lower, and a higher proportion in the retail trades, particularly among tailors, butchers, bakers, and saddlers. Overall, they had similar proportions of industrial and building craftsmen and professionals. However, within these categories, Germans had proportionally more clerks, weavers, shoemakers, distillers, and turners, fewer merchants, carpenters, and blacksmiths, and similar percentages of doctors, coopers, smiths, millers, masons, tanners, and joiners. By the end of the eighteenth century, German immigrants were apparently more skilled than the American population that they entered. Relative to the American population, German immigrants cannot be characterized as largely peasant farmers, but, relative to English immigrants, they were more agricultural. Among the skilled occupations, German immigrants possessed a distribution of trades which differed significantly from that of the Pennsylvania population. The occupational composition of German immigrants is not representative of the southern Rhineland, the source of most German immigrants. For example, the occupational distribution in Baden was over 50 percent agricultural as late as 1810. This comparison suggests that German immigrants were drawn disproportionately from the nonagricultural classes. The declining proportion of farmers and rising proportion of retail tradesmen among German immigrants reflected changing economic opportunities among occupational classes, rather than an altered occupational distribution, in the Rhineland during the eighteenth and early nineteenth centuries.21 Quantitative evidence from ship passenger lists offers an alternative and less biased estimate of the social composition of German immigrants between 1709 and 1820 than has been previously gathered from literary sources and quantitative evidence based on the postmigration experiences of German immigrants. The composition changed substantially over the century. The occupational mix shifted away from farmers and building crafts toward retail trades, industrial crafts, and the professions. There was little change in the adult age distribution. The educational and occupational skills of German immigrants were compar­able if not superior to those of their English and American counterparts. They were over 80 percent literate by the 1770s. German immigration injected greater numbers of women and children and proportionally more skilled workmen into the American labor market than British immigration, and, because adult Germans were older, the skill and work experience which they possessed per occupation was greater than that of British immigrants. They complemented rather than duplicated the British and American occupational mix, and thus added diversity to the American marketplace. Family composition also changed substantially. Ironically, as German immigration became less family oriented (by shifting toward single adult males),

102   German immigration to America, 1709–1835 family size grew, primarily through reductions in the proportion of childless couples. Relative economic opportunities for single adults versus married persons and for farmers versus tradesmen changed substantially in Germany compared with America during the eighteenth century. However, these trends were small relative to German differences with English immigrants, who had a much higher proportion of single adult males. The family composition of German immigration more closely resembled English emigration to the colonial periphery, such as to Nova Scotia, or seventeenth-­century Puritan immigration, than eighteenth-­century British immigration to the mid-­Atlantic and Chesapeake regions.22 Compared with British immigrants, the greater family orientation of German immigrants implies a greater contribution to subsequent native population growth per immigrant. The higher proportion of married couples, particularly newly married childless couples, among German immigrants injected relatively more families into the American population who were just beginning their fertility cycle. In addition, the relatively balanced sex ratio may have increased the frequency of marriage, particularly within the Pennsylvania-­German community. Independent of language and custom, the greater family orientation also implies more social and community cohesion and slower assimilation into American society than for other immigrants. Roughly half of all of the immigrants entered servitude upon arrival in Pennsylvania. Many Germans entered servitude as intact families, whereas virtually all British immigrants entered servitude as single adults. Because servants were not allowed to marry during their contract, the impact on subsequent population growth, ethnic assimilation, and community cohesion caused by the relative greater family orientation of German immigrants was accentuated by the prevalence of servitude before 1800. Finally, the decline of servitude and the rise in the proportion of immigrants in the retail trades after 1800 led to greater urban concentrations of German immigrants and, coupled with the lower migration totals and lower proportion of families among German immigrants between 1785 and 1815, helps to explain the accelerated assimilation of Germans into the American mainstream after the American Revolution.23

Notes   * Originally published as Farley Grubb, “German immigration to Pennsylvania, 1709–1820,” Journal of Interdisciplinary History, vol. 20, no. 3 (Winter 1987): 417–36. Copyright 1990. © 1990 by the Massachusetts Institute of Technology and the editors of The Journal of Interdisciplinary History. Reprinted with permission. Farley Grubb is Associate Professor of Economics at the University of Delaware. The author gratefully acknowledges the helpful comments of David W. Galenson, Henry Gemery, and Robert Selig as well as financial support from the University of Delaware Economics Department for the purchase of microfilmed manuscripts.   1 See Frank R. Diffenderffer, “The German Immigration into Pennsylvania Through the Port of Philadelphia, and ‘The Redemptioners,’ “Pennsylvania German Society, XX (1899), 97–106; Henry E. Jacobs, “The German Emigration to America 1709–1740,” Pennsylvania German Society, VIII (1897), 148. Sizable German colonies could also

Age, occupation, and family composition   103 be found elsewhere in America, most notably in Georgia and the Carolinas. The contribution of German immigrants to the population growth of the Delaware Valley was similar to the contribution of all immigrants to total U.S. population growth during the great migration wave of 1840 to 1860 (see table 5.1); United States, Department of Commerce, Bureau of the Census, Historical Statistics of the United States Colonial Times to 1970 (Washington, D.C., 1975), 1,168; Jonathan Hughes, American Economic History (Glenview, Ill., 1983), 115. The estimated contribution of total net migration to population growth was smaller, 7 to 9 percent for 1840–60 and 1 to 3 percent for 1800–40, making the earlier impact of German immigration look even more impressive. See Lance Davis, et al., American Economic Growth (New York, 1972), 123.   2 See the discussions in Lutz K. Berkner, “Inheritance, Land Tenure and Peasant Family Structure: A German Regional Comparison,” in Jack Goody, Joan Thirsk, and Edward P. Thompson (eds.), Family and Inheritance (New York, 1976), 71–95; Carl Bridenbaugh, Cities in Revolt (London, 1955); Hans Fenske, “International Migration: Germany in the Eighteenth Century,” Central European History, XIII (1980), 332–47; James A. Henretta, “Families and Farms: Mentalité in Pre-­Industrial America,” William and Mary Quarterly, XXXV (1978), 3–32; Richard Hofstadter, America at 1750 (New York, 1971); James O. Knauss, Jr., “Social Conditions among the Pennsylvania Germans in the Eighteenth Century,” Pennsylvania German Society, XXIX (1918), 1–217; James T. Lemon, “Comment on James A. Henretta’s Families and Farms: Mentalité in Pre-­Industrial America,” William and Mary Quarterly, XXXVII (1980), 688–700; Hans Medick and David Warren Sabean, “Interest and Emotion in Family and Kinship Studies: A Critique of Social History and Anthropology,” in Medick and Sabean (eds.), Interest and Emotion (New York, 1984), 9–27; A. Gregg Roeber, “In German Ways? Problems and Potentials of Eighteenth-­Century German Social and Emigration History,” William and Mary Quarterly, XLIV (1987), 750–74; idem, “The Origins and Transfer of German-­American Concepts of Property and Inheritance,” Perspectives in American History, III (1987), 115–71; Sabean, “Aspects of Kinship Behavior and Property in Rural Western Europe Before 1800,” in Goody, Thirsk, and Thompson (eds.), Family Inheritance, 96–111; Jared Sparks (ed.), The Works of Benjamin Franklin (Boston, 1838), VII, 71–3; Frank Trommler and Joseph McVeigh (eds.), America and the Germans (Philadelphia, 1985), I; Stephanie Grauman Wolf, Urban Village (Princeton, 1976); Louis B. Wright, The Cultural Life of the American Colonies, 1607–1763 (New York, 1957).   3 Henry M. Muhlenberg, The Journals of Henry Melchior Muhlenberg (Philadelphia, 1942), II, 423–4. See also Bridenbaugh, Cities in Revolt, 333; Donald F. Durnbaugh, “Christopher Sauer Pennsylvania-­German Printer,” Pennsylvania Magazine of History  and Biography, LXXXIII (1959), 316–25; Robert H. Billigmeier and Fred A. Picard (ed.), The Old Land and the New (Minneapolis, 1965), 35–147. The construction of a large, representative sample linking the German emigration records extracted by Werner Hacker with American records on German immigration still waits to be done.   4 The 19,072 immigrants in the evidence for 1785–1820 are particularly interesting because of the general lack of immigration statistics and quantitative studies of immigration for that period. See Abbot E. Smith, Colonists in Bondage (New York, 1947), 320; Henry Gemery, “European Emigration to North America, 1700–1820: Numbers and Quasi-­Numbers,” Perspectives in American History, I (1984), 283–342. The 1727–1808 passenger lists are in Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, Pa., 1934), I, III; the 1800–1820 lists are in “Philadelphia Custom House Records, 1800–83” (ms. held at the National Archives, Washington, D.C., microfilm # 425). Analysis of the yearly flow of German immigrants based on these lists is presented for 1727–75 in Marianne S. Wokeck, “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History

104   German immigration to America, 1709–1835 and Biography, CV (1981), 249–78, and for 1785–1820 in Grubb, “The Disappearance of European Immigrant Servitude in the United States: An Economic Analysis of Market Collapse, 1785–1835,” unpub. ms. (Univ. of Delaware, 1988). After 1820, the social characteristics of immigrants can be gathered from U.S. State Department passenger lists. See William J. Bromwell, History of Immigration to the United States (New York, 1856); Grubb, “The Reliability of U.S. Federal Immigration Statistics: The Case of Philadelphia, 1815–1830,” unpub. ms. (Univ. of Delaware, 1989).   5 For discussions of the 1709 migration, see Diffenderffer, “The German Exodus to England in 1709,” Pennsylvania German Society, VII (1897), 255–414; Walter A. Knittle, Early Eighteenth Century Palatine Emigration (Philadelphia, 1937); Jacobs, “German Emigration,” 31–134.   6 The English evidence has been extensively analyzed in Bernard Bailyn, Voyagers to the West (New York, 1986). This evidence was reorganized here to make it comparable to the German evidence.   7 Although some ship lists separated passengers into family groups, many ship lists recorded men, women, and children separately, with no indication of family membership. Social composition was estimated by matching names, ages, and the relative positioning of names recorded in the subcategories of each ship list. Table 5.2 may overstate the proportion of single migrants because some of the unmarried adults, although making independent migration decisions, may have been traveling with other relatives, such as parents, grandparents, or siblings.   8 See Sharon V. Salinger, “Artisans, Journeymen, and the Transformation of Labor in Late Eighteenth-­Century Philadelphia,” William and Mary Quarterly, XL (1983), 62–84; Grubb, “Servant Auction Records and Immigration into the Delaware Valley, 1745–1831: The Proportion of Females Among Immigrant Servants,” Proceedings of the American Philosophical Society, CXXXIII (1989), 154–69. Differences between the ethnic sex ratios were greater among emigrant servants. See David W. Galenson, White Servitude in Colonial America (Cambridge, Mass., 1981), 23–5; Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, XXII (1985), 316–39.   9 The samples in table 5.3 are smaller than in table 5.2 because passenger lists that did not record enough information to group children and parents into separate family units were eliminated. Counterfactual analysis is a controlled simulation exercise. For example, by replacing the family size estimated in the 1730s evidence with the family sizes estimated in the 1709 evidence, 1.5 is replaced by 2.76, the number of single adults and married persons in the 1730s sample are held constant, but the number of children in the 1730s sample is hypothetically increased because the number of children per married couple was hypothetically increased. The result is a different set of hypothetical proportions of single adults, married persons, and children among the 1730s immigrants and consequently a hypothetically different dependency ratio. These counterfactual estimates can be compared to the real estimates to determine the percentage change in the dependency ratio from 1709 to the 1730s caused by the change in family size between the two dates. 10 These counterfactual estimates hold the number of married persons per sample constant and show how the number of children per couple would change if the proportion of childless couples remained the same in subsequent samples. 11 On the influence of inheritance systems, see Lutz K. Berkner, “Peasant Household Organization and Demographic Change in Lower Saxony (1689–1766),” in Ronald Demos Lee (ed.), Population Patterns in the Past (New York, 1977), 53–69; Berkner and Franklin F. Mendels, “Inheritance Systems, Family Structure, and Demographic Patterns in Western Europe, 1700–1900,” in Charles Tilly (ed.), Historical Studies of Changing Fertility (Princeton, 1978), 209–23. Migration to America was less controlled by state and corporate businesses than colonization schemes involving Eastern Europe. See Fenske, “International Migration,” 332–47. The migration–marriage

Age, occupation, and family composition   105 connection in Eastern colonization was suggested by Robert Selig in a personal communication based on Selig, Rautige Schafe und Geizige Hirten: Studien zur Auswanderung aus dem Hochstift Wurzburg im 18. Jahrhundert und ihre Ursachen (Wurzburg, 1988); J. Michael Phayer, “Germans Emigrate to a French Colony—An Eighteenth Century Account,” Societas, VIII (1978), 243–8; Selig, “Mangy Sheep and Greedy Shepherds: Emigration from the Prince-­Bishopric of Wuerzburg in the Year 1764,” Proceedings of the Eleventh European Studies Conference, XI (1986), 244. 12 See Wolfang Kollman and Peter Marschalck, “German Emigration to the United States,” Perspectives in American History, VII (1973), 529–44; Charlotte Erickson, “Emigration from the British Isles to the U.S.A. in 1831,” Population Studies, XXXV (1981), 183–5. 13 The age pattern among English immigrants to Pennsylvania was similar to the pattern among all English indentured immigrants. See Galenson, White Servitude, 26–33. 14 Bernard Bailyn, Education in the Forming of American Society (Chapel Hill, 1960), 83–4. See the discussions in Roger S. Schofield, “The Measurement of Literacy in Pre-­Industrial England,” in Goody (ed.), Literacy in Traditional Societies (Cambridge, 1968), 311–25; Grubb, “The Growth of Literacy in Colonial America: Longitudinal Patterns and Economic Models,” unpub. ms. (Univ. of Delaware, 1989). 15 Idem, “Colonial Immigrant Literacy: An Economic Analysis of Pennsylvania-­German Evidence, 1727–1775,” Explorations in Economic History, XXIV (1987), 63–7. 16 See Muhlenberg, Journals, I, 140; Samuel E. Weber, The Charity School Movement in Colonial America (Philadelphia, 1905), 17; Grubb, “Growth of Literacy.” 17 Durnbaugh, “Christopher Sauer,” 316–25. For an early nineteenth-­century example, see Billigmeier and Picard (eds.), Old Land and New, 35–147. The passenger lists were not used for taxation, employment, or other purposes that might cause immigrants or shippers to misrepresent occupations. 18 For example, the majority of eastern emigrants from the Prince-­Bishopric of Wurzburg in 1764 were farmers. See Selig, “Mangy Sheep and Greedy Shepherds,” 245. 19 The decade prior to the American Revolution may have seen an influx of German miners. Muhlenberg reported that over 100 miners debarked in September 1764 at Philadelphia, or about 10 percent of the adult male Germans arriving in that year. See Muhlenberg, Journals, II, 121. 20 Even if German immigrants with unlisted occupations were laborers, the number would still be relatively small. The unlisted category included a number of cabin passengers who were not asked to report an occupation. Gentlemen were more likely to be cabin passengers than laborers. 21 See Selig, “Mangy Sheep and Greedy Shepherds,” 238–49; Wolfram Fischer, Der Staat und der Anfang der Industrialisierung in Baden, 1800–1850 (Berlin, 1962), 277–97; Mary Jo Maynes, “Schooling the Masses: A Comparative Social History of Education in France and Germany, 1750–1850,” unpub. Ph.D. diss. (Univ. of Michigan, 1977), 31. 22 See Bailyn, Voyagers to the West, 89–232, 361–604. For a discussion of the determinants of family size in colonial America, see Daniel Levy, “The Economic Determinants of Family Size in the Colonial Chesapeake: Evidence from the Colonial Legislators of Maryland,” unpub. Ph.D. diss. (Univ. of Chicago, 1989). 23 See Grubb, “Incidence of Servitude,” 316–39; idem, “British Immigration to Philadelphia: The Reconstruction of Ship Passenger Lists from May 1772 to October 1773,” Pennsylvania History, LV (1988), 118–41; idem, “The Auction of Redemptioner Servants, Philadelphia, 1771–1804: An Economic Analysis,” Journal of Economic History, XLVIII (1988), 595–8.

6 Literacy Longitudinal patterns and market forces*

Literacy underwent revolutionary growth in northwestern Europe during the seventeenth and eighteenth centuries. This revolution coincided with other dramatic changes in European society, such as the industrial, demographic, agricultural, political, and religious revolutions (Deane 1969: 20–84). While the relationships between literacy and these other revolutions are not fully understood, their association is apparent and many potential influences exist (Cipolla 1969; Cremin 1970; Graff 1981: 232–60; 1987a, 1987b; Jensen 1986: 114–28; Maynes 1985: 117–31; Mitch 1984, 1988; Sanderson 1983; West 1978). The transplantation of European society across the Atlantic brought the literacy revolution to the American periphery. While numerous studies have shown that colonial America participated in this expansion of literacy, the common longitudinal patterns of literacy growth across the various regions and populations of colonial America have received less attention.1 The accumulation of literacy studies over the last 30 years has made possible the examination of common patterns of growth in colonial literacy, as well as systematic comparisons with European literacy. The evidence from these studies is organized here to investigate regional similarities in the growth paths of colonial literacy and to explain how conditions and constraints peculiar to colonial America altered these growth paths from that of north-­western Europe. In particular, the evidence is used to explore the general pattern of temporary declines in colonial literacy which followed initial European settlement, a pattern called here “creolean degeneracy.” Implicit economic models have been frequently used to explain the growth in literacy. However, a consensus on the dominant explanation for the growth of colonial literacy has not emerged. These implicit models are made explicit here and used to analyze the common patterns of growth in colonial literacy, in particular the nonlinear pattern of creolean degeneracy found in the evidence. By making the models explicit, the implications of alternative economic explanations are made apparent. These explicit models are used to suggest directions for future research on colonial literacy that may resolve the debate over competing causes of the peculiar growth paths of colonial literacy, including the creolean degeneracy in literacy.

Literacy   107

Literacy estimates Numerous studies have measured literacy in selected areas of colonial America and Europe during the seventeenth and eighteenth centuries. The usual method of quantifying literacy is to measure the percentage of people in a population who could sign their names. Signature literacy is the only universal, standard, and direct measure available for the colonial period which provides a substantial quantity of evidence. Although a dichotomous measure, the ability to affix a signature is believed to correspond to a middle range of literacy skills or roughly to the ability to read fluently (Craig 1981; Cressy 1980: 1–19, 42–62; Lockridge 1974: 7; Schofield 1968; Soltow and Stevens 1981: 3–6).2 Tables 6.1 and 6.2 present the evidence from these studies for adult male literacy, organized by country or colony, time period, and urban versus rural location.3 Because different sources of evidence possess different potential biases and because sample sizes vary, comparisons of the patterns of literacy across these studies cannot be exact. For example, signature evidence gathered from wills, deeds, petitions, or jury lists is thought to overstate a population’s literacy, possibly by different degrees.4 However, the similarity of the evidence among the colonial studies, particularly over the growth paths of individual colonial subregions, suggests that comparing the longitudinal patterns across colonial regions may be done with some confidence. If the cross-­sectional biases are relatively constant over time, then the longitudinal pattern will be relatively unbiased; that is, the slope and change in slope of the estimated pattern over time will be the same as the true underlying pattern in the population. Given the above caveats, several observations can be made about the common patterns of literacy growth. First, differences in urban versus rural literacy were significant throughout Europe and the colonies. Urban areas were systematically more literate, though rural areas were beginning to close the literacy gap by the end of the eighteenth century. The differential was 10 to 25 percentage points. A positive correlation between population density and literacy has been noted or measured in other studies (Graff 1987a, 1987b; Lockridge 1974; Soltow and Stevens 1981). Second, the level of literacy was higher in the colonies than in Europe.5 Of all European countries perhaps only Scotland surpassed America in literacy by 1800. Not only had the European literacy revolution been transplanted to the American periphery during the colonial period, but colonial literacy had somehow leaped past that of northwestern Europe. Not only was the United States born wealthy (Jones 1980), it was born literate. Nevertheless, while the levels of colonial literacy were relatively high, the long-­run trend appears to be slightly steeper in northwestern Europe, both in urban and rural areas. European literacy was catching up to American literacy by the nineteenth century. Finally, while literacy levels did vary across the colonies, high levels of literacy were distributed much more evenly across the colonies than across Europe. High levels of literacy in the colonies were not confined to New England, nor were they just a product of Puritan motives or produced selectively by New

108   German immigration to America, 1709–1835 Table 6.1 Studies of colonial adult male literacy, 1650–1840 Location, years I. rural areas New England 1650–70 1705–15 1758–62 1799–1829 Windsor Dist., nh 1755–1830 Pennsylvania Chester County 1729–44 1745–54 1755–64 1765–74 1840 Lancaster County 1729–44 1745–54 1755–64 1765–74 1729–64 1765–74

Percentage literate

Evidence source

55.0 64.0 78.0 75.0

500 750 1,000 395

91.3

10,289

All records

74.3 72.6 69.4 73.4 >99.0

253 285 288 338 57,515

Wills Wills Wills Wills Census

63.4 60.1 64.6 63.4 60.0 66.3

93 168 260 388 80 95

Wills Wills Wills Wills Wills in German Wills in German

574

Army enlistees

New York, New Jersey, Pennsylvania 1799–1829 60.0 Maryland Baltimore County 1768 Virginia 1600s 1600s 1705–62 1762–7

Sample size

80.0

2,030

54.0 60.0 66.0 68.0

2,165 12,445 467 648

North Carolina Perquimans County 1661–95 1696–1721 1722–47 1748–76

67.0 57.0 64.0 79.0

51 187 306 435

South Carolina 1729–65

80.0

II. urban areas New England Boston 1650–70 1705–10 1758–62

77.0 74.0 82.0

Wills Wills Wills Army enlistees

Petitions Jury lists Deeds, depositions Wills Wills

Deeds Deeds Deeds Deeds –

100 200 150

Deeds, wills

Wills Wills Wills

Literacy   109 Location, years Grafton 1747

Percentage literate

Sample size

Evidence source

93.0

95

All records

74.8 92.6

254 95

All records All records

90.9 83.7 87.3

44 766 110

Wills Petitions Wills

79.0

360

–

Pennsylvania Philadelphia 1699–1706 1773–5 1798–1840

80.0 81.6 72.0

40 98 930

Wills Wills Merchant seamen

Virginia Elizabeth City 1693–9 1763–71

75.0 91.0

190 177

Deeds, depositions All records

New York New York City (Dutch) 1675–98 1706–38 New York City 1699–1703 1701–2 1760–75 Albany 1654–1775

III. point of entry (immigrants) English servants 1683–4 41.2 1718–59 70.3 German immigrants Radical Protestants 1727–45 Other Protestants 1727–45 1746–56 1761–75 French immigrants 1632–63 1680–99

590 2,693

Contract register Contract register

71.3

334

Loyalty oaths

63.4 72.1 80.0

6,600 12,525 4,536

Loyalty oaths Loyalty oaths Loyalty oaths

62.2 44.1

– –

– Marriage register

Sources: Beales 1978: 98; Cremin 1970: 533–43; Galenson 1979: 79; Gallman 1988: 574; Gilmore 1989: 120; Grubb 1987; Jensen 1986: 225; Kilpatrick 1912: 197, 228–9; Lockridge 1974: 24, 77; Moogk 1989: 467; Soltow and Stevens 1981: 37–8, 50–2; Strassburger 1934; Tully 1972: 304. Notes Radical German Protestants were ships with Dunkers, Schwenkfelders, and Moravians present. Other German Protestants were Lutheran and Dutch Reformed. The German immigrant evidence is for Pennsylvania arrivals, and the French immigrant evidence is for Canadian arrivals.

110   German immigration to America, 1709–1835 Table 6.2 Studies of European adult male literacy, 1633–1840 Location, years I. countries England 1641–2 Suffolk County 1696 1754–1800 Nine parishes 1754–79 1780–1809 1810–40

Percentage literate

Sample size

Evidence source

30

9,428

Loyalty oaths

47 62

10,056 –

Loyalty oaths Marriage register

57 57 58

1,896 2,543 4,596

Marriage register Marriage register Marriage register

Scotland 1638–43 1650–60 1690–1710 1740–50 1800

36 44 76 78 90

5,897 152 952 233 –

Loyalty oaths Depositions Depositions Depositions –

Rural France 1686–90 1740–59 1760–89 1790–1809

19 36 42 45

– – – –

Marriage register Marriage register Marriage register Marriage register

Continental Europe Protestant 1700 Catholic 1700

40–45

–

–

20–30

–

–

East Prussia 1750 1765 1800

10 25 40

– – –

– – –

Rural Belgium 1778–92

60

Sweden and Finland 1800

80

–

–

Italian Piedmont 1800

45

–

–

Netherlands 1800

70

–

–

Geneva countryside 1800

60

–

Marriage register

Germanic areas 1800

55–60

–

–

Castile, Spain 1800

10–15

–

–

Literacy   111 Location, years

Percentage literate

Sample size

Evidence source

II. towns and cities England Norwich 1633–7 London 1641–4 Chester, Bridgewater 1642 Seven towns 1754–62 1799–1804

67 70

– –

Marriage register Marriage register

Netherlands Amsterdam 1630 1680 1730 1780

57 70 76 85

– – – –

Marriage register Marriage register Marriage register Marriage register

Urban France 1740–59 1760–89 1790–1809

59 66 73

– – –

Marriage register Marriage register Marriage register

44

494

Wills

78

609

Loyalty oaths

50

1,728

Loyalty oaths

Germany Schriesheim 1800 Koblenz 1800

80

–

Bürger list

86

–

Marriage register

Geneva 1800

90

–

Marriage register

Brussels 1820

57

–

–

Sources: Cipolla 1969: 20–2, 61–4; Craig 1981: 170; Cressy 1980: 72–4, 100–7, 181; Gawthrop and Strauss 1984: 53; Graff 1987b: 195, 223; Houston 1982: 86–90; Maynes 1985: 15–18, 126; Schofield 1973: 445; 1989; Stone 1969: 100–4. Notes The French estimates are averages for 76 departments. In the English estimates, the reconstituted nine parishes were Ash, Austrey, Bottesford, Bridford, Dawlish, Gedling, Morchard Bishop, Odiham, and Shepshed, and the seven towns were an unweighted average of the estimates for Penzance, Oxford, Northampton, Nottingham, King’s Lynn, Bristol, and Halifax.

England schooling legislation (Graff 1987b: 163–72, 249–57; Lockridge 1974: 72–101). Nor were they confined to English-­speaking populations, as can be seen from the Pennsylvania German evidence. The levels and patterns of growth in literacy were similar across the rural areas of New England, Pennsylvania, Virginia, and North Carolina, and similar across the urban areas of Boston, New York, Philadelphia, and Elizabeth City, Virginia. Mass literacy (defined here as over 70 percent literate) was achieved early in urban areas and increased little

112   German immigration to America, 1709–1835 through the eighteenth century. Literacy in rural areas increased until mass literacy was achieved by the middle of the eighteenth century.

Causes of high initial levels of colonial literacy What caused the high initial level and long-­run growth of colonial literacy is still a matter of debate. There are two potential sources of high colonial literacy: above-­average literacy of immigrants and above-­average transmission of literacy to American-­born colonists. Most literacy studies do not distinguish between European and American-­born colonists, and there are no studies which measure literacy only for the latter. How the education of American-­born colonists influenced overall colonial literacy has not been disentangled from the influence of immigrants on colonial literacy. The history of colonial literacy would be enhanced if future studies distinguished European from American-­born colonists in the evidence. Direct evidence on the literacy of European immigrants suggests that immigration was an important force in shaping colonial literacy patterns. The literacy of three immigrant populations, English indentured servants, German colonists arriving in Pennsylvania, and French-­Canadian immigrants, has been measured at the point of entry into the colonies. These immigrant groups were more literate than their European home populations and were comparable in literacy to the colonial populations they entered (see Tables 6.1 and 6.2). English indentured servants went predominantly to the Chesapeake region (Galenson 1981: 220–3). Their 70 percent literacy rate compares well with the Virginia rate of 68 percent. German immigrants to Pennsylvania were 70 percent to 80 percent literate between 1750 and 1775, whereas German residents of Lancaster County, Pennsylvania, were 66 percent literate.6 While limited in coverage, these three studies of immigrant literacy support the view that immigration per se may have been one of the important causes of high initial levels of colonial literacy, because of the disproportionate propensity of literate Europeans to migrate to America (see also Graff 1987a, 1987b: 163, 170, 251; Grubb 1987; Lockridge 1974: 5, 46–8, 74–5, 78, 82; 1981: 190, 193). Explaining the long-­run trend in colonial literacy, however, requires incorporating the relative influence of American-­born colonists. Over time, the proportion of American-­born in the colonial population rose, and thus the evidence used to measure colonial literacy would be increasingly dominated by the literacy of this group. Variation in the timing and rate of increase in the proportion of American-­born in the colonial population might explain regional differences in long-­run literacy patterns. How the literacy of immigrants compared with or influenced that of their American-­born offspring is unclear. Few studies measure the extent to which literacy was transmitted from immigrant parents to their American-­born children during the seventeenth and eighteenth centuries. The transmission of literacy could occur through two avenues: literate parents could teach their children to read and write, or parents, regardless of their own literacy, could provide their

Literacy   113 children with access to instruction by others, such as at a local church or formal school or at work, through the child’s servant or apprentice contract. Two studies have measured the literacy both of parents and of their children, among German immigrants to Pennsylvania (Grubb 1987: 74) and among families in Shepshed parish, England (Levine 1979). These studies find a positive and statistically significant correlation between the literacy of parents and the literacy of their offspring. However, the correlation is less than 1.7 Some children of literate parents failed to learn to read and write, and many children of illiterate parents nevertheless learned to read and write. These findings suggest that if the intergenerational transmission of literacy depended solely on literate parents teaching their children to read and write, or if providing instruction outside the family, whether at church, school, or work, became more difficult, as during a migration to the colonial frontier, then the level or growth of literacy in subsequent generations might degenerate. Measuring the ability to affix a signature or measuring how this ability correlated between parents and their children does not reveal whether that ability was learned at home, at work, at church, or at a formal school. Measuring signature literacy reveals little about the methods available or the methods actually used by parents to transmit literacy to their children. Nor do these measurements reveal whether the social and economic circumstances which enabled parents to acquire literacy were still present when their children were of educable age. In a world where formal schooling was not always available, the transmission of literacy from immigrant parents to their American-­born offspring may have been difficult and certainly was less than automatic. The history of colonial literacy would be greatly enhanced if future research were directed towards the quantitative significance of the different mechanisms of literacy transmission.

The nonlinear trend in regional colonial literacy The long-­run growth of colonial literacy was not smooth. Underlying the positive trend were spells of retrogression. In newly settled areas literacy started high. As these areas developed, literacy declined, but it soon recovered and eventually surpassed the initial level. A few scholars have noted creolean degeneracy in literacy for isolated places and times in colonial America (Bailyn 1960: 78–83; Knauss 1918: 74). Typically, some obstacles to literacy transmission from immigrants to subsequent generations of colonists are assumed to have caused temporary declines in literacy. However, this nonlinear pattern has not been seen as a phenomenon affecting all colonial regions, nor has it been systematically measured.8 Quantitative evidence of this nonlinear pattern exists for several colonies (see Figure 6.1). Soltow and Stevens (1981: 34–7) found such a pattern between 1640 and 1760 for the regions of Maine, New Hampshire, and Essex County and Hartford County, Massachusetts. Literacy started high, 60–70 percent, then declined by as much as 20 percentage points by 1680, and finally recovered to initial levels by 1710. Because immigration to New England declined precipitously after the middle of the seventeenth century, the decline in literacy may represent

114   German immigration to America, 1709–1835

Percentage literate

100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

(1)

(2)

(3)

Initial immigration possessing above average literacy

“Creolean degeneracy” in subsequent generation

Return to positive long run growth trend in literacy

1650

1690

1670

1710

1730

1750

1770

Year New Hampshire, NE Maine, NE

Essex Co., NE Hartford Co., NE

Perquimans Co., NC Chester Co., PA Lancaster Co., PA

Figure 6.1  Patterns of regional literacy growth in rural colonial America.

some retrogression among American-­born New Englanders from the literacy level of their immigrant parents. Tully (1972) found a nonlinear pattern for the rural counties of Chester and Lancaster, Pennsylvania. Literacy declined by 3 to 5 percentage points between 1729 and midcentury and then recovered to the 1729 level by 1774. Gallman (1988: 574) found a similar pattern for Perquimans County, North Carolina. In the decades after initial colonization, literacy declined by over 10 percentage points. It recovered to initial levels by the middle of the eighteenth century. Only one study attempts to measure colonial literacy by where literacy was acquired. Kilpatrick (1912: 197) separated the New York City Dutch population between 1675 and 1738 into men who spent their school-­age years in Europe and men who spent them in America. In 1675 the American group was 38 percentage points less literate. The literacy differential steadily declined until by 1698 the two groups had similar literacy rates. Kilpatrick’s evidence is consistent with the nonlinear patterns in Figure 6.1 and suggests that the cause involves differences in European and American education.

Literacy   115 The nonlinear patterns of literacy growth in Figure 6.1 are consistent with the creolean degeneracy hypothesis. Typically, a township, county, or subregion was first settled by European immigrants, but subsequent population growth was dominated by American-­born colonists. This pattern of population growth, coupled with the observed decline in literacy after initial settlement, implies that European immigrants failed to transfer their above-­average literacy to their American-­born children. Therefore, as the share of American-­born children in a county’s population increased, the average measured literacy of the county’s population would temporarily decline before resuming positive long-­run growth. While the quantitative estimates in Figure 6.1 are consistent with the creolean degeneracy hypothesis, they are not conclusive proof of creolean degeneracy. They do not directly measure the literacy of immigrant parents and then compare it with the literacy of their American-­born offspring, nor do they show whether literacy was initially acquired in Europe or in the colonies. A simulation of overlapping generations illustrates that the observed nonlinear trend in literacy could be attributed to creolean degeneracy.9 Suppose the literacy of a county’s population is measured at four 30-year intervals, from initial colonization by European immigrants in 1700 until 1790, using signatures on wills. Let each interval represent a new generation who are 30 years of age at the date of measurement, let each new generation be twice the size of the preceding generation, let population growth after initial colonization be only through indigenous increase, and for each generation let a random 25 percent sign wills at age 30 and the remaining 75 percent survive to sign wills at age 60. Acquisition of literacy is assumed to be complete by age 30 for each generation. Finally, let European immigrants be 70 percent literate, first-­generation American-­born 50 percent literate, second-­generation American-­born 70 percent literate, and third-­ generation American-­born 80 percent literate. These conditions simulate the relevant features of the evidence (see Figure 6.1), namely, that immigrants were relatively literate, the proportion of American- to European-­born in the population rose over time, colonial population density rose over time, and among American-­born children literacy rose across generations. For example, suppose 100 European immigrants who are age 30 and 70 percent literate arrive in 1700. Of this group, 25 percent sign wills in 1700, so that the measured literacy of the population in 1700 is 70 percent. In 1730, 200 first-­generation American-­born residents, who are age 30 and 50 percent literate, and 75 European immigrants, who are age 60 and 70 percent literate, are present. Of this group, 25 percent of the first generation sign wills and all 75 remaining immigrants sign wills, so that the measured literacy of the population in 1730 is 62 percent. In 1760, 400 second-­generation American-­born residents, who are age 30 and 70 percent literate, and 150 first-­generation residents, who are age 60 and 50 percent literate, are present. Of this group, 25 percent of the second-­ generation sign wills and all 150 remaining first-­generation sign wills, so that the measured literacy of the population in 1760 is 58 percent. The same algorithm produces a measured literacy of the population in 1790 of 74 percent. Thus, literacy follows the nonlinear pattern over time, falling from 70 percent in 1700 to

116   German immigration to America, 1709–1835 62 percent in 1730, then to 58 percent in 1760, and finally rising to 74 percent literate by 1790.10 Qualitative evidence from contemporary observers in several colonies is also consistent with the nonlinear pattern of literacy growth and with the creolean degeneracy explanation of that pattern. For example, Bailyn (1960) provided accounts of educational decline on the New England frontier following initial colonization in the late seventeenth century. From this evidence Bailyn concluded that the high cost of schooling on the frontier left the responsibility of transmitting literacy to the family, and the family failed to maintain literacy levels. Similar accounts can be found for German communities in Pennsylvania for the middle of the eighteenth century. In 1747, Henry M. Muhlenberg (1942, I: 140–1), head of the Lutheran church in Pennsylvania, wrote in his diary: I had on several occasions been urged to visit a number of Lutherans who lived in the northwestern mountains 46 miles from my house. The people there are making a poor and precarious living and lack both bodily and spiritual nourishment. Some of them are growing up wild and hence have no further interest in churches and schools. I arranged . . . for a grant of a small tract of land for a church. On this land they were to build a wooden schoolhouse or church and, if in time they became able to support a schoolmaster, the children at least would be given some help. Gottlieb Mittelberger (1960: 46–7), a German teacher and pastor visiting Pennsylvania between 1750 and 1754, commented: In Pennsylvania preachers receive no salary or tithes. Their total income is what they get annually from their church members, and that sum varies a good deal. Schoolteachers are in the same situation. Many hundreds of children, however, are unable to go to school partly because of the extent of forest land they would have to traverse. Thus many planters lead a rather wild and heathenish life. The situation of rural churches and schools is similar. In general, churches and schools are located only in those places where several neighbors or church members live close together. The problem of educating American-­born Germans was also debated in the Pennsylvania German press. For example, Christopher Sauer’s Almanac, published in Philadelphia in 1752, presented a debate between a new German settler and an old Pennsylvania German inhabitant (reprinted in Weber 1905: 17): new-­comer: 

A matter that is of very great importance to me is, that, in Germany, one is able to send his children to school to have them instructed in reading and writing. Here it is well nigh impossible to get such instruction; especially, where people live so far apart. O, how fortunate are they who have access to a good teacher by whom the children are well taught and trained!

Literacy   117 inhabitant: 

It is true. On that account many children living on our frontiers grow up like trees. But since the conditions are such that few people live in cities and villages as they do in Germany it is natural that one meets with certain inconveniences. Where is there a place in this world where one does not meet with some objectionable feature during his natural life? new-­comer:  But this is an exceptional want, for if children are thus brought up in ignorance it is an injury to their soul’s welfare,—an eternal injury. Muhlenberg, Mittelberger, Sauer, and other members of the German-­American immigrant community thought that schooling their offspring was more difficult on the American frontier than obtaining their own schooling was in Europe when they were children. On the American frontier, family instruction was usually the only option for transmitting literacy from parent to child, and it proved inad­ equate for maintaining literacy levels. Without access to formal schooling, children on the frontier grew up “wild,” “like trees.” Thus, the American-­born offspring of German immigrants were in danger of literacy degeneration. The quantitative and qualitative evidence which describes regional nonlinear trends in colonial literacy is wide-­ranging, emerging from many different colonial regions. However, it is also fragmentary, composed of small samples and less than comprehensive in its coverage of colonial regions. The evidence appears to support a nonlinear growth path only in rural areas and not in urban areas. More evidence needs to be gathered to firmly establish the extent and character of these nonlinear literacy patterns. Finally, the current evidence is consistent with but does not conclusively prove creolean degeneracy as the cause of the nonlinear pattern. However, the evidence is strong enough to support creolean degeneracy as a working hypothesis for future research into colonial literacy. The history of colonial literacy would be enhanced if future studies measured these nonlinear patterns and did so in such a way that the location and method of acquiring literacy might be discerned.

A supply versus demand model of colonial growth in literacy The evidence can be used to explore the causes of the nonlinear trend in colonial literacy and to develop models to guide future research. The difference in literacy between rural and urban areas suggests that changes in population density may have influenced literacy trends. European migration to the colonial frontier represented a dramatic change in the surrounding population density and therefore may be an important factor. Differential literacy between immigrants and their American-­born children may also be an important factor. The combination of these two influences could produce a nonlinear literacy trend. For example, immigrants may have come from a more densely populated environment, in which the costs and benefits of acquiring literacy differed from what their children experienced on the colonial frontier. These different costs and benefits may have led immigrant parents to transmit a lower level of literacy to their children in America than what they had acquired as children in Europe. If so, colonial

118   German immigration to America, 1709–1835 l­iteracy started out relatively high because immigrant literacy was high. The failure of immigrant parents to transfer literacy to their American-­born offspring temporarily lowered colonial literacy. Finally, as American-­born colonists came to dominate the population, measured literacy was dominated by the long-­run rise in literacy in subsequent generations. A positive correlation between population density and literacy has been noted in several studies, but exactly how rising population density causes literacy to increase is less clear (Graff 1987a, 1987b; Lockridge 1974; Soltow and Stevens 1981). Typically, the arguments in the literature have used an implicit supply-­ and-demand discussion to relate changes in population density to changes in literacy. Literacy is thought of as a market: some individuals demand it and others supply it. Changes in population density shift either the entire demand schedule or the entire supply schedule of literacy, per given cost of acquiring literacy, and thus the quantity of literacy acquired by the population changes. Whether rising population density leads to higher literacy through a shift in the demand schedule or through a shift in the supply schedule has remained unresolved. The literature has used these economic ideas loosely, without formulating explicit models. Typically, supply and demand shifts are mentioned simultaneously, without much discussion of the possible contradictions or alternative implications of a supply explanation versus a demand explanation.11 Discovering whether population density works through shifts in supply or in demand is important for understanding the causes of growth in literacy, for tracing its ramifications for other aspects of economic and social life, and for assessing its benefits in the colonial economy. The implicit models in the literature are made explicit here, and their implications are systematically explored. What makes economic theory empirically useful is that supply and demand schedules do not shift by themselves or for no reason. Observable parameters shift these curves, and typically a given parameter shifts only the demand curve or the supply curve, not both. For example, changes in income shift the demand curve but not the supply curve; changes in technology shift the supply curve but not the demand curve. Population density is an atypical parameter in that it may affect either the supply or the demand curve. This possibility gives rise to two different models of how changing population density affects the literacy market, the demand model and the supply model. The two alternative models, presented in Figure 6.2, are drawn to illustrate the nonlinear trend in colonial literacy. The initial position, Demand1 or Supply1, represents the conditions experienced by immigrant parents when they were children of educable age in Europe. The second position, Demand2 or Supply2, represents the conditions experienced by immigrant parents with regard to educating their American-­born offspring. The final position, Demand3 or Supply3, represents the conditions experienced by subsequent generations of American-­born children.12 Both the supply model and the demand model are drawn to trace the same longitudinal pattern in the quantity of literacy illustrated in Figure 6.1: starting from high initial levels at Q1, literacy temporarily declines to Q2 and then finally

Literacy   119 Supply explanation (increasing population density shifts supply right)

Demand explanation (increasing population density shifts demand right)

Supply1

P2 P1

Supply3

P3 Demand Q2

Q1

Q3

Quantity of literacy (percentage literate)

Cost of acquiring literacy

Cost of aquiring literacy

Supply Supply2

P3 P1

Demand3

P2 Demand1

Demand2 Q2

Q1

Q3

Quantity of literacy (percentage literate)

Figure 6.2 Supply and demand model of colonial literacy decline and growth caused by changes in population density.

rises past the starting level to a long-­run higher level at Q3. Immigrants arrived in America with Q1 literacy, but because they left a densely populated Europe for a sparsely populated America, demand and/or supply shifted left to Q2 for their American-­born children. Finally, as population density in America increased over time, demand and/or supply shifted right and returned literacy to higher levels at Q3 for subsequent generations of Americans. The demand model typically assumes that rising population density increased the demand for literacy by increasing its commercial value or by increasing the need to socialize the young politically and religiously. The benefits of being literate increased with the frequency of commercial transactions and social intercourse, which is assumed to increase with population density. Literacy also helped in the development of other trade skills which may have been more important in urban areas (farmers are assumed to have little commercial need for literacy). Temptations and vices are assumed to become relatively more avail­ able as population density increases. Thus, as population density rose, parents would demand relatively more literacy for their children to maintain social control or to provide their children with social and marketable skills. The demand model argues that the different levels of literacy between urban and rural areas was caused by differences in the relative demands for literacy rather than by differences in the relative costs (supply) of literacy. Over time, as population density rose, the commercial value of literacy rose or the need for social control rose, and thus the demand for literacy shifted right, causing long-­run growth in literacy (Craig 1981: 157–68; Cremin 1970: 545–59; Cressy 1980:

120   German immigration to America, 1709–1835 88–95; Gilmore 1982, 1989; Graff 1987a, 1987b: 249–51; Lockridge 1974: 43–108; 1981: 183–98). Shifts in the demand for literacy caused by changes in population density might explain the nonlinear pattern in colonial literacy. For example, German immigrants to Pennsylvania came from a relatively dense population environment, where literacy had a relatively high commercial value. Thus, they had a high level of literacy because they had a high demand for literacy when growing up in Germany, Demand1. On moving to the colonial frontier, they entered a less dense population environment, where literacy had a relatively low commercial value. Therefore, immigrants were overeducated, given the value of literacy in their new environment. Given the lower commercial value of literacy on the colonial frontier, and hence a lower demand for literacy, immigrant parents would transmit a lower level of literacy to their American-­born children than they possessed, Demand2. Finally, as population density rose, the commercial value of literacy would rise, and hence the demand for literacy among subsequent American-­born generations would rise, returning society to a long-­run growth path in literacy, Demand3.13 Some support for the demand explanation comes from the positive correlation between literacy and both wealth and occupation found in cross-­sectional evidence (for example, see Beales 1978; Cremin 1970; Cressy 1980; Graff 1987b; Lockridge 1974; Tully 1972). The observed rural–urban literacy differential might be caused by income or wealth differentials rather than by population densities. If the acquisition of literacy is thought of as a consumption activity rather than an investment activity, and given a positive income elasticity of demand for literacy, then secular increases (or decreases) in income or wealth shift the demand schedule out (or in), causing the population’s literacy level to rise (or fall). This interpretation of the demand model has two problems. First, the estimated correlations between wealth and literacy relate the wealth and literacy of the same individual. Because individuals acquire literacy long before their occupation and wealth are determined, this evidence suggests that any causality flows from literacy to wealth rather than from wealth to literacy. Literacy would appear to be more an investment good than a consumption good. To estimate the effect of wealth on the demand for literacy, future studies should compare the wealth of parents with the literacy of their children, other things being equal. Second, the nonlinear trend in colonial literacy is difficult to explain as demand shifts driven by changes in income or wealth. Such an explanation of creolean degeneracy in literacy leads to the conclusion that the wealth of immigrants in America was significantly below the wealth their parents possessed when they were children in Europe. Evidence on income and wealth differentials, and the fact that immigration flows remained positive, suggests that this scenario is unlikely (Jones 1980; Perkins 1988: 1–18, 212–40). In contrast, the supply model assumes that rising population density increased the quantity of literacy by lowering the supply cost of acquiring it. Costs are assumed to have declined with population density because of economies of scale

Literacy   121 in the provision of education. One teacher, one schoolhouse, one lesson could accommodate several students as easily as one student. The economies of scale would spread the fixed initial costs of the schoolhouse and teacher over several students and thus lower the unit cost for each student. In other words, as population density rose and group instruction became feasible, the average cost per student would decline over some limited range. Thus, the short-­run supply of literacy would shift to the right as population density increased. These economies of scale would be exhausted quickly in the city. Therefore, falling education costs caused by rising population density were more salient in the countryside or on the frontier; they had little continuing impact in urban areas. The initial difference in literacy levels between urban and rural areas would have been caused by the difference in the cost of education rather than by differences in the demand for literacy (Bailyn 1960: 78–84; Cremin 1970: 517–43; Gilmore 1982, 1989; Graff 1987b: 249–51; Lockridge 1974: 43–108; Soltow and Stevens 1981: 39–42). Shifts in the supply of literacy might explain the nonlinear pattern of colonial literacy. For example, German immigrants to Pennsylvania grew up in a relatively dense population environment in Germany, where the cost of acquiring literacy was relatively low, Supply1. On moving to the colonial frontier, they entered a less dense population environment and consequently experienced a relatively high cost of transmitting literacy to their American-­born offspring, Supply2. The higher cost of acquiring literacy on the frontier led immigrant parents to transmit a lower level of literacy to their American-­born children than they had acquired as children in Germany. As population density in America increased, the cost of education declined, shifting the supply curve to the right and returning society to a long-­run growth path in literacy for subsequent generations of Americans, Supply3. Although rising population density may have affected literacy through shifts both in supply and in demand, discovering the dominant effect is important for interpreting the impact of literacy on colonial America. If the demand model is dominant, then the high initial levels of colonial literacy were of little benefit to the colonial economy. Immigrants were overeducated. Resources used to acquire literacy prior to emigration were wasted, because literacy was of little value on the colonial frontier. Creolean degeneracy in literacy would represent no loss to colonial society. Maintaining the literacy levels of the overeducated immigrants in subsequent generations would involve investing in skills (or in social control) that had little value relative to other frontier activities. By contrast, if the supply model is dominant, then the high initial levels of colonial literacy were a net wealth transfer from Europe to America. Immigrants acquired useful skills at low cost in Europe and then brought them to the Amer­ ican frontier, where learning them would have cost more. Creolean degeneracy in literacy would represent a productive loss to the economy as subsequent generations had to adapt to a higher cost of acquiring literacy than their immigrant parents faced prior to leaving for America.

122   German immigration to America, 1709–1835

Testing the supply versus demand models Although the movements in the quantity of literacy are similar for both the supply and the demand models, the two explanations differ in regard to the movements in the cost of acquiring literacy (see Figure 6.2). If rising population density shifted the demand curve for literacy to the right, then the cost of acquiring literacy would be bid up. Alternatively, if rising population density shifted the supply curve to the right, then the cost of acquiring literacy would fall. Therefore, the dominant explanation of the nonlinear pattern in colonial literacy can be determined through measuring movements in the cost of acquiring literacy. If the cost were initially higher in Europe than on the colonial frontier and subsequently rose in the colonies over time, then the demand explanation would dominate the supply explanation. If the cost were initially lower in Europe than on the colonial frontier and subsequently fell in the colonies over time, then the supply explanation would dominate the demand explanation.14 While measuring changes in the cost of literacy would be the simple way to discover whether changing population density affected literacy more through shifts in supply or through shifts in demand, quantitative evidence on changes in the cost of acquiring literacy is not readily available. Research on the expansion of literacy in the seventeenth and eighteenth centuries has largely ignored cost measurements in favor of quantity measurements. One of the strengths of the supply-­and-demand model is that it shows that gathering more quantitative evidence on signature literacy cannot by itself answer some of the important questions involving the causes and consequences of the growth in colonial literacy. Future research should be directed towards measuring changes in the cost of acquiring literacy. Estimation of the cost of literacy should include not only the direct costs of tuition, books, and teacher salaries but also the opportunity cost of the lost work time that students, instructors, and parents experienced during the education process. Part of this new research should include studying shifts in methods of transmitting literacy from parents to children, because the method chosen might affect the overall cost of literacy. Currently, the portion of literacy acquired at home, at work, at church, or at school is not known. Finding quantitative evidence on the cost of acquiring literacy and on the methods used to transmit literacy in the colonial period may seem impossible, but then so was finding significant quantitative evidence on signature literacy for the colonial period 30 years ago. As with the accumulation of quantitative evidence on signature literacy, it may take 30 years and many new dissertations to develop adequate evidence on the cost and methods of transmitting literacy in the colonial period. Several research directions might prove fruitful, such as estimating the threshold settlement size for setting up a full-­time school, estimating the greatest distance a student could travel and still maintain school enrollment, or taking ranges of estimates of schooling fees, building costs, and labor costs and estimating the minimum enrollment needed to support a full-­time school, and then using demographic information on age structure to estimate the threshold settlement size

Literacy   123 needed to support a full-­time school. Developing and expanding alternative evidence, for example, from contract labor markets, might also be useful for estimating education costs. For example, Grubb (1988: 599) found that the cost of acquiring employer-­provided education through an indentured servant contract for German children in Pennsylvania was a 22 percent increase in contract length, independent of productivity differences relating to age. Developing and expanding alternative evidence might also be useful for estimating the method used to transmit literacy. For example, servant contracts negotiated in Philadelphia for the children of German immigrants frequently stipulated the method of education to be provided by the employer. Out of 243 contracts in the servant records housed in the City Archives of Philadelphia (1771–73) for male German children, 33.3 percent stipulated the provision of some kind of education. Given that adult German immigrants in the 1770s were 80 percent literate (see Table 6.1), this relatively infrequent stipulation is consistent with the creolean degeneracy hypothesis. The rural–urban difference in literacy is also present in this evidence. Among servants going to Philadelphia masters, 39 percent had education stipulated in their contracts, whereas among servants going to masters in the surrounding countryside, 28 percent had education stipulated in their contracts. However, this evidence on creolean degeneracy and rural–urban differentials by itself does not reveal whether the underlying causes were supply- or demand-­driven. What is interesting in this servant evidence, and illustrative of the new directions research in colonial literacy should take, is that the method of providing education was also stipulated in the servant contracts. Parents would want to choose the method of education which was the least costly to their child’s employer, because the cost of the education was paid by their child through additions to his or her contract length. Thus, by choosing the least costly method  of education for that employer, the parent also reduced the labor time needed by their child to acquire literacy. In the contract evidence, parents either  specified a number of months of formal schooling or required only that reading and writing be taught (informal education), leaving the method of instruction up to the employer. The ratio of formal schooling to informal instruction in the contracts which stipulated education was systematically higher for children going to Philadelphia employers than for children going to employers in the surrounding countryside, 0.5 compared with 0.1. This comparison implies that formal schooling was less costly in the city than in the surrounding countryside and thus supports a supply-­dominated explanation of the nonlinear trend in colonial literacy. Finally, qualitative evidence might help sort out the demand and supply explanations. While the conclusions based on the opinions of contemporary observers must remain tentative, this evidence would appear to support the supply model as the dominant explanation. For example, the German community in colonial Pennsylvania repeatedly stressed the higher cost of schooling on the colonial frontier relative to Europe. They related the cost difference to relative population density (see the quotes above).

124   German immigration to America, 1709–1835 There appears to be less qualitative evidence which indicates that literacy was not in demand on the frontier. The colonial countryside was not a subsistence economy. Trade, exchange, and production for the market permeated the rural landscape (for example, see Jensen 1986: 79–91; Rothenberg 1981; Schweitzer 1987). It is not certain that the commercial value of literacy, and hence the demand for literacy, was necessarily lower in the countryside. Therefore, a supply-­dominated explanation would appear to be more consistent with the quantitative and qualitative evidence than a demand explanation.15

Conclusion Colonial America was a literate society by European standards. The accumulation of physical and human capital may have been part of the same process that was rapidly propelling Americans from frontier backwardness to competitiveness with the economies of western Europe. The rise of mass literacy may have affected the cultural, political, and economic direction of the new nation (for example, see Cremin 1970: 359–75, 545–59; Cressy 1987; Gilmore 1989; Jensen 1986: 114–28; Knauss 1918; Steele 1986; Tully 1972). However, the path to mass literacy was neither simple nor direct. Both quantitative and qualitative evidence suggests that local or regional literacy growth was nonlinear. While this nonlinear pattern appears to have been a general phenomenon across the colonies, more research is needed to establish its extent. The causes and economic consequences of the nonlinear pattern of literacy growth are less clear. Changes in population density and the dynamics of transmitting literacy from immigrant parents to American-­born offspring appear to be important factors. The evidence is consistent with the creolean degeneracy hypothesis. Colonial immigrants were more literate than their European home populations. However, on the colonial frontier, immigrant parents failed to transmit their high level of literacy to their children. Finally, as colonial populations became dominated by American-­born colonists, literacy growth returned to a positive long-­run growth trend. Why colonial literacy should have behaved in this fashion is also unclear. The quantitative evidence on signature literacy is consistent with both a supply and a demand explanation. Further progress on these issues calls for a change in research strategies. The previous emphasis on gathering quantitative evidence of signature literacy should be redirected towards gathering quantitative evidence of the changing cost of acquiring literacy. Evidence on costs will help solve the problem of whether changes in the quantity of literacy was supply- or demand-­ driven. Finally, research should be redirected towards gathering quantitative evidence of the methods of transmitting literacy. The choice of method, such as informal education at home or work versus formal education at school or church, will help explain how colonists reacted to changes in the cost of acquiring literacy and changes in the value of being literate during the colonial era.

Literacy   125

Notes   * Originally published as Farley Grubb, “Growth of literacy in colonial America: longitudinal patterns, economic models, and the direction of future research,” Social Science History, vol. 14, no. 4 (Winter 1990): 451–82. Copyright 1990. © 1990 by the Social Science History Association. All rights reserved. Reprinted with permission of the publisher, Duke University Press. F. W. Grubb is associate professor of economics at the University of Delaware. His field of research is immigration and immigrant labor markets in North America from 1650 through 1835, and he has recent publications in this and many other journals. Earlier versions of this article were presented at the History Seminar, University of Delaware, and at the Mellon (parss) Seminar on Work and Population at the University of Pennsylvania. The author wishes to thank the participants in these seminars and David Galenson, Henry Gemery, Claudia Goldin, Lynn Lees, David Mitch, Erik Monkkonen, and anonymous referees for helpful comments on earlier drafts. Social Science History 14:4 (Winter 1990). Copyright © 1990 by the Social Science History Association. ccc 0145–5532/90/$1.50.   1 Most quantitative studies have concentrated on estimating cross-­sectional variance in literacy within a given population. Quantitative measures of literacy for the seventeenth and eighteenth centuries were positively co-­related with population density, wealth, occupational status, and religion (Cressy 1980: 104–41; Lockridge 1974: 13–27; Galenson 1979; Schofield 1973: 450; Stone 1969; Houston 1982). In many cases, these correlations have been interpreted as causes of variation in literacy across societies. As these societies moved towards mass literacy in the late eighteenth century, many of these correlations were weakened or eliminated. The study here will focus on common longitudinal patterns in literacy across societies or colonies rather than on cross-­sectional patterns found within societies.   2 This correspondence between reading and writing was less strong in Sweden (Johansson 1981).   3 The study here will be confined to adult male literacy. While there are studies of adult female literacy (for example, see Auwers 1980; Gilmore 1989), they are too few, with sample sizes that are too small, to allow examination of common longitudinal patterns across colonial regions or to allow adequate comparison with northwestern Europeans. Studies of literacy among children and young teens are virtually nonexistent.   4 Probated wills tend to be drawn from the segment of the population with higher occupational status and greater wealth. Because literacy is positively correlated with wealth and occupation, signature evidence from wills may overstate literacy in the general population. However, probated wills also measure literacy late in life, when literacy skills may have deteriorated due to feebleness or senility. The net direction of the bias is unclear, but the wealth bias has usually been assumed to dominate. Recently, however, tests for wealth biases in samples of colonial probated wills have shown the bias to be insignificant (Main and Main 1988: 29–34). Deeds, petitions, and depositions suffer from double counting of some groups in the population, typically the more literate, and also generate inconsistent evidence for some individuals. For more discussions of the biases in the evidence and other measurement problems, see Beales 1978: 94–5; Cremin 1970: 575–6; Cressy 1980: 105–8; Gallman 1987; Lockridge 1974: 7–14, 133, 152; Soltow and Stevens 1981: 38, 207; Tully 1972: 304.   5 The typical sources of signature literacy are different between the colonies and Europe. Probated wills are the most common evidence used for the colonies, and marriage records are the most common evidence used for Europe. The relative direction of any biases is unclear. Probated wills may overrepresent the wealthy and higher occupational groups, which tended to be more literate relative to the population measured by marriage records. However, probated wills also measure literacy late in life, when it may have declined (see n. 4), relative to when marriage records measure liter-

126   German immigration to America, 1709–1835 acy. In addition, the average age difference between those signing wills and those marrying, some 25 years, may understate the relative superiority of colonial literacy to that of Europe, other things being equal. In a given year, signers of wills should be older on average and thus members of an earlier, less literate generation than signers of marriage registers. Given a positive long-­run trend in literacy, this age difference will understate the relative literacy of a given population in a given year when evidence from wills is used rather than evidence from marriage registers. My best guess is that correcting for these biases will not change the relative literacy ranking by region or country. See also notes 1 and 4.   6 Because the measured estimates of literacy for colonial populations mix European and American-­born evidence, the difference between immigrant literacy and American-­ born literacy may be greater than indicated in table 6.1. The French-­Canadian evidence is less supportive and harder to evaluate regarding this pattern. English indentured servants, in general, were low in the occupational hierarchy. Clearly, since they could not pay the £10 that passage to America cost, they possessed relatively little wealth. Given that literacy is positively correlated with wealth and occupational status, the relatively high levels of literacy among English indentured servants suggests that the migration process positively selected colonists in the New World from the literate in the European population. The German evidence from Lancaster County, Pennsylvania, was based on wills and was thought to be biased high (Tully 1972: 305). The German immigrant evidence was relatively complete and so no representative biases exist. Therefore, German immigrant literacy may have exceeded resident colonial German literacy by an even greater margin than indicated in table 6.1 (Grubb 1987: 66).   7 In Grubb’s study, among Germans who had emigrated to Philadelphia between 1727 and 1775 a sample of 192 fathers with 252 adult sons (above age 16) was taken. The average age of the fathers was 50 and the average age of the sons was 20. The signature literacy of fathers and their sons had a correlation coefficient of 0.42. It was significantly different from 0 at the 0.0001 level, a t statistic of 6.39. The proportion of literate fathers with all literate sons and illiterate fathers with all illiterate sons was 69 percent.   8 Nonlinear patterns are found only when literacy is measured at frequent intervals and only when samples are confined geographically, such as to counties. For example, Lockridge (1974) did not find a nonlinear pattern in his estimates for New England (see table 6.1). However, his estimates are too infrequent and taken from too wide a geographical area. His sample points span the nonlinear pattern and thus miss the intervening decline in literacy, as suggested by the more recent estimates of Soltow and Stevens (1981).   9 While this exercise is less ambitious than a fully specified computer simulation model, the point of the simulation, namely, that under reasonable conditions creolean degeneracy may explain the nonlinear literacy trend, can be made with this simple two-­ generation overlapping simulation. Thirty-­year intervals are used because acquisition of literacy was usually complete by age 30 and few individuals survived to age 90. 10 The calculation of the population’s literacy in this simulation model for 1700 is (100 × 0.25 signing wills × 0.7 literate)/(25 signing wills) = 0.7 literate; for 1730, [(200 × 0.25 signing wills × 0.5 literate) + (100 × 0.75 signing wills × 0.7 literate)]/ (125 signing wills) = 0.62 literate; for 1760, [(400 × 0.25 signing wills × 0.7 literate) + (200 × 0.75 signing wills × 0.5 literate)]/(250 signing wills) = 0.58 literate; and for 1790, [(800 × 0.25 signing wills × 0.8 literate) + (400 × 0.75 signing wills × 0.7 literate)]/(500 signing wills) = 0.74 literate. In the actual historical environment there are several complicating factors abstracted from in this simulation, such as the possibility of continued stochastic additions of European immigrants to a given frontier region as well as stochastic migrations of American-­born offspring into or out of that region. 11 Changes in the observed quantity of literacy in a population should not be equated with changes in the demand for (or supply of ) literacy. This equation represents a

Literacy   127 classic error of economics, namely, confusing movements along a demand (or supply) schedule with shifts in the entire schedule. A shift in demand or supply represents an entirely new schedule of quantity of literacy per cost of acquiring literacy. Similarly, changes in parameters, such as population density, which affect supply (or demand) conditions cannot be interpreted as movements along the given supply (or demand) curve, because the supply and demand curves are plotted only with cost and quantity on the axes. Population density is not plotted along one of the axes. Changes in parameters that are not plotted along one of the two axes must be interpreted as shifts in the entire supply (or demand) schedule. 12 It should be noted that Demand1 and Supply1 represent the position of potential immigrants in Europe and not the average position for the whole European population. Immigrants to America were more literate than their European home populations (see Tables 6.1 and 6.2). In addition, while Q2 is below Q1 in Figure 6.2, it is not necessarily below the average literacy of their respective European home populations. In other words, creolean degeneracy would not necessarily lead to lower literacy levels in America than in Europe. The overall supply and demand for literacy in America may have been higher than the overall supply and demand for literacy in Europe for a number of reasons, such as relatively high religious fervor; a relatively high proportion of land ownership, which required knowledge of property rights and accounting techniques; a relatively high proportion of literate adults, who could then teach literacy to children; and so on. 13 Alternatively, one could argue that German immigrants came from a relatively dense population environment in Europe where the demand for literacy was high because social control of one’s children—keeping them away from vice and competing religious doctrines—was important, Demand1. On moving to the colonial frontier, these German immigrants entered a less dense population environment, where vice and competing religious doctrines were less of a threat. Therefore, immigrants were overeducated, given the value of literacy in their new environment. Given less need for social control on the colonial frontier, and hence a lower demand for literacy, immigrant parents would pass on a lower level of literacy to their American-­born children than they possessed, Demand2. Finally, as population density in America rose over time, the need for social control would also rise, and hence the demand for literacy among subsequent American-­born generations would rise, returning society to a long-­ run growth path in literacy, Demand3. 14 The opposite cost predictions of the supply and demand models illustrate the different possible impacts of literacy on the colonial economy. If literacy increased because the supply curve shifted right, then the economy gained in literacy without sacrificing as many resources as previously. Expansion of literacy was accompanied by the freeing of resources that could be put to other productive uses. The initial creolean degeneracy in colonial literacy represented an economizing reaction to the increase in the real costs of literacy, which were draining the productive resources of the immigrants. Alternatively, if literacy increased because the demand curve shifted right, then the economy had to sacrifice more resources than previously to produce the higher desired literacy level. Resources were withdrawn from other productive uses in the economy to produce the higher literacy. The initial creolean degeneracy in colonial literacy represented a reaction in the colonies to the lower value of literacy, which freed educational resources for employment elsewhere during initial settlement. 15 This is not to say that demand factors were never important. For example, the difference in literacy between radical Protestants, mainstream Protestants, and Catholics is probably best explained as a difference in the demand for literacy, in terms of the strength of the desire to read the Bible for oneself. However, the emphasis in the qualitative evidence on the religious value of literacy presents another problem for the demand model. The connection between rising population density and increasing demand for literacy comes in part from the assumed higher commercial value of

128   German immigration to America, 1709–1835 literacy in urban areas. However, other components of the demand for literacy, like religious motives, may work in the opposite direction. The religious demand for literacy may have been lower in urban areas because access to churches and pastors was easier there than in frontier areas, other things being equal. In town, unlike on the frontier, a person had ways of hearing God’s Word other than reading the Bible. Religious revival and awakening in the colonies was important in the countryside.

References Auwers, L. (1980) “Reading the marks of the past: Exploring female literacy in colonial Windsor, Connecticut.” Historical Methods 13: 204–14. Bailyn, B. (1960) Education in the Forming of American Society. New York: W. W. Norton. Beales, R. W., Jr. (1978) “Studying literacy at the community level: A research note.” Journal of Interdisciplinary History 9: 93–102. Cipolla, C. M. (1969) Literacy and Development in the West. Baltimore: Penguin Books. City Archives of Philadelphia (1771–3) “Record of indentures of individuals bound out as apprentices, servants, etc. and of German and other redemptioners in the Office of the Mayor of the City of Philadelphia, October 3, 1771 to October 5, 1773.” Craig, J. E. (1981) “The expansion of education.” Research in Education 9: 151–213. Cremin, L. A. (1970) American Education: The Colonial Experience, 1607–1783. New York: Harper and Row. Cressy, D. (1980) Literacy and the Social Order. Cambridge: Cambridge University Press. Cressy, D. (1987) Coming Over: Migration and Communication between England and New England in the Seventeenth Century. New York: Cambridge University Press. Deane, P. (1969) The First Industrial Revolution. Cambridge: Cambridge University Press. Galenson, D. W. (1979) “Literacy and the social origins of some early Americans.” Historical Journal 22: 75–91. Galenson, D. W. (1981) White Servitude in Colonial America. Cambridge: Cambridge University Press. Gallman, R. E. (1987) “Two problems in the measurement of colonial literacy.” Historical Methods 20: 137–41. Gallman, R. E. (1988) “Changes in the level of literacy in a new community in early America.” Journal of Economic History 48: 567–82. Gawthrop, R. and G. Strauss (1984) “Protestantism and literacy in early modern Germany.” Past and Present 104: 31–55. Gilmore, W. J. (1982) “Elementary literacy on the eve of the industrial revolution: Trends in rural New England, 1760–1830.” Proceedings of the American Antiquarian Society 92: 87–178. Gilmore, W. J. (1989) Reading Becomes a Necessity of Life. Knoxville: University of Tennessee Press. Graff, H. J., ed. (1981) Literacy and Social Development in the West. London: Cambridge University Press. Graff, H. J. (1987a) The Labyrinths of Literacy. London: Falmer. Graff, H. J. (1987b) The Legacies of Literacy. Bloomington: Indiana University Press. Grubb, F. (1987) “Colonial immigrant literacy: An economic analysis of Pennsylvania-­ German evidence, 1727–1775.” Explorations in Economic History 24: 63–76. Grubb, F. (1988) “The auction of redemptioner servants, Philadelphia, 1771–1804: An economic analysis.” Journal of Economic History 48: 583–603.

Literacy   129 Houston, R. (1982) “The literacy myth? Illiteracy in Scotland, 1630–1760.” Past and Present 96: 81–102. Jensen, J. M. (1986) Loosening the Bonds: Mid-­Atlantic Farm Women, 1750–1850. New Haven, ct: Yale University Press. Johansson, E. (1981) “The history of literacy in Sweden,” in H. J. Graff (ed.) Literacy and Social Development in the West. London: Cambridge University Press: 151–82. Jones, A. H. (1980) Wealth of a Nation to Be. New York: Columbia University Press. Kilpatrick, W. H. (1912) The Dutch Schools of the New Netherlands and Colonial New York. Washington, DC: U.S. Government Printing Office. Knauss, J. O., Jr. (1918) “Social conditions among the Pennsylvania Germans in the eighteenth century, as related in German newspapers published in America.” Pennsylvania German Society 28: 1–217. Levine, D. (1979) “Education and family life in early industrial England.” Journal of Family History 4: 368–80. Lockridge, K. A. (1974) Literacy in Colonial New England. New York: W. W. Norton. Lockridge, K. A. (1981) “Literacy in early America, 1650–1800,” in H. J. Graff (ed.) Literacy and Social Development in the West. London: Cambridge University Press: 183–200. Main, G. L. and J. T. Main (1988) “Economic growth and the standard of living in southern New England, 1640–1774.” Journal of Economic History 48: 27–46. Maynes, M. J. (1985) Schooling in Western Europe. Albany: State University of New York Press. Mitch, D. (1984) “Underinvestment in literacy? The potential contribution of government involvement in elementary education to economic growth in nineteenth-­century England.” Journal of Economic History 44: 557–66. Mitch, D. (1988) “The rise of popular literacy in modern times and its explanation.” Unpublished manuscript, Economics Department, University of Maryland Baltimore County, Catonsville. Mittelberger, G. (1960) Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754, trans. and ed. O. Handlin and J. Clive. Cambridge, MA: Harvard University Press. Moogk, P. N. (1989) “Reluctant exiles: The problems of colonization in French North America.” William and Mary Quarterly, 3d ser., 46: 463–505. Muhlenberg, H. M. (1942) The Journals of Henry Melchior Muhlenberg, trans. T. G. Tappert and J. W. Doberstein. 3 vols., Philadelphia: Muhlenberg. Perkins, E. J. (1988) The Economy of Colonial America. New York: Columbia University Press. Rothenberg, W. B. (1981) “The market and Massachusetts farmers, 1750–1855.” Journal of Economic History 41: 283–314. Sanderson, M. (1983) Education, Economic Change, and Society in England, 1780–1870. London: Macmillan. Schofield, R. S. (1968) “The measurement of literacy in pre-­industrial England,” in J. Goody (ed.) Literacy in Traditional Societies. Cambridge: Cambridge University Press: 311–25. Schofield, R. S. (1973) “Dimensions of illiteracy, 1750–1850.” Explorations in Economic History 10: 437–54. Schofield, R. S. (1989) “Fertility, nuptiality, and mortality according to literacy: 1754–1840.” Unpublished manuscript, Cambridge Group for the History of Population, Cambridge, U.K.

130   German immigration to America, 1709–1835 Schweitzer, M. M. (1987) Custom and Contract: Household, Government, and Economy in Colonial Pennsylvania. New York: Columbia University Press. Soltow, L. and E. Stevens (1981) The Rise of Literacy and the Common School in the United States. Chicago: University of Chicago Press. Steele, I. K. (1986) The English Atlantic, 1675–1740: An Exploration of Communication and Community. New York: Oxford University Press. Stone, L. (1969) “Literacy and education in England.” Past and Present 42: 68–139. Strassburger, R. B. (1934) Pennsylvania German Pioneers. 3 vols., Norristown: Pennsylvania German Society. Tully, A. (1972) “Literacy levels and educational development in rural Pennsylvania, 1729–1775.” Pennsylvania History 39: 301–12. Weber, S. E. (1905) The Charity School Movement in Colonial Pennsylvania. Philadelphia: George F. Lasher Press. West, E. G. (1978) “Literacy and the industrial revolution.” Economic History Review 31: 369–83.

7 The age structure of German immigrant literacy*

Introduction Colonial Americans were relatively literate by European standards, based on signature evidence.1 The high rate of literacy may have contributed to the rapid economic, political, and cultural development of the colonies. The source of the literacy has been attributed both to colonial efforts to encourage mass schooling and to the possibility that immigrants were more literate than the European populations from which they came.2 However, literacy at the point of immigration has not been directly estimated. Thus the source of high colonial literacy has remained obscure. For Germans arriving in Pennsylvania during the middle of the eighteenth century, a substantial quantity of signature evidence was gathered at the point of immigration. Between 1727 and 1775 male German immigrants above age 15 were required to sign an oath of allegiance to the British Crown and an oath of abjuration to the Pope and the House of Stuart. The sample of 23,988 signatures is larger and has fewer biases than evidence used in previous studies of colonial literacy.3 Because the German evidence is measured at the point of immigration, immigrant literacy can be compared with the literacy of resident colonists and Europeans, thus determining whether immigrants were a source of high colonial literacy. This evidence can also be used to determine the time trend and the adult age structure of immigrant literacy, something which has seldom been done. The relationships are important for determining both the age when literacy was acquired and the propensity at different ages for literates to migrate.

The trend and relative position of immigrant literacy German immigrants have been regarded as less educated and less literate than other colonials and Northern Europeans. Wright (1957: 108) concluded, “the German mystics who made up the early waves of settlers cared little for learning. Later German settlers of the Reformed and Lutheran churches were for the most part peasants with only a slight interest in education.” In 1753 Benjamin Franklin commented (Sparks, 1838, Vol. VII: 71–3), “Those who came hither

132   German immigration to America, 1709–1835 are generally the most stupid of their own nation”. However, Franklin later stated, “They import many books from Germany, and of the six printing houses in the Province, two are entirely German, two half German and half English, and but two are entirely English.” Assessing such judgments across the entire spectrum of educational achievements is difficult. However, one component can be easily compared across groups and that is the level of signature literacy. The overall illiteracy rate for the 23,988 adult male Germans arriving in Philadelphia between 1727 and 1775 was 29 percent.4 There was a downward trend over the half century, from about 35 percent illiterate in the 1730s to under 20 percent by the 1770s (see Figure 7.1). This decline represented real gains by the immigrants because their age structure changed little over the century (Grubb, 1985). By the American Revolution German immigrants were approaching a level of mass literacy. There can be little doubt that German immigrants contributed to the general superiority of colonial literacy vis-­à-vis the rest of Europe. Craig (1981: 170) ranked adult male illiteracy by European states in 1800 as follows: Scotland, 10 percent; Netherlands, 30 percent; Belgium, 39 percent; England and Wales, 40 percent; and France and the German-­speaking areas, 45 percent. Both Cressy (1980: 117) and Schofield (1973: 450) found adult male English illiteracy in the middle of the eighteenth century to be about 35 percent. Therefore, German immigrants were perhaps second only to Calvinist Scotland in terms of European literacy, suggesting some positive self-­selection by literates into the German immigrant stream.5 In the third quarter of the eighteenth century the illiteracy of adult male residents in the rural areas of Pennsylvania, Virginia, and New England ranged from 37 to 22 percent (Lockridge 1974: 24, 77; Tully 1972: 304). This evidence was based on signatures from wills. Adjusting for the age and wealth biases in the 50 45 Percentage illiterate

40

1727–76

23,988 observations

Men age 16 or older

35 30 25 20 15 10

Illiteracy percentage by year

5 0

1728 1732 1736 1740 1744 1748 1752 1756 1760 1764 1768 1772 1776 Year

Figure 7.1 Illiteracy percentage by year: German immigration to Philadelphia, 1727–75 (source: Derived from Strassburger, 1934, Vols. I and II).

The age structure of German immigrant literacy   133 evidence, German immigrants were at least similar if not less illiterate than these rural colonials.6 Late eighteenth century adult male illiteracy in the cities of New York, Philadelphia, Boston, and Elizabeth, Virginia, ranged between 20 and 10 percent (Cremin 1970: 533, 540; Kilpatrick 1912: 197, 228–9; Lockridge 1974: 24, 77). However, this evidence came from very small samples of wills, deeds, and petitions, which may have systematically selected for the literate in the population.7 Whether these biases were large enough to reverse the apparent 0 to 10 percentage point advantage urban colonial literacy had over German immigrant literacy is unknown. The impact of immigration on total American literacy declined as the native-­ born population grew in relative importance, but it may still have been important by the middle of the eighteenth century. A very crude estimate of the impact of German immigrant literacy on Pennsylvania literacy can be made with the following equation: Total colonial literacy = (native-­born literacy) (share of native-­born in the population) + (immigrant literacy) (share of immigrants in the population). A literacy rate of 48 percent for native born is implied from the 71 percent literacy of German immigrants, the 55 percent total colonial literacy (bias adjusted), and an immigrant share in the population of 30 percent (Tully 1972: 305; Wokeck 1981: 260–1; Department of Commerce 1975: 1168).8 This implies that German immigration may have increased Pennsylvanian literacy by some 15 percent. Earlier in the colonial era the impact of immigration on literacy may have been even greater because the share of immigrants in the population was higher. Therefore, it appears that immigration per se was an important force in propelling the colonies ahead of Europe in terms of literacy. This was because literates had a higher propensity to migrate. To investigate the nature of this positive self-­selection by literates into the German immigrant stream, age-­ specific literacy rates will be analyzed.

Estimating the age structure of literacy Within any generation literacy is a cumulative process. As a generation passes from adolescence to maturity more and more of its members acquire the ability to sign their names, thus raising the measured literacy of the generation. Once literacy is acquired, it is seldom lost. Only in old age might feebleness cause a measurable deaccumulation of signature literacy. This accumulation process is the age structure of literacy. Age-­specific literacy rates can be derived from a subset of the evidence presented in Figure 7.1. Between 1730 and 1754, 64 German immigrant ships recorded separate accounts of the names and ages of the immigrants who signed the loyalty oaths. The two lists were merged to create a set of 4,518 male immigrants, ages 14 to 78, with the distinction of whether or not they could sign their name at the time of immigration.9

134   German immigration to America, 1709–1835 The age structure of literacy is a generation-­specific or life-­cycle phenomenon and should be estimated by measuring the same sample at successive ages for each generation. However, each individual is only observed once, at the point of entry into the colony. Observations are recorded across generations or cross-­ sectionally for each year, and the same sample cannot be measured at successive ages. This mixes life-­cycle effects within each generation with secular trends across generations. Therefore, the age structure of literacy must be estimated by constructing artificial generations or birth cohorts. This is a typical problem for historical data sets used to analyze life-­cycle effects (Wells 1978). The age structure of literacy is estimated using regressions which control the secular trend in literacy across generations. The simplest and most obvious secular control, the year of arrival, produces a biased approximation.10 To correct this bias a secular control, the birth cohort, is constructed by subtracting the age at arrival from the year of arrival to derive the birth year. This allows the regression to observe the literacy of immigrants born in the same year but at different ages as they chose to migrate in different years, thus producing a generation-­ specific estimate of the age structure of literacy. The bias correction can be seen in the following specifications: Corrected specification. Illiteracy = constant + a1 (Age) + a2 (Age2) + a3 (year of arrival – age) + other variables. Biased specification. Illiteracy = constant + a′1 (Age) + a′2 (Age2) + a′3 (year of arrival) + other variables. The negative slope coefficient on Age when the birth cohort is controlled is a1 which is equal to (a′1 + a′3) when only the year of observation is controlled. Therefore, controlling only the year of observation will systematically understate the true decline in illiteracy with age, a1 < a′1, when illiteracy is declining secularly, a′3 < 0. Correctly specifying the secular control is critical for interpreting life-­cycle phenomena. Table 7.1 presents the estimated age structure of literacy for both specifications.11 Controlling only the year of arrival significantly understates the true decline in illiteracy from adolescence to maturity (see Figure 7.2 which plots the age–illiteracy curves estimated in Table 7.1).12 Figure 7.2 also shows that estimating the age–illiteracy curve without secular controls by pooling the observations cross-­sectionally for each age produces an even larger bias understating the decline in the age–illiteracy curve. Seldom can the importance of specification errors be demonstrated so clearly as with the evidence in Figure 7.2. The following discussions will use only the corrected specification.

Interpreting the age structure of literacy The estimated age–illiteracy curve from the corrected specification exhibits a sharp fall in the early ages, from 73 percent illiterate at age 14 to 10 percent illiterate by age 44 or a fall of 2.2 percentage points per age. The decline is long,

The age structure of German immigrant literacy   135 Table 7.1 The age structure of illiteracy for male German immigrants to Philadelphia, 1730–54 Explanatory variables

Birth cohort controlled (corrected specification)

Year of arrival controlled (biased specification)

Constant Age Age2 Birth cohorta Year of arrival Dynastic change in the Palatinateb Total German immigration per yearc Number of observations Iterations to convergence –2 Log likelihood Model (χ2) Correct prediction rate

  0.9452 (0.8001) –0.0640*** (0.0170)   0.0072*** (0.00024) –0.0172** (0.0103)

  25.2759* (17.9865) –0.0432** (0.0212)   0.00065*** (0.00023)

  0.3248** (0.1592)   0.000008 (0.000037) 4,518 5 5,747.0 85.41*** 65.3%

–0.0148* (0.0103)   0.5250*** (0.1366)   0.00007** (0.00003) 4,518 5 5,740.4 92.01*** 65.3%

Source: Strassburger 1934, Vols. I and II, and text. Notes The estimation technique is Logit, cumulative logistic probability function, as computed by the SAS program. Standard errors are in parentheses. The dependent variable is observed illiteracy coded as one if the immigrant made his mark and zero if he signed his name. Significance Levels are based on χ2 tests. a Constructed by subtracting the age at arrival from the year of arrival and forming an index of individuals born in the same year coded from 0 for those born before 1680 to 60 for those born after 1738. The index only alters the constant term from what it would be using the numerical year. This accounts for the different constant terms in the two regressions. b Coded as 1 for the years before 1743 and 0 otherwise, representing the ascendance of Charles Theobald to the throne of the Palatinate (see Schlosser 1844, Vols. III and IV). c Taken from Wokeck 1981: 260–1. * Indicates significance above the 0.16 level. **  Indicates significance above the 0.1 level. ***  Indicates significance above the 0.01 level.

reaching a minimum at age 44. Finally, there is a sharp upswing in the older ages. From age 44 to age 60 illiteracy rises 17 percentage points or about 1 percentage point per year lived.13 Because the estimation samples immigrants born in the same year as they chose to migrate at different ages, i.e., 30-year-­olds on the curve were a different sample than 15-year-­olds, two explanations of the curve are possible. If the propensity of literates to migrate is invariant across age then the immigrants’ age– illiteracy curve would be representative of the home population’s curve. This implies an exceptional amount of literacy acquired after the school-­leaving ages and then lost during old age (Galenson 1981a: 820–1).14 Such an interpretation would alter the accepted explanation of how literacy was transmitted and retained in a preindustrial economy (Bailyn 1960; Craig 1981; Cremin 1970; Lockridge 1974; Soltow and Stevens 1981). Although some literacy was acquired in the post-­teen years outside of formal schools and some was lost in old age, the magnitude of the estimated results

136   German immigration to America, 1709–1835

1730–54 4,518 observations Ages 14–78 Logit regression fits evaluated at means

Percentage illiterate

100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Biased specification (year held constant)

Corrected specification (birth cohort held constant)

Pooled cross-sectional estimate 10

15

20

25

30

35

40 45 Age

50

55

60

65

70

75

Figure 7.2 Age structure of illiteracy: Logit regressions, German male immigration to Philadelphia, 1730–54 (sources: see Table 7.1 and the text).

cannot be fully explained by assuming the immigrants’ age–illiteracy curve was representative of the home population’s curve. Literary sources seldom mentioned postadolescent acquisition of literacy and usually stressed the importance of schools in the transmission of literacy. Pastor Muhlenberg (1942), head of the Lutheran church in eighteenth-­century Pennsylvania, occasionally accepted older teenagers for instruction in reading and writing but indicated that these were extraordinary cases. German servant contracts commonly had provisions for schooling or being taught to read and write for those under age 15 but not for those over age 15 (“Book of Redemptioners” and “Record of Indentures”). The second possible explanation of the estimated age–illiteracy curve assumes that the propensity of literates to migrate varied with age. A falling proportion of illiterates choosing to migrate as their generation approached middle age would cause declining illiteracy among young adults. A rising proportion of illiterates choosing to migrate as their generation passed beyond middle age would cause rising illiteracy among the old. This process would bow or drag the estimated age–illiteracy curve for immigrants down in the middle ages from what would be characteristic for the home population. At 10 percent illiterate, middle-­aged German immigrants were substantially less illiterate than the Rhineland population, implying that immigrant literacy had a strong age-­specific component. Why would literates increasingly select themselves into the migrating population as they passed through their teens and twenties? Young adults typically invested in general forms of human capital, such as literacy and trade skills, which could migrate with them and be readily marketed in a new location. The marginal migrant would offer more labor to the market and consume less leisure

The age structure of German immigrant literacy   137 in the new location relative to the old.15 In addition, premiums were paid in the colonies for possession of literacy, trade skills, etc. (Adams 1970; Galenson 1981b: 97–113). This would raise the relative return to investing in general human capital because more time would be spent in market activities in the new location. Non-­marginal migrants would also invest in non-­location specific skills before migrating, because the alternative of migrating first and then acquiring these skills would have a lower present value in equilibrium, given positive migration costs. Therefore, literates among young adults might tend to have a higher propensity to migrate, other things being equal.16 This may account for the decline in the age–illiteracy curve in the late teens and twenties as independent migration decisions became possible. Why would the relative propensity of literates to migrate decline after middle age? In the extreme older ages immigrants were largely dependents of younger migrants. Their migration objective may have been simply to remain united with offspring and grandchildren rather than any personal economic aim. In addition, there may have been disincentives for literates to migrate after reaching middle age. Literacy in the seventeenth and eighteenth centuries was highly correlated with other rankings of the social structure among older persons (Craig 1981: 170–2; Cressy 1980: 104–41; Galenson 1979: 75–91; 1981a: 813–29; Lockridge 1974: 13–27; Schofield 1973: 450). Therefore, literates tended to make relatively more investments in other forms of human capital as they grew older; many being location specific, such as building up a good business reputation, a high standing in the community, local market relationships, or some other “stake” in the community. Because migration is an investment in location, migrants lose the location-­ specific aspects of their human capital.17 Unlike physical capital such as a house, location-­specific human capital cannot be readily sold before moving or readily purchased in a new location.18 Therefore, the old should be relatively less likely to migrate, both because the benefits of investing in migration are realized over a shorter period and because the old face a greater loss of accumulated stocks of location specific human capital.19 Because literates made significantly more investments in location-­specific human capital as they aged, they might have a declining relative propensity to migrate as they aged. This may explain the rise in the estimated age–illiteracy curve after age 44. Combining the two effects, the rising propensity of literates to migrate early in life with the falling propensity of literates to migrate late in life, implies that a relatively flat age–illiteracy curve for the home population would take on a U shape for immigrants. This outcome was a result of investing in location versus non-­location-specific human capital at different ages, relative to the incentives to migrate.20

Other determinants of immigrant literacy The trend in immigrant literacy was measured across successive generations while controlling for changes in the age structure of the immigrants. The

138   German immigration to America, 1709–1835 estimated growth in literacy was 1.72 percentage points from birth cohort to birth cohort. This rapid pace suggests either rapid growth in literacy in the Rhineland or that the propensity of literates to migrate was rising over time across all age groups. There was little relationship between literacy and the magnitude of migration. However, immigrant literacy increased after Charles Theobald ascended to the throne of the Palatinate in 1742. Theobald was associated with intolerance toward protestants and rising emigration, especially early in his reign (Schlosser 1844, Vols. III–VI; Wokeck 1981). By lowering the prospects of Rhineland protestants, Theobald may have caused the migration of more literate protestants and nonfarmers (Grubb 1985). This may help explain why German immigrant literacy was higher than most other Europeans. Traditionally, farmers and Catholics were from less literate classes (Cressy 1980; Lockridge 1974).21 Finally, the evidence on age and literacy can be used to illustrate the intergenerational persistence of literacy. The sample in Table 7.1 had a subset of 192 fathers with 252 sons. Fathers and sons had about the same literacy rate, 60 percent. The correlation of the literacy of fathers with the literacy of their sons was positive, 0.42, but small.22 The proportion of literate fathers with all literate sons and illiterate fathers with all illiterate sons was 69 percent. This suggests some persistence of literacy across generations within the family, but the persistence was far from absolute.

Conclusions The signature literacy of eighteenth century German immigrants was at least comparable if not better than the established colonial populations and better than the European populations from which they came. This suggests that immigration per se was a major force propelling the colonies past Europe in terms of literacy. The disproportionate propensity of literates to migrate was age specific and may have been caused by an increasing proportion choosing to migrate as they approached middle age, the effect being reversed after middle age. This behavior may have been the result of systematically investing in general human capital in younger ages compared with investing in location-­specific human capital later in life. Finally, the trend in immigrant literacy was rising across generations, suggesting either rapid literacy growth in the Rhineland or a rising propensity of literates to migrate in part because of political turmoil. Immigration was important in raising the level of human capital in colonial America.

Notes   * Reprinted from Farley Grubb, “Colonial immigrant literacy: an economic analysis of Pennsylvania-­German evidence, 1727–1775,” Explorations in Economic History, vol. 24, no. 1 (January 1987): 63–76, Copyright 1987, © 1987, with permission from Elsevier. The author thanks Rick Agnello, John Craig, Stanley Engerman, Robert Fogel, David Galenson, Saul Hoffman, Ken Lewis, Jeff Miller, David Mitch, Jon Moen, Jim

The age structure of German immigrant literacy   139 Mulligan, John Pritchet, and Karl Scheunes for helpful comments on earlier drafts. A version of this paper was presented in the Economic History Workshop, University of Chicago, and in the Economics Seminar, University of Delaware. I thank the participants of these seminars for their comments. © 1987 Academic Press, Inc.   1 The generally accepted and most used measure of literacy is the ability to affix a signature. This is the only universal, standard, and direct measure available for the colonial period that provides a substantial quantity of evidence. Although a dichotomous measure, the ability to affix a signature is believed to correspond to a middle range of literary skills or roughly to the ability to read fluently. See Craig (1981: 168–70), Cressy (1977: 1–3; 1980: 1–19, 42–62), Lockridge (1974: 7), Schofield (1968: 311–25, and 1973: 438–43), and Soltow and Stevens (1981: 3–6).   2 The introductory discussion is based on Bailyn (1960), Craig (1981: 151–74), Cremin (1970: 419–550), Galenson (1979: 78–83), Kilpatrick (1912), Lockridge (1974), Soltow and Stevens (1981: 1–57), Tully (1972: 301–12), Weber (1905), West (1978), and Wright (1957: 98–125). For skepticism regarding the connection between literacy and economic growth see Schofield (1973).   3 See Diffenderffer (1899: 191–2), Grubb (1985), Strassburger (1934), and Wokeck (1981). Oaths were required only of non-­British adult male immigrants before being released from the captain’s custody (no female evidence). They covered 90 to 95 percent of the entire adult male population of Germans immigrating to Pennsylvania. German immigrants signed in German. Literacy in German was not a handicap because Pennsylvania was almost half German. There are no unrepresentative sample selection problems, and the oaths were signed only once. Other studies have relied on signatures from wills, deeds, petitions, and jury lists which suffer from double counting and correlation with other rankings of the social structure, thus producing biased measures and typically overstating the literacy of the population under study (Beales 1978: 94–5; Cremin 1970: 575–6; Cressy 1980: 105–8; Lockridge 1974: 7–14, 133, 152; Soltow and Stevens 1981: 38, 207–8; Tully 1972: 304).   4 Prior estimates of German immigrant illiteracy were even lower (Weber 1905: 14–15; Kilpatrick 1912: 229). Neither of these studies disaggregated literacy by year or related literacy to age. Weber and Kilpatrick also took their evidence from the older and inferior Rupp (1898) compilation of ship lists (see Wokeck 1981: 250–8; and Strassburger 1934, Vol. I: vii–xxvi). When an immigrant made his “mark” instead of his signature the recorder wrote the immigrant’s first name to the left of the mark and his last name to the right of the mark, giving the mark an appearance of a middle initial. A hasty reading of the names may have led Weber and Kilpatrick to undercount illiterates.   5 A direct comparison of German immigrant literacy with the literacy of their home population was not possible because of the lack of signature evidence for the home  population (see Maynes 1977: 291–2). Galenson (1979, 1981a) measured the literacy of English indentured immigrants leaving London between 1718 and 1759. Those above age 15 had a 30 percent illiteracy rate. A subset of 271 bound for Pennsylvania were 32.5 percent illiterate (Kaminkow and Kaminkow 1964; Galenson 1977). Comparing across the same birth cohorts, German immigrants were only a few percentage points more illiterate than these English emigrants. This evidence also suggests some positive self-­selection by literates into the English emigrant stream. Lockridge (1974: 5, 46–8, 74–5, 78, 82) also alludes to a positive self-­selection by literates into the migration stream, suggesting as much as a 10 percentage point differential.   6 See footnote 3. Those with probated wills were relatively wealthier, older, and from a higher occupational class. This group was more literate than the general population. Although wills were made late in life when feebleness may have robbed a person of his ability to sign his name, the wealth bias overstating literacy is generally assumed to dominate.

140   German immigration to America, 1709–1835   7 See footnote 3. Deeds, petitions, and jury lists overstate the literacy of the general population because they sample from the wealthier parts of society and double count due to multiple participations.   8 Tully (1972: 305) lowers his raw estimate to 40–55 percent literate for Pennsylvanians to account for the biases in the will evidence. The implied native-­born literacy estimate may be biased upward because the total colonial literacy estimate may include other immigrant groups and because the estimated immigrant share in the population may be too small considering the relevant populations. Thus the contribution of German immigrant literacy to total literacy may have been even greater.   9 The evidence was derived from Strassburger (1934, Vols I and II) and matched 91 percent of the names on the two lists. The sample comprised 22 percent of all adult male German immigrants and 27.5 percent of all German immigrant ships arriving between 1730 and 1754 (Wokeck 1981: 260–1). The illiteracy of the sample was 34.5 percent, which was representative of the illiteracy rate of all adult male German immigrants in these years. The sample had a mean age of 29 and a median age of 26. Roughly 19 percent of the sample fell between the ages of 16 and 20, 24.6 percent between 21 and 25, 20.6 percent between 26 and 30, 20.7 percent between 31 and 40, 11.2 percent between 41 and 50, and only 4 percent were over age 50. 10 This specification observes the literacy of immigrants of all ages arriving in a given year. It constructs an age–illiteracy curve by comparing the literacy of 40-year-­olds, not with their literacy when they were 16, but with 16-year-­olds arriving in the same year who are from a later, more literate generation. This will systematically understate the decline in illiteracy with age when illiteracy is declining secularly. 11 Above the quadratic term, higher orders of the age polynomial proved to be statistically insignificant for all specifications and were dropped. Because the dependent variable is binary, either signing or making a mark, logit or cumulative logistic probability function was used. Ordinary least squares (OLS) has problems when the dependent variable is binary (Pindyck and Rubinfeld 1976: 237–51). Biased slope coefficients from using OLS can alter the interpretation of the estimates, i.e., the OLS coefficient on Age was biased to over 75 percent below that reported for Logit in table 7.1. 12 The estimated coefficients in table 7.1 do not exactly relate to the algebra of the two specifications, a1 does not exactly equal (a′1 + a′3), because of the approximations used by the computer programs. There are more birth cohorts than years of observations and both could not be controlled in the same regression because of near-­perfect multicolinearity. 13 The results are consistent with Galenson (1981a). He found illiteracy falling with age for English indentured emigrants in their late teens and early twenties at 1.3 percentage points per age in 1683 and 1.8 percentage points per age in 1718–59. However, Galenson biased his estimates by controlling the year of emigration rather than the year of birth. The correct slope coefficient on age can be approximated by adding his Year coefficient to his Age coefficient, raising his estimate by 25 percent or to the same magnitude as those found for German immigrants. 14 School-­leaving age in Germany was around 13 (Maynes 1977: 262). For Tudor and Stuart England the schooling ages were 5 to 15, and boys were assumed to read and write by age 9 to 11 (Cressy 1977: 11). 15 Given positive migration costs and using the standard labor–leisure model, the marginal migrant must have the higher wage or income opportunity set of the new country tangent to the same indifference curve as the income opportunity set of the home country, resulting in less consumption of leisure in the new country despite being indifferent between the two locations. If nonleisure time is assumed to be market time then the marginal migrant will offer more labor to the market in the colonies than he will if he stays in the home country. The effect becomes larger as the cost of migration increases.

The age structure of German immigrant literacy   141 16 The positive connection between education and the propensity to migrate has been well established. However, the age-­specific aspects of the connection have not been investigated. The relationship is usually attributed to lower information costs faced by more educated individuals (see Greenwood 1975: 406–7). 17 This analysis is based on the distinction between general and specific human capital (see Becker 1980: 19–40). Investment in general human capital involves acquiring skills or knowledge that can be readily sold in the market place to any number of buyers or in any number of locations. Investment in specific human capital involves acquiring skills or knowledge whose possession only has value to a particular firm, person, or in a particular location. 18 This analysis may also explain the higher propensity of blue-­collar workers to migrate relative to white-­collar workers in pre-­twentieth century America (Thernstrom 1973: 39–42). If white-­collar workers made relatively more investments in location-­specific human capital in this era, they would have a lower propensity to migrate. The falling cost of reliable long-­distance information in the twentieth century has decreased the location-­specific elements to white-­collar investments in human capital thus making white-­collar workers more mobile. 19 The effects of age on lowering the returns to migrating by shortening the period of benefits are well established. However, the connection to different types of investment in human capital has not been generally made. Costs similar to losing investments in location-­specific human capital have been placed under the rubric of “psychic costs” such as leaving friends or family (see Greenwood 1975: 404–7). 20 The process can be directly observed among German religious ministers migrating to Pennsylvania in the eighteenth century. Ministers invested in education or general human capital early in life and then migrated in search of a congregation. Once established with a congregation, location-­specific investments in human capital were made by building up a rapport, reputation, and following with the congregation. Thus older ministers were reluctant to migrate, unless it was with their congregation. Only young ministered migrated without a flock. A number of examples can be found in Strassburger (1934) and Muhlenberg (1942). 21 The proportion of farmers among German immigrants fell from around 63 percent in 1709 to approximately 26 percent by 1800, and the proportion of Catholics was around 8.9 percent (see Grubb 1985). Radical protestant sects were even more literate than mainstream protestants. The colony of Schwenkfelders arriving on the St. Andrew in 1734 had a 26 percent illiteracy rate, and the colony of Moravians arriving on the Catherine in 1742 had a 7 percent illiteracy rate. The spot arrival of some of these radical sects in 1729, 1730, 1733, 1734, 1741, and 1742 may explain some of the variance in the illiteracy rates measured in figure 7.1 (Strassburger 1934). 22 The correlation coefficient is statistically different from zero above the 0.0001 level, a t statistic of 6.39. The criteria used to assert a father–son relationship were the same last name, at least 20 years apart in age, and listed alongside each other on the lists. The average age of fathers was 50 and the average age of sons was 20.

References Adams, D. R., Jr. (1970), “Some Evidence on English and American Wage Rates, 1790–1830.” Journal of Economic History 30, 499–520. Bailyn, B. (1960), Education in the Forming of American Society. Chapel Hill: University of North Carolina Press. Beales, R. W., Jr. (1978), “Studying Literacy at the Community Level: A Research Note.” Journal of Interdisciplinary History 9, 93–102. Becker, G. S. (1980), Human Capital, 2nd ed. Chicago: University of Chicago Press.

142   German immigration to America, 1709–1835 Book of Redemptioners, 1785–1804. Manuscript held at the Pennsylvania Historical Society, Philadelphia. Craig, J. E. (1981), “The Expansion of Education.” Review of Research in Education 9, 151–213. Cremin, L. A. (1970), American Education, The Colonial Experience 1607–1783. New York: Harper and Row. Cressy, D. (1977), “Levels of Illiteracy in England, 1530–1730.” Historical Journal 20, 1–23. Cressy, D. (1980), Literacy and the Social Order. Cambridge, MA: Cambridge University Press. Department of Commerce (1975), Historical Statistics of the United States Colonial Times to 1970. Washington, D.C.: Dept. of Commerce. Diffenderffer, F. R. (1899), “The German Immigration into Pennsylvania Through the Port of Philadelphia, and ‘The Redemptioners’,” pt. 7 of “The German Influence in its Settlement and Development.” Pennsylvania German Society 10, 1–328. Galenson, D. W. (1977), “Agreements to Serve in America and the West Indies, 1727–1731,” Genealogists’ Magazine 19, 40–4. Galenson, D. W. (1979), “Literacy and the Social Origins of Some Early Americans.” Historical Journal 22, 75–91. Galenson, D. W. (1981a), “Literacy and Age in Preindustrial England: Quantitative Evidence and Implications.” Economic Development and Cultural Change 29, 813–39. Galenson, D. W. (1981b), White Servitude in Colonial America. Cambridge, Mass.: Cambridge Univ. Press. Greenwood, M. J. (1975), “Research on Internal Migration in the United States: A Survey.” Journal of Economic Literature 13, 397–433. Grubb, F. (1985), The Social Origins of Eighteenth-­Century German Immigrants. Paper presented at the annual conference of the Society for Historians of the Early American Republic, Georgetown University, University of Delaware (mimeo). Kaminkow, J. and Kaminkow, M. (1964), A List of Emigrants From England to America 1718–1759. Baltimore: Magna Charta. Kilpatrick, W. H. (1912), The Dutch Schools of the New Netherlands and Colonial New York. Washington, D.C.: Govt. Printing Office. Lockridge, K. A. (1974), Literacy in Colonial New England. New York: Norton. Maynes, M. J. (1977), Schooling the Masses: A Comparative Social History of Education in France and Germany, 1750–1850. Ph.D. dissertation, Univ. of Michigan. Muhlenberg, H. M. (1942), The Journals of Henry Melchior Muhlenberg. Philadelphia: Muhlenberg Press. Pindyck, R. S. and Rubinfeld, D. L. (1976), Economic Models and Economic Forecasts. New York: McGraw-­Hill. Record of Indentures of Individuals Bound Out as Apprentices, Servants, etc. and of Germans and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773. Manuscript held at the City Archives of Philadelphia. Rupp, I. D. (1898), A Collection of 30,000 Names of German and Other Immigrants to Pennsylvania, 1727–1776, 2nd ed. Philadelphia. Schlosser, F. C. (1844), History of the Eighteenth Century. London: Chapman and Hall. Schofield, R. S. (1968), “The Measurement of Literacy in Pre-­Industrial England.” In Literacy in Traditional Societies (J. Goody, ed.): 311–25. Cambridge, MA: Cambridge University Press.

The age structure of German immigrant literacy   143 Schofield, R. S. (1973), “Dimensions of Illiteracy, 1750–1850.” Explorations in Economic History 10, 437–54. Soltow, L. and Stevens, E. (1981), The Rise of Literacy and the Common School in the United States. Chicago: University of Chicago Press. Sparks, J. (ed.) (1838), The Works of Benjamin Franklin. Boston: Tappan and Whittemore. Strassburger, R. B. (1934), Pennsylvania German Pioneers (W. J. Hinke, ed.). Norristown: Pennsylvania German Society. Thernstrom, S. (1973), The Other Bostonians. Cambridge, MA: Harvard University Press. Tully, A. (1972), “Literacy Levels and Educational Development in Rural Pennsylvania 1729–1775.” Pennsylvania History 39, 301–12. Weber, S. E. (1905), The Charity School Movement in Colonial Pennsylvania. Philadelphia: G. G. Lasher Press. Wells, R. V. (1978), “On the Danger of Constructing Artificial Cohorts in Times of Rapid Social Change.” Journal of Interdisciplinary History 9, 103–11. West, E. G. (1978), “Literacy and the Industrial Revolution.” Economic History Review 31, 369–83. Wokeck, M. (1981), “The Flow and Composition of German Immigration to Philadelphia, 1727–1775.” Pennsylvania Magazine of History and Biography 105, 249–78. Wright, L. B. (1957), The Cultural Life of the American Colonies 1607–1763. New York: Harper and Row.

8 Educational choice in the era before free public schooling*

The growth of literacy and the expansion of the educational system was important to the formation of human capital, the socialization of the American work force, and the economic development of nineteenth-­century America.1 How, when, and why education expanded is not well understood, however. Mass literacy among white males (more than 65 percent in rural areas and 80 percent in urban areas) had been achieved in most colonies by the Revolution. Although less is known about white female literacy, it appears to have lagged behind male literacy in most colonies prior to the Revolution. By the 1840 census, the first to measure literacy, universal literacy (more than 90 percent) had been achieved in the northeast by white males and females in both urban and rural locations.2 Exactly when female literacy caught up with male literacy—and when rural literacy achieved parity with urban—is unknown. How much literacy was achieved through formal schooling—as opposed to informal instruction at home or at work—is likewise unknown. Neither do we know when formal schooling became available and when it replaced informal instruction. Recent research has suggested that the increase in female literacy, rural literacy, and formal schooling predated the common-­school reformation movement. Tax-­ supported free public schools were uncommon before 1840. Given that mass literacy was achieved before the Revolution largely without government assistance, and that universal literacy was achieved by 1840 without free public schooling, an expansion of pay- or tuition-­supported common schools may deserve the credit. Although the early Republic appears to have been a period of educational transformation, the paucity of quantifiable evidence makes the timing and character of that transformation difficult to identify. This paucity of evidence also allowed reformers in the early Republic to claim that the provision of schooling, and hence education, was sorely lacking, particularly among immigrants and the poor.3 Most of these questions cannot be resolved with the quantitative evidence typically used to study education: literacy rates, school enrollment rates, and the number of schools, books, and newspapers per capita.4 Signature literacy rates and books per capital do not reveal how or where literacy was acquired. Enrollment rates do not reveal what share of those attending school nevertheless acquired literacy at home or whether those not attending school nevertheless did so elsewhere.5

Educational choice before free public schooling   145

New evidence from immigrant servant contracts Evidence on the range of educational choices can be found in servant contracts for German children arriving in Philadelphia between 1771 and 1804. Roughly 45 percent of German immigrants in this period sold themselves into servitude to finance their voyage, and children comprised 25 percent of their number. The salient features of these servant contracts were recorded by Philadelphia magistrates for the periods from 1771–3, 1785–1804, and 1817–31. German servants were redemptioners who negotiated their labor contracts directly with prospective American masters on landing in Philadelphia. Parents negotiated their child’s contract, after which the child left the family (probably for the first time) and entered an employer’s household for 6 to 13 years. The amount and type of education negotiated into the child’s contract may have been the last influence parents had over their child’s future. The child paid for the education by serving a longer contract: 22 percent longer, other things being equal. Because literacy added little to an adult servant’s value, it was largely a consumption good; thus the expense of education was the servant’s responsibility, not the master’s.6 Although the contract evidence is limited in size, scope, and representativeness, all other quantitative data on education from this period suffer from the same problem. This evidence does have certain unique advantages, though. First, it covers new arrivals, arguably the most alien of the large free migrations to America during the early Republic. Second, immigrants sold themselves into servitude only when they lacked enough accumulated wealth to pay for their transatlantic passage; so these were all relatively poor. Given these facts, this evidence can be used to investigate the crisis in the education of immigrants and the poor asserted by contemporary reformers. Third, the evidence spans the Revolution and includes females. It may thus help determine when female literacy reached parity with that of males, and what impact republican fervor had on educational change. Finally and most important, the evidence identifies different methods of education—namely, formal schooling versus informal instruction outside of schools—that can be disaggregated by sex, age, location, and time span. Education clauses in the servant contracts were distinct from those stipulating that occupational training be provided. For males under the age of 17, training was specified in only 7 percent of the contracts. In 82 percent of those that did stipulate training, however, separate education clauses were included for the acquisition of literacy. Thus, occupational training was different from and not a substitute for being taught to read and write. Immigrant male children typically completed their initial servant contract, which paid for their passage to America, before entering a new apprenticeship contract to learn their chosen occupation.7 Whether employers fulfilled their contractual obligations and what quality of education they provided is unknown, but the contracts were recorded by the government and were legally binding. Private organizations such as local German societies oversaw contract compliance. Cases of contract noncompliance that found their way into newspapers and courts usually involved the servant running

146   German immigration to America, 1709–1835 away or the master providing deficient freedom dues or excessive punishment. Failure to provide the contracted education was seldom mentioned. This kind of servant and apprenticeship contracting had persisted throughout society for centuries, which suggests that contract fraud could not have been pervasive.8 Other education evidence such as schooling enrollment rates likewise reveals little about the quantity and quality of education provided in school or even whether the child was becoming literate at school rather than at home. Employers had no financial incentive to provide more education and legally could not provide a different form of education to their servants than that directly contracted for. Thus, the contract evidence may represent a good measure of the relative incidence of different methods of education. Types of education contracted for by immigrant servants Table 8.1 presents the relative amounts and methods of instruction specified in the contracts of German immigrant servant children for 1771–3 and 1787–1804, disaggregated by sex and their master’s location. The total percentage of children receiving some form of employer-­provided education was 34 percent between 1771 and 1773. This percentage rose to 70 percent in the decades just after the Revolution. The last decades of the eighteenth century therefore appear to have been a period of considerable expansion in the provision of education. Given the probability that some children acquired literacy before emigrating, the total percentage of children contracting for education was high, and the demand for education among these immigrants was strong. Table 8.1 also presents a percentage distribution of the methods of instruction. In some cases only reading was stipulated, usually phrased as “to be taught to read in the Bible.” Writing was never stipulated by itself, only in conjunction with reading. It was usually phrased as “to be caused to write a legible hand.” Arithmetic, or ciphering, was specified in only a few cases. It always appeared in conjunction with reading and writing, and was phrased as “to be taught to cipher to the rule of three.” (The rule of three was being able to do proportions.)9 The implied sequence of instruction—reading first, writing second, and ciphering last—supports the assumptions made about education in most literacy studies: that being able to affix a signature represented some middle range of literacy skill, or at least the ability to read.10 Alternatively, many contracts stipulated formal schooling—evening, day, or winter school. A specific number of months was always included: “nine months’ evening school,” for example. Because formal schooling was stipulated when it was the desired method of education, the lack of its stipulation in an education clause implies some informal or nonschool method, such as being instructed directly by the employer. (Formal schooling and informal instruction were never stipulated together in the same contract.) The geographic pattern of these two educational methods has cost implications that will be tested further on. A considerable amount of literacy was acquired outside of formal schools. The total number of children being taught to read and write informally by their

0.3 0.0 0.0 0.8

13.5 9.4 14.3 18.2

1.7 0.9 4.2 2.4

1787–1804 Total [N = 365] Philadelphia [N = 184] Other towns [N = 53] Countryside [N = 128]

FEMALES 1771–3 Total [N = 126] Philadelphia [N = 64] Other towns [N = 7] Countryside [N = 55]

1787–1804 Total [N = 173] Philadelphia [N = 107] Other towns [N = 24] Countryside [N = 42] 20.8 21.5 20.8 19.1

19.1 15.6 14.3 23.6

20.6 21.1 20.8 19.1

21.8 22.4 33.3 20.0

Reading and writing only

0.0 0.0 0.0 0.0

0.8 0.0 0.0 1.8

4.9 4.9 0.0 1.6

1.7 0.0 0.0 2.9

Reading, writing, and ciphering only

48.0 (9.5) 41.1 (9.7) 66.7 (9.2) 54.8 (9.8)

2.4 (10.0) 1.6 (6.0) 14.3 (12.0) 1.8 (12.0)

43.3 (12.4) 43.3 (12.5) 41.5 (14.0) 46.1 (11.6)

6.6 (6.4) 12.9 (7.1) 5.6 (6.0) 2.9 (4.5)

Education in formal schools (months)

70.5 63.5 91.7 76.2

35.7 26.6 42.9 45.5

69.0 69.6 69.8 68.0

33.3 38.8 50.0 27.9

Total receiving any education

2.13 1.83 2.67 2.56

0.07 0.06 0.50 0.04

1.68 1.51 1.47 2.11

0.25 0.50 0.13 0.11

Ratio of school to informal education

Notes Philadelphia refers to masters residing in the city. Employers residing in Philadelphia County outside of Philadelphia City but not in other towns were included in the countryside category. The average number of months of schooling contracted for is in parentheses. Immigrant servants under age 16 were considered children, which corresponded to contract lengths of five or more years. See text; Grubb, “The Auction of Redemptioner Servants,” 595–9; and Grubb, “The Long-run Trend.”

Sources: “Record of Indentures”; and “Book A of Redemptioners.”

3.3 3.5 11.1 2.1

MALES 1771–3 Total [N = 243] Philadelphia [N = 85] Other towns [N = 18] Countryside[N = 140]

Reading only

Informal education

Table 8.1 Education specified in servant contracts for German immigrant children in Pennsylvania, 1771–1804 (in percentages)

148   German immigration to America, 1709–1835 employers stayed constant at 20 percent between 1771 and 1804. Relative to other methods of education, being taught to read and write informally represented 67 percent of all education stipulations in the 1770s but fell to 30 percent by the 1790s, as formal schooling became more common. Employer instruction in reading alone was the second-­most-common type of education in the 1770s for females. Its relative incidence also lessened after the Revolution. Ciphering instruction by the employer was of minor importance and almost exclusively provided to males. Relative to other methods of education, formal schooling accounted for only 18 percent of all education stipulations in the 1770s, but rose to 64 percent by the 1790s. The duration of schooling was typically short, averaging between six and 12 months—perhaps just long enough to teach the elements of reading and writing but not much more.11 The expansion in education between the 1770s and 1790s must be accounted for by an expansion in formal schooling. Because the percentage of all children receiving informal instruction stayed constant over time, the increase in the proportion of servants receiving education had to occur through a net addition of formal schooling stipulations, not through a substitution of formal schooling for informal instruction. In addition, the average duration of schooling increased for males from six months in the 1770s to roughly a year by the 1790s. There must have been a significant expansion of the availability of schools in the Delaware Valley in the first two decades after the Revolution.12 The magnitude of informal education in the 1770s indicates that high levels of colonial literacy could have been achieved without extensive reliance on organized schools; however, the latter may have been important for the final surge to universal literacy. By 1817 the provision for education in immigrant servant contracts for German children had been standardized to six weeks’ schooling for every year of service. Informal education clauses do not appear in the contracts after 1817.13 Geographic cost factors and the choice of educational methods The geographic distribution of educational methods may provide further insight into the educational changes occurring during the early Republic. If formal schooling and informal education were equally capable of transmitting literacy, then which one employers chose may have depended on their circumstances. Both rural and urban employers competed for the same servants, who were sold directly off ships docked in Philadelphia.14 Presumably, employers who lived far from schools, such as farmers in the countryside, would only agree to provide formal schooling if they were compensated for the lost work time, tuition, and other expenses of sending the servant to a distant school. That compensation would have been a longer contract.15 Because servants preferred a shorter contract, other things being equal, these employers would offer the servant an equally effective but cheaper substitute for formal schooling—for instance, informal instruction in reading and writing by someone in their household. This substitute would cost less than schooling and thus require less extra service time from the servant. The offer would enhance the ability of rural employers to

Educational choice before free public schooling   149 compete with urban employers for the same servants. If rural masters and their servants systematically chose informal education over formal schooling, then informal instruction must have had a relative cost advantage in the countryside. Employers who had schools nearby, such as those who lived in large towns, had the same options. To compete with rural employers, they would offer ser­ vants the least costly method of providing the same effective education, thus requiring the least amount of extra service time from the servant. If urban employers and their servants chose formal schools over informal instruction, then formal schools must have had a relative cost advantage in urban areas. Table 8.1 separates the evidence into servants bound to masters in Philadelphia, other towns, and the countryside. Although the percentage of males receiving any kind of education in the countryside lagged behind that for urban areas in the 1770s, informal education was proportionally more common in the countryside than in towns. The choice of educational methods in small towns fell between that found in Philadelphia and the countryside, suggesting that small towns had only recently acquired access to formal schools. For example, the Borough of Chester did not advertise that it had “erected a convenient schoolhouse” until 1771.16 Assuming that the cost of informal instruction was similar in urban and rural areas, urban areas’ choice of more education overall and of proportionally more formal education in particular indicates that (1) the cost of education in general was lower in urban areas and (2) the cost of formal schooling in particular was lower in urban areas than in rural areas as late as the 1770s.17 At least for males, rural areas caught up with urban in both level of instruction and relative use of schools by the 1790s, suggesting that the urban cost advantage in education and formal schooling disappeared soon after the Revolution. A significant increase in the availability of schools in the countryside must have occurred in the first two decades after the Revolution. Perhaps population densities increased enough in the rural counties surrounding Philadelphia to realize economies of scale for formal schooling. The increasing acceptance of female school-­teachers, who were paid less than males, may also have lowered rural schooling costs after the Revolution.18 These rural/urban findings support a supply-­driven interpretation of the nonlinear long-­run trend in colonial literacy rates. In colonial America literacy rates typically fell after the initial settlement of a region before resuming positive secular growth. The evidence on rural versus urban educational methods is consistent with a supply interpretation of this initial “creolean degeneracy” in literacy.19 Formal schooling was so costly in the sparsely settled countryside that informal instruction was substituted for it in an effort to economize on education costs. Because this substitution did not reduce the cost of education to that of more densely settled areas, the quantity of literacy acquired in newly colonized areas was less. Because informal education in the countryside was still more expensive than schooling in urban areas, parents that moved to recently settled, sparsely populated colonial areas transmitted lower amounts of literacy to their children than they had typically received as children in Europe.

150   German immigration to America, 1709–1835 The effect of age and gender on educational choice The evidence can also suggest the age at which instruction was most likely. Specific ages were only recorded in a few contracts. For children the length of service specified was inversely correlated with age, declining almost one for one between ages 3 and 15.20 Using contract length as a proxy for age, Table 8.2 presents the distribution of education stipulations per age for the 1787–1804 sample. For males over 16 and for females over 12, few contracts stipulated any form of education. The few that did usually specified being taught English, indicating that those youth may already have been literate in German. The proportion of younger males contracting for education was fairly constant—77 to 84 percent between the ages of 3 and 10—before steadily declining to 26 percent by age 16. The decline in the proportion of younger females contracting for education was more rapid with the rise in age. It fell from 86 to 64 percent between the ages of 6 and 10, then to 23 percent by age 12, and finally to zero thereafter. Servant contracts did not state the age at which the servant would receive the education. Because children had long contracts, it is difficult to determine the typical age at which instruction would have been given. The relationship between contract length and age indicates that males under 12 served until they were about 19, whereas females under 12 served until they were about 17. Contract lengths leveled off at about three years by age 20. Therefore, education was almost never stipulated after age 20. There was no adult education. Informal instruction coexisted with formal schooling for children and teenagers, but was not a method by which post-­school-age teens and adults learned to read and write. The proportion of formal schooling to informal instruction shows no consistent age pattern. The method chosen was a matter of cost rather than of age. In a steady state equilibrium, subtracting the proportion contracting for education at each age from the adult literacy level measures the proportion of children who had already acquired education before emigrating to America and therefore did not need it stipulated in their contracts. The percentage of contracts stipulating education stayed constant at roughly 80 percent for children up to age 10. Therefore, 80 percent represents the steady state education level achieved by that population. This level also matches the estimated adult literacy rate of that population.21 Subtracting the percentage contracting education for those over age 10 from this 80 percent level reveals the percentage at each age that had already acquired literacy before emigration. This algorithm indicates that, by the age of 12, roughly 21 percent of the males who would acquire literacy had done so prior to emigration; by age 13, roughly 29 percent; by age 15, roughly 49 percent; and by age 16, roughly 67 percent. Comparable female figures are roughly 19 percent by age 10 and roughly 71 percent by age 12. Consequently most instruction must have been given to males between 12 and 16 and to females between 10 and 12.22 In addition, though the proportion contracting for formal schooling does not change much before age 14, the duration of schooling declines steadily with age, from 18.5 months’ schooling at age 6 to about 9.5 months’ at age 13 for males, and from 12 months’ at age 6 to about seven

Educational choice before free public schooling   151 Table 8.2 Education and age of German immigrant servants in Pennsylvania, 1787–1804 (in percentages) Estimated age

Informal education Reading only

Reading and writing only

Reading, writing, and ciphering only

Education in formal schools (months)

Total receiving any education

Adults (males over 16 and females over 12): Males [N = 1,802] 0.1 0.3 Females [N = 605] 0.0 0.2

0.1 0.0

  2.2 (5.7)   0.3 (7.5)

  2.7   0.5

Ages: 15–16 Males [N = 75]

0.0

5.3

1.3

20.0 (8.1)

26.7

Ages: 13–15 Males [N = 74]

0.0

9.5

0.0

31.0 (9.0)

40.5

Ages: 12–13 Males [N = 49]

0.0

6.1

2.0

49.0 (9.8)

57.1

Ages: 10–12 Males [N = 38] Females [N = 102]

0.0 1.0

13.2 2.9

0.0 0.0

50.0 (9.3) 19.6 (6.8)

63.2 23.5

Ages: 9–10 Males [N = 40] Females [N = 31]

0.0 0.0

20.0 19.4

15.0 0.0

42.5 (10.4) 45.2 (9.5)

77.5 64.5

Ages: 8–9 Males [N = 30] Females [N = 22]

0.0 0.0

13.3 9.1

16.7 0.0

50.0 (10.0) 63.6 (8.1)

80.0 72.7

Ages: 7–8 Males [N = 37] Females [N = 20]

2.7 0.0

24.3 25.0

5.4 0.0

48.7 (14.2) 55.0 (10.4)

81.1 80.0

Ages: 6–7 Males [N = 25] Females [N = 18]

4.0 0.0

36.0 11.1

4.0 0.0

40.0 (15.9) 66.7 (9.0)

84.0 77.8

Ages: 6 and under Males [N = 90] Females [N = 51]

0.0 3.9

34.4 41.2

3.3 0.0

41.1 (18.5) 41.2 (11.9)

79.0 86.3

Sources: See sources for Table 8.1. Notes Age can be approximated by contract length in the following manner: contracts shorter than five years correspond to males over 16 and females over 12, contracts of six years to males age 13 to 15 and females age 10 to 11, seven years to males age 12 to 13 and females age 9 to 10, eight years to males age 10 to 12 and females age 8 to 9, nine years to males age 9 to 10 and females age 7 to 8, 10 years to males age 8 to 9 and females age 6 to 7, 11 years to males age 7 to 8 and females age 5 to 6, 12 years to males age 6 to 7 and females age 5, and contracts over 12 years to males under age 6 and females under age 5.

152   German immigration to America, 1709–1835 months’ by age 12 for females. This decline in the duration of schooling is consistent with most schooling being received before age 16. The extent of female education is surprising. The evidence indicates rough equality in the provision of education to both sexes by the early 1770s. There is little evidence here of the traditional female lag in the acquisition of literacy. Parity may have been a recent achievement, however. A schooling advertisement in the Pennsylvania Gazette in 1771 boldly emphasized that females would be admitted along with the usual male scholars, as though this were an unusual or new development.23 The only significant gender difference in the data was in the type of instruction. A higher proportion of females than males was instructed in reading only, and females were almost never instructed in ciphering. Therefore, measures of signature literacy may not accurately measure the difference in literacy rates between males and females, because a higher percentage of females may have been taught to read but not write.24

Conclusion Because the expansion in education between 1771 and 1804 was observed among new arrivals of both sexes, and because female education had nearly reached parity with male education by the 1770s, it is difficult to ascribe this educational expansion solely to a Revolution-­inspired patriotic demand response or to an ideology of republican motherhood response among American-­born citizens. The high rates of education among these immigrant servants does not support the crisis in the education of immigrants and the poor asserted by contemporary education reformers. If there were such a crisis after the 1820s, it may have been due in part to the disappearance of immigrant servitude.25 Ser­vitude may have played an important role in educating, integrating, and socializing poor immigrants to American society.

Notes   * Originally published as Farley Grubb, “Educational choice in the era before free public schooling: evidence from German immigrant children in Pennsylvania, 1771–1817,” Journal of Economic History, vol. 52, no. 2 (June 1992): 363–75. Copy­ right 1992. © 1992 by Cambridge University Press and the editors of The Journal of Economic History. Reprinted with permission. The Journal of Economic History, Vol. 52, No. 2 (June 1992). © The Economic History Association. All rights reserved. ISSN 0022–0507. The author is Associate Professor of Economics at the University of Delaware, Newark, DE 19716. Preliminary versions of this article were presented in seminars at the University of Delaware and the University of Pennsylvania. The author thanks Joseph P. Ferrie, Claudia Goldin, Lynn Lees, Saul Hoffman, T. V. Jackson, David Mitch, John Monroe, Anne Pfaelzer De Ortiz, and seminar participants for helpful comments.   1 For example, see Field, “Educational Expansion”; Fishlow, “Levels of Ninteenth-­ Century American Investment in Education”; and Graff, The Labyrinths of Literacy.   2 See Grubb, “Growth of Literacy”; Perlmann and Shirley, “When Did New England Women Acquire Literacy?”; and Soltow and Stevens, The Rise of Literacy.

Educational choice before free public schooling   153   3 See sources in notes 2 and 5; and Cremin, American Education: The National Experience.   4 See sources in note 2; Cremin, American Education: The Colonial Experience: 225–48, 517–43; and Gilmore, Reading Becomes a Necessity.   5 Early nineteenth-­century reformers claimed that schools were more interested in discipline and socialization than in teaching literacy. They also stressed the importance of parental instruction. See Fishlow, “The American Common School”; and Kaestle, Pillars of the Republic.   6 See Grubb; “The Incidence of Servitude”; Grubb, “The Auction of Redemptioner Servants”; Grubb, “The Long-­run Trend”; and Grubb, “Growth of Literacy.”   7 This pattern is consistent with the age at which education occurred, as discussed later. See also Morris, Government and Labor: 378–83.   8 Ibid.: 381, 390–512; Jacoby, “The Transformation of Industrial Apprenticeship”; and Steinfeld, The Invention of Free Labor.   9 See the description in Cremin, American Education: The Colonial Experience: 501–3; and in Kaestle, Pillars of the Republic: 17. 10 See Cressy, Literacy and the Social Order: 1–19, 42–62; and Schofield, “The Measurement of Literacy.” 11 On the typical length of formal schooling see Cressy, Literacy and the Social Order: 19–61; Kaestle, Pillars of the Republic: 15–16, 111; and Soltow and Stevens, The Rise of Literacy: 89–147. 12 These findings are consistent with those of Jensen, Loosening the Bonds: 167–83, for Pennsylvania, and with evidence on New York and Massachusetts enrollment rates in Kaestle, Pillars of the Republic: 25–9. 13 Derived from “Book C of Redemptioners.” By contrast, a sample of 264 non-­ immigrant servants in Delaware between 1818 and 1839 shows that whereas 84 percent contracted for some type of education, only a third were educated in formal schools. See Jensen, Loosening the Bonds: 227. 14 On the auction of these servants see Grubb, “The Auction of Redemptioner Servants.” 15 To the extent that literacy was not a consumption good but an investment good it was an investment in general human capital, which lasted beyond the servant’s brief bondage and would be useful to many future employers. Therefore, the servant and not the employer would pay the cost of instruction. See the sources in note 6; and Becker, Human Capital: 15–45. 16 Pennsylvania Gazette, December 26, 1771. Commercial agriculture was well developed in Chester County, which only ranked behind Philadelphia and possibly Lancaster counties in the employment of immigrant servants. The rural/urban differential in the method of instruction was not as pronounced for females. Females were less likely than males to be purchased by farmers and more likely to be purchased by storekeepers, innkeepers, merchants, and residents from densely populated rural areas such as Philadelphia County. Thus, females may have been located closer to rural churches and schools than males. See Grubb, “Immigrant Servant Labor.” 17 Informal education is assumed to involve individual instruction in the home or workplace. The primary cost would be the opportunity cost of instruction time for teachers, students, and employers. Whether rural or urban employers had more marginal idle time for home instruction is unclear. The common presumption that farm labor was more idle during the winter off-­season than urban labor may be incorrect. Farmers had winter chores in preparation for spring planting and alternative home production in winter. For example, see Jensen, Loosening the Bonds: 36–56, 79–91. Urban employers had considerable nonproductive marginal time, given customary business hours and the seasonality of trade. Because the net difference in marginal idle time between farmers and urban employers is unclear, the opportunity cost of home instruction was assumed constant across locations.

154   German immigration to America, 1709–1835 18 Grubb, “Growth of Literacy,” 461–76; Jensen, Loosening the Bonds: 167–83; and Kaestle, Pillars of the Republic: 123–34. 19 See Grubb, “Growth of Literacy,” 461–75. 20 See Grubb, “The Auction of Redemptioner Servants,” 595–9; and Grubb, “The Long-­ run Trend.” 21 See Grubb, “Colonial Immigrant Literacy”; and Grubb, “German Immigration to Pennsylvania.” 22 See Kaestle, Pillars of the Republic: 107; Maynes, Schooling in Western Europe: 83–102; and Soltow and Stevens, The Rise of Literacy: 89–147. 23 Pennsylvania Gazette, September 19, 1771. 24 On female education see Jensen, Loosening the Bonds: 167–83; Kaestle, Pillars of the Republic: 29, 62; and Perlmann and Shirley, “When Did New England Women Acquire Literacy?” 25 On the disappearance of immigrant servitude, see Erickson, “Why Did Contract Labour Not Work?” 34–46; and Grubb, “The Disappearance of European Immigrant Servitude.”

References Becker, Gary S., Human Capital (2nd edn., Chicago, 1980). “Book A of Redemptioners, 1785–1804” (unpublished manuscript, Pennsylvania Historical Society, Philadelphia). “Book C of Redemptioners, 1817–1831” (unpublished manuscript, Pennsylvania Historical Society, Philadelphia). Cremin, Lawrence A., American Education: The Colonial Experience, 1607–1783 (New York, 1970). Cremin, Lawrence A., American Education: The National Experience, 1783–1876 (New York, 1980). Cressy, David, Literacy and the Social Order (Cambridge, MA, 1980). Erickson, Charlotte, “Why Did Contract Labour Not Work in Nineteenth-­Century United States?” in Shula Marks and Peter Richardson, eds., International Labour Migration, Historical Perspectives, Vol. 24 (Hounslow, England, 1984): 34–56. Field, Alexander James, “Educational Expansion in Mid-­Nineteenth-Century Massachusetts: Human-­Capital Formation or Structural Reinforcement?” Harvard Educational Review, 46 (Nov. 1976): 521–52. Fishlow, Albert, “The American Common School Revival: Fact or Fancy?” in Henry Rosovsky, ed., Industrialization in Two Systems (New York, 1966): 40–67. Fishlow, Albert, “Levels of Nineteenth-­Century American Investment in Education,” this Journal, 26 (Dec. 1966): 418–36. Gilmore, William J., Reading Becomes a Necessity of Life (Knoxville, 1989). Graff, Harvey J., The Labyrinths of Literacy (London, 1987). Grubb, Farley, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, 22 (July 1985): 316–39. Grubb, Farley, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­Atlantic Economy,” Social Science History, 9 (Summer 1985): 249–75. Grubb, Farley, “Colonial Immigrant Literacy: An Economic Analysis of Pennsylvania-­ German Evidence, 1727–1775,” Explorations in Economic History, 24 (Jan. 1987): 63–76.

Educational choice before free public schooling   155 Grubb, Farley, “The Auction of Redemptioner Servants, Philadelphia, 1771–1804: An Economic Analysis,” this Journal, 48 (Sept. 1988): 583–603. Grubb, Farley, “The Disappearance of European Immigrant Servitude in the United States: An Economic Analysis of Market Collapse, 1785–1835” (Department of Economics Working Paper No. 89–3, University of Delaware, 1988). Grubb, Farley, “German Immigration to Pennsylvania, 1709 to 1820,” Journal of Interdisciplinary History, 20 (Winter 1990): 417–36. Grubb, Farley, “Growth of Literacy in Colonial America: Longitudinal Patterns, Economic Models, and the Direction of Future Research,” Social Science History, 14 (Winter 1990): 451–82. Grubb, Farley, “The Long-­run Trend in the Value of European Immigrant Servants, 1654–1831: New Measurements and Interpretations,” Research in Economic History, 14 (1992), forthcoming. Jacoby, Daniel, “The Transformation of Industrial Apprenticeship in the United States,” Journal of Economic History, 51 (Dec. 1991): 887–910. Jensen, Joan M., Loosening the Bonds: Mid-­Atlantic Farm Women, 1750–1850 (London, 1986). Kaestle, Carl F., Pillars of the Republic: Common Schools and American Society, 1780–1860 (New York, 1983). Maynes, Mary J., Schooling in Western Europe (Albany, NY, 1985). Morris, Richard B., Government and Labor in Early America (Boston, 1981). Pennsylvania Gazette, 1729–1789 (Philadelphia, 1968). Perlmann, Joel and Dennis Shirley, “When Did New England Women Acquire Literacy?” William and Mary Quarterly, 48 (Jan. 1991): 50–67. “Record of Indentures of Individuals Bound Out As Apprentices, Servants, etc., and of Germans and Other Redemptioners in the Office of the Mayor of the City of Philadelphia, October 3, 1771, to October 5, 1773” (unpublished manuscript, City Archives of Philadelphia). Schofield, Roger S., “The Measurement of Literacy in Pre-­Industrial England,” in Jack Goody, ed., Literacy in Traditional Societies (Cambridge, MA, 1968): 311–25. Soltow, Lee and Edward Stevens, The Rise of Literacy and the Common School in the United States (Chicago, 1981). Steinfeld, Robert J., The Invention of Free Labor (Chapel Hill, NC, 1991).

Part II

German immigrant servitude in America, 1745–1835

9 The incidence of servitude in transatlantic migration, 1771–1804*

Introduction The New World held out the prospects of a better life to many Europeans. However, even by the eighteenth century the cost of transatlantic migration was still quite substantial. The price of ocean passage alone might easily have exceeded the value of half a year’s income for the typical British emigrant and the value of a year’s income for the typical German immigrant.1 It should not be surprising then to find that many prospective emigrants could not meet the cost of passage from their accumulated savings. Throughout the colonial period many of these emigrants purchased passage to America by voluntarily selling some of their future labor, this being the only other asset at their disposal. Some emigrants became indentured servants by signing fixed-­term future labor agreements before embarkation, exchanging this contract for transportation. Others borrowed the passage fare from their respective shippers pledging to sell themselves as servants in America, if necessary, to repay the loan. This process was known as redemptioner servitude. These institutions of immigrant servitude were the private market’s solution to financing the voyage of those who could not pay in advance. It gave the poorer part of the population the ability to finance moving from low-­productivity areas to high-­productivity areas, thereby improving their own welfare as well as the overall welfare of society. Without these labor markets it is unlikely that pre-­nineteenth-century America would have been populated as fast or grown as rapidly as it did. Little has been done quantitatively to determine the importance of immigrant servitude, i.e., to establish the incidence of servitude among immigrants, or to discover what influenced the likelihood that an immigrant would resort to servitude for financing transatlantic migration. Virtually nothing is known about the variance in the incidence of servitude over time, by age or family status, or across colonies or national groups. A couple of authors have presented broad estimates of the proportion of servants among the immigrants. These estimates are summarized in the often quoted assessment of Abbot E. Smith, “If we exclude the Puritan migration of the 1630’s, it is safe to say that not less than one-­half, nor more than two thirds, of all white immigrants to the colonies were indentured servants or redemptioners or convicts.”2 The aggregative generality and imprecision of this

160   German immigrant servitude in America, 1745–1835 measure comes from the estimating procedure which compares aggregate migration estimates to estimates of the number of servants recorded in colonial government records.3 Nevertheless, this estimate has formed the authoritative basis for subsequent studies of the flow of immigrant servants into Colonial America.4 In this study the percentage of servants within the immigrating population will be estimated by directly matching individual servant sales in the colonies to the immigrant vessels which carried them. Then for all ships carrying immigrants in the sample which recorded the total number of passengers, a servant percentage can be calculated. Estimating the incidence of servitude by using individual observations will allow for disaggregating the incidence of servitude by various immigrant characteristics. This was done for the port of Philadelphia for the period from 1771 to 1804. Philadelphia possessed a large and active market in immigrant servant contracts. By the late eighteenth century it was perhaps the single largest servant market in North America. Buyers from Pennsylvania, New Jersey, New York, Maryland, Virginia, the Carolinas, Georgia, Florida, Ohio, Kentucky, Massachusetts, and New Hampshire purchased ser­ vants at its auctions. The major supply of servants was provided by Irish and German immigrants.5 The Philadelphia market was also unique in that periodically the civil authorities recorded individual servant purchases. The surviving records of servant transactions, for the years 1771–3 and 1785–1804, were matched to immigrant passenger records from these years uncovering the incidence of servitude among immigrants arriving in Philadelphia in this period. The 1785–1804 evidence allows for determining the variance in the incidence of servitude by family status and the 1771–3 evidence allows for disaggregating the incidence of servitude by nationality. Additional information was taken from English emigration records for the years 1773–6 which covers all colonial destinations. This information allows for estimating the incidence of servitude by age and across colonial destinations.6 Analysis of this information may lead to a better understanding of the underlying causes of servitude and thereby a way of further depicting the social and economic differences across colonies and across the national groups which comprised the American colonists.

Family status, age, and life cycle wealth as determinants of the incidence of servitude Perhaps the most comprehensive and accurate records in the Philadelphia market are for German immigration between 1785 and 1804. The passenger lists from German immigrant vessels record the names of all aboard, predominantly by family groups. The record of servant transactions for these years was designed exclusively to record German immigrant servant sales. Therefore, matching the names between these two sources makes it possible to derive both the incidence of servitude and its family structure for this immigrating population (see Table 9.1). The overall incidence of servitude among the 7,837 German immigrants arriving in Philadelphia over this period of 19 years was 44.8 percent.7 This is 21 percentage points lower than previous aggregate estimates for this same population.8 The

The incidence of servitude, 1771–1804   161 Table 9.1 The distribution by family status of the incidence of servant contracting among German immigrants to Pennsylvania, 1785–1804 Total immigration Number

Percentage who became servants

(%)

Single adult males Single adult females Married persons Single male parents Single female parents Dependent children

3,074 605 1,854 50 69 2,185

(39.2) (7.7) (23.7) (0.6) (0.9) (27.9)

51.3* (44.9)a 58.5* (10.1) 35.2* (18.6) 30.0** (0.4) 50.7** (1.0) 40.1* (25.0)

Total

7,837

(100.0)

44.8 (100.0)

Sources: The information was derived by cross-matching the names of servants in “Book of Redemptioners, 1875–1804,” with the names of passengers listed on the ship manifests compiled in Strassburger (1934). Notes In many cases the family status was indicated in the passenger lists. Where the status was not indicated it was inferred from the positioning of the last names. Families were always listed together with the father named first then the mother and finally followed by any children. a This is the percentage of all German servants that comprised this status. * These means are all statistically different from one another above the 0.001 level, based on onetailed t tests. ** These means are statistically different from one another above the 0.025 level. In addition, the difference between single adult females and single female parents, between single adult males and single male parents, and between dependent children and single parents are statistically significant above the 0.1 level.

composition of these German immigrant servants was roughly split between singles and those migrating with other family members. Single, independent servants comprised 55 percent of the total with single males being the largest category at 45 percent. A quarter of the sales were composed of dependent children while parents and married persons made up the remaining 20 percent.9 There was considerable variance in the incidence of servitude depending on the immigrant’s family status. The relative probability of entering servitude among independent migrants should tend to reflect their relative wealth position, i.e., the relative ability to meet the cost of passage out of accumulated savings. These relative wealth differences may in part explain the difference in the incidence of servitude between single adult males and single adult females. Among single adult migrants, who were mostly in their late teens and early twenties, females may have had less opportunity to accumulate wealth. This may account for why single adult females had a 14 percent greater probability of resorting to servitude relative to their male counterparts. There exists an even larger difference in the incidence of servitude between single adults and married persons. The servant percentage was over 50 percent for single adults, whereas among married persons it was approximately one-­third less at around 35 percent. This  difference would appear to be a result of individuals being at different stages in their life cycle accumulation of wealth.10 Married persons were

162   German immigrant servitude in America, 1745–1835 generally older and therefore may have been relatively wealthier, having had more time to save. Thus married persons may have had a greater likelihood of being able to meet the cost of passage in advance out of accumulated savings. In order to investigate further the life cycle wealth effect on the incidence of servitude for this sample of German immigrants, the proportion of married immigrants entering servitude was disaggregated by the number of children in the family. The number of children should crudely approximate the age structure of household heads and hence their relative position in the life cycle accumulation of wealth. The incidence of servitude for household heads declines with family size, suggesting that the accumulation of wealth over a lifetime may have raised the probability of being able to pay cash in advance when securing ocean passage (see Table 9.2). Parents with few children, presumably the younger and so less wealthy, had a higher percentage entering servitude than those with many children. Among married couples the probability of entering servitude was just over 40 percent for those with less than two children, just over 30 percent for those with two to five children, and down to only 15 percent for those with over five children. The relative constancy of the servant percentage for married couples with between two and five children may be the result of their rising wealth being offset by increasing migration costs as the number of dependent children increased. Finally, the most dramatic decline in the incidence of servitude was for single female household heads. Those with one child experienced a 70 percent servitude rate whereas those with over three children experienced only a 10 percent rate. The sensitivity of the incidence of servitude to changing wealth over the life cycle is also suggested by evidence in English emigration records for the years 1773 to 1776. This evidence allows the incidence of servitude to be disaggregated directly by age, something that is not possible with the German evidence. Among these English emigrants, bound for all North American colonies, the proportion of adult males listed as servants peaks at age 21 and then declines steadily until age 25, by some 25 percent. The servant percentage stays rather constant from age 25 to age 40 and then radically declines to under 5 percent servant by age 50 (see Table 9.3).11 The shape of this servant percentage by age profile is very similar to the shape of the servant percentage by family size for married couples profile found in Table 9.2. The relationship between age and the likelihood of resorting to servitude suggests that the incidence of servitude among immigrants may be very sensitive to the determinants of individual wealth, in particular, the life cycle accumulation of wealth. These results also suggest that within a given emigrating population the division between servants and free emigrants may have been as much one of being at relatively different stages along a similar life cycle accumulation of wealth as being from different absolute income classes. The life cycle explanation may be relatively more important for German immigration since it was more family oriented with a wider age distribution than English immigration.12 But even though English immigration had less age variation, 40 percent were between ages 21 and 25, table 9.3 indicates that the incidence of servitude was sensitive to age even within this narrow age range.

The incidence of servitude, 1771–1804   163 Table 9.2 The incidence of servitude for adult parents by family size: German immigration to Pennsylvania, 1785–1804 Total immigration Number Married persons with: No children 1 child 2 children 3 children 4 children 5 children 6 children 7 + children

Percentage who became servants

(%)

510 360 326 218 186 132 62 60

(6.5) (4.6) (4.2) (2.8) (2.4) (1.7) (0.8) (0.8)

42.6 (6.2)a 40.0 (4.1) 31.9* (3.0) 31.2 (1.9) 30.1 (1.6) 34.9 (1.3) 14.3* (0.2) 16.7 (0.3)

1,854

(23.7)

35.2 (18.6)

Single females with: 1 child 2 children 3 children 4 + children

33 15 11 10

(0.4) (0.2) (0.1) (0.1)

69.7 (0.7) 46.7* (0.2) 36.4 (0.1) 10.0* (0.03)

Total

Total

69

(0.9)

50.7 (1.0)

Single males with: 1 child 2 children 3 children 4 + children

21 12 7 10

(0.3) (0.2) (0.1) (0.1)

38.1 (0.2) 33.3 (0.1) 14.3 (0.03) 20.0 (0.1)

Total

50

(0.6)

30.0 (0.4)

Sources. See the source note to Table 9.1. Notes a This is the percentage of all German servants in this category. * This indicates that the mean is statistically different from the preceding mean above the 0.1 level for one-tailed t tests. The sample size used to compute the t statistics was the number of couples as opposed to individuals in the case of married persons.

Determinants of servitude among dependent children The financial resources of parents seem to have had a major impact on the incidence of servitude among dependent children. For German immigrants arriving in Philadelphia between 1785 and 1804 the proportion of dependent children who entered servitude was 40 percent, over 7 percentage points higher than their parents but still well below the rate for single adults (Table 9.1). The probability of entering service among these children varied positively with family size and with whether their parents also entered service (Table 9.4). When parents did not have enough resources to pay for their own passage fare  the likelihood that they would pay the passage fares of their children might  be expected to be significantly reduced. This is consistent with the

164   German immigrant servitude in America, 1745–1835 Table 9.3 The incidence of servitude by age for adult male English emigrants, December 1773 to April 1776 Age

Total observations

15–20 21 22 23 24 25 26–30 31–35 36–40 41–45 46–50 51 + Total

947 555 430 266 240 266 812 371 307 124 70 25 4,440

Percentage listed as servants 77.2 91.1* 83.7* 80.8*** 79.2 69.2** 69.6 68.2 66.1 43.5* 25.7** 4.0*** 74.4

Source: Fothergill (1908–11). Notes This sample is for all destinations in British Colonial America. * This indicates that this mean is statistically different from the preceding mean above the 0.001 level for a one-tailed t test. ** This indicates that this mean is statistically different from the preceding mean above the 0.1 level. *** This indicates that this mean is statistically different from the preceding mean above the 0.15 level.

evidence, which indicates that if the parents went into service their children had a 60 percent probability of also becoming servants, about twice the chance than if their parents remained free. Similarly, as family size increased, total family debt for passage also increased, thus putting pressure on the family’s limited resources. Children from large families therefore might be expected to have a higher probability of entering servitude. This also is consistent with the evidence in Table 9.4 which indicates that an only child had a 24 percent chance of entering service whereas possessing three or more siblings nearly doubled this chance. Children from large families where the parents also entered service, therefore, had the greatest likelihood of becoming servants, from 60 to 85 percent. German immigrant parents have been much maligned for supposedly selling their children in order to pay their own passage charges. The historical literature tends to follow the assessments of Gottlieb Mittelberger who described the servitude of German immigrant children in 1756 as follows: Many parents in order to pay their fares in this way and get off the ship must barter and sell their children as if they were cattle. People who arrive without funds to pay their way and who have children under the age of five, cannot settle their debts by selling them. They must

The incidence of servitude, 1771–1804   165 Table 9.4 The incidence of servitude among German immigrant children to Pennsylvania, 1785–1804 Number of siblings

From all households 0 1 2 3 4 5 6+ Total

Total immigration of dependent children

Percentage of dependent children entering servitude

Number

(%)

234 380 381 416 340 216 218

(3.9) (4.9) (4.9) (5.3) (4.3) (2.8) (2.8)

23.9 (1.6)a 32.9* (3.6) 38.6* (4.2) 45.4* (5.4) 45.6 (4.4) 45.4 (2.8) 49.1 (3.1)

(27.9)

40.1** (25.0)

2,185

From households where the parents also became servants 0 103 (1.3) 1 126 (1.6) 2 116 (1.5) 3 120 (1.5) 4 120 (1.5) 5 18 (0.2) 6+ 36 (0.5)

41.8 (1.2) 55.6*† (2.0) 62.9 (2.1) 62.5 (2.1) 70.0† (2.4) 83.3 (0.4) 58.3* (0.6)

Total

59.6** (10.9)

639

(8.2)

From households with single parents who also became servants 0 31 (0.4) 61.3 (0.5) 1 22 (0.3) 86.4* (0.5) 2 15 (0.2) 73.3 (0.3) 3+ 13 (0.2) 69.2 (0.3) Total

81

(1.0)

71.7** (1.7)

Source: See the source note to Table 9.1. Note Zero siblings implies the observation of an only child migrating in the company of his parents. One sibling counts all children from families with two children, and so on. a This is the percentage of all servants in this category. * This indicates that this mean is statistically different from the preceding mean above the 0.05 level, based on a one-tailed t test. ** These means are statistically different from one another above the 0.05 level. † These means are statistically different from one another above the 0.05 level.

give away these children for nothing to be brought up by strangers; and in return these children must stay in service until they are twenty-­one years old. Children between five and ten who owe half-­fare, that is, thirty florins, must also go into service in return until they are twenty-­one years old, and can neither set free their parents nor take their debts upon themselves. On the other hand, the sale of children older than ten can help to settle a part of their parents’ passage charges.13

166   German immigrant servitude in America, 1745–1835 This same impression might also be gathered by comparing the evidence between Tables 9.2 and 9.4. There is a marked divergence in the incidence of servitude between parents and children as family size increases. For example, only one-­sixth of parents with seven or more children entered servitude whereas almost 50 percent of their children entered service. However, in order to assess truly whether children were sold so that their parents could go free, the sale prices of their children must also be considered. The evidence from the 1785–1804 record of German servant sales with regards to contract prices and other contract parameters tends not to support the above description. Although some children entered service while their parents remained free, virtually none were sold for prices above their own passage fare. Therefore, children were almost never used to pay for their parents’ passage obligations. In fact, most of the transfer of passage debts within the family seems to be from child to parent, with parents serving in the child’s place or paying for the child’s passage outright. Among all free parents some 70 percent of their children also remained free, suggesting that free parents were paying for over two-­thirds of the migration expenses of their children. Among married servants, roughly half had explicit contracts exempting some or all of their children from service. These children were allowed to stay with their parents while still being provisioned by their parents’ master. Servitude among dependent children was disproportionately concentrated in the older ages, ages 7 to 15, compared to the age structure of all immigrant children.14 Children under age 5 were seldom sold or given away. In general, it was the oldest child who was pressed into service first. Many of these children were at the age where entering into apprenticeship contracts might have been expected of them anyway. This may help explain the divergence in the probability of entering ser­ vitude among children compared to their parents as family size increased. As parents grew older they would tend to accumulate both more children and more wealth. But this life cycle accumulation of wealth apparently was not enough to fully cover the added migration cost of extra children. Thus although parents of large families had more wealth and so had a higher probability of at least being able to pay their own freight expense, they also had a larger total family passage expense and so tended disproportionately to press their older, apprenticeship age children into paying their individual share of the family debt by entering servitude.

The incidence of servitude across national groups Little is known about how the proportion of servants to free immigrants varied across national groups. Differences in the incidence of servitude across European immigrant populations may provide valuable information on how the nature of these populations differed. Unfortunately, only one year of evidence is complete enough to allow for comprehensive and detailed estimates of the incidence of servitude across national groups. This evidence, however, appears to be relatively representative of the period under study. For the year 1773 the percentage of servants among Irish, English, and Scots immigrants was estimated for the port of Philadelphia (Table 9.5). This was the only year for which a list of

The incidence of servitude, 1771–1804   167 Table 9.5 The incidence of servitude among immigrants to Pennsylvania, 1771–3

Irish, 1773 Northern Irish Londonderry (6 ships) Belfast (3 ships) Newry (5 ships) Subtotal Southern Irish Dublin (3 ships) Waterford (1 ship) Cork (1 ship) Subtotal Irish subtotal

Total passengers

Percentage who became servants

765 154 511 1,430

12.94 26.62* 23.29 18.11**†

235 81 81 387

69.94 58.02* 41.98* 61.27**†

1,827

27.48

Scots, 1773 Glasgow (2 ships) Isle of Lewis (1 ship) Scots subtotal

82 295 382

8.54 30.54* 25.39**

English, 1773 Liverpool (1 ship) Bristol (2 ships) London (5 ships) English subtotal

8 36 111 174

0.00 83.33 54.95* 52.30**

2,378

29.02**†

British total

German, 1771–3 (adult men only) Rotterdam (10 ships) 594 London (7 ships) 153

60.94 48.37*

German total

58.37**

747

Sources: The servant contracts were taken from “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia, October 3, 1771, to October 5, 1773.” For the British, these contracts were matched to ship arrival information in Tepper (1978: 180–1). For the Germans, these contracts were matched to the loyalty oath lists preserved in Strassburger (1934). For the Germans the common evidence runs from October 31, 1771 to August 23, 1773. For the British the common evidence runs from May 13, 1773 to September 2, 1773. Notes * This indicates that the mean is statistically different from the preceding mean above the 0.05 level, based on one-tailed t tests. ** These means are all statistically different from one another above the 0.05 level. † These means are all statistically different from one another above the 0.05 level.

British ship arrivals could be found which coincided with the port’s records on bound servant sales and also had enough information to allow both the matching of servants to ships and the determination of the servant percentage among the ships’ passengers. The British passenger records did not record the names of those aboard so the matching of servant contracts to immigrant vessels was

168   German immigrant servitude in America, 1745–1835 p­ erformed by cross-­referencing the following information: the date of arrival with the contract’s commencement date, the contract’s sales agent with the ship’s captain, the ship’s name, and the ship’s port of departure.15 The incidence of servitude among German colonists arriving in Philadelphia between 1771 and 1773 was also estimated. German servants could only be matched to their respective vessels by comparing the names in the contract records to the names recorded in the German passenger lists. Unlike the British evidence, there were no other clues for matching German servants to their respective vessels. In addition, the incidence of servitude among these Germans can only be determined for males above age 15 since only the names of this group were recorded in the surviving German passenger records.16 Therefore, the German and British samples in Table 9.5 are not strictly comparable both because the historical samples were gathered differently and because the sex, age, and family status between the two immigrant populations were substantially different. The British sample has only five married couples and no children among the servants. Among English immigrants in this period, there were over five times fewer dependents than among German immigrants. Almost all British servants were young unattached adults and 26 percent of the sample were females. The German sample in Table 9.5 consists of only males above age 15 with about 9 percent being husbands jointly contracting with their wives. The estimated results uncover several interesting features about the incidence of servitude in the British immigrating populations. Only 29 percent of the 2,378 British subjects arriving in Philadelphia in 1773 were sold as servants. This suggests that either prior estimates may have been inflated or the proportion of ser­vants among British immigrants may have fallen precipitously by the 1770s.17 The regional difference in the servant percentage are equally as striking. English immigrants led the way with a little over half becoming servants.18 However, the English were a relatively small subset of the British immigrant stream, only 7.3 percent of those arriving in Philadelphia in 1773. British immigration, at least to Pennsylvania, was dominated by Irish colonists. The proportion of servants among Irish and Scots immigrants was roughly half the English proportion, i.e., 27 and 25 percent, respectively. This difference is even more surprising considering that the percentage of female servants among the Irish and Scots was twice that of the English, single adult females being more likely to enter servitude. The Irish migration presents a special case in that there was a significant difference between the northern and the southern Irish. Ulster immigrants possessed the lowest regional incidence of immigrant servitude at 18 percent. On the other hand, the southern Irish experienced the highest rate. At 61 percent, it was over three times the Ulster rate. Northern Ireland may have supplied three and a half times as many immigrants in 1773 as did southern Ireland but among servants the two contributed equally.19 Finally, in the year 1735 the American Weekly Mercury reported the arrival of 177 servants out of 372 Irish immigrants arriving in Philadelphia for a servant percentage of 47.6.20 If the southern to northern Irish migration was relatively constant over time, the decline in the incidence of

The incidence of servitude, 1771–1804   169 servitude among the Irish from 1735 to 1773 would be over 40 percent. This suggests that there may have been a gradual decline in servitude among Irish immigrants over the century. The proportion of servants among the Germans from Rotterdam in the early 1770s was roughly equal to that experienced by the southern Irish, 61 percent. However, Germans sailing from London had a 20 percent lower incidence of servitude. This was also 12 percent lower than that experienced by Englishmen sailing from the same port. Comparing overall British and German immigration to Philadelphia in this period suggests that the British had a 50 percent lower incidence of servitude. Finally, comparing the servant percentage among adult male Germans before and after the Revolution shows a decline in the incidence of servitude of about 10 percentage points over this 20-year period. This suggests that immigrant servitude may have been gradually disappearing among the Germans as well. The relative incidence of servitude among national groups or across regions of emigration should tend to reflect the relative ability to secure passage to America through paying cash in advance. For the British, the primary cost of migration would be the ship’s passage fare since most were within a short distance of a port. If the cost of passage to Philadelphia did not vary significantly between British ports, say due to the competitive nature of the shipping market, then the regional differences in the incidence of servitude among the British should tend to reflect, for the most part, the average relative wealth positions of the respective emigrants. This suggests that immigrants from Scotland and Ulster, particularly the large migration from Londonderry, must have been relatively well off compared to immigrants from southern Ireland and England. This interpretation also suggests that the extreme southern Irish were better off than Irish from Dublin, mainland Scots were better off than Scots from the Isle of Lewis, and English from London were better off than English from Bristol. One possible explanation of these inferred wealth differences among British immigrants is that they may reflect average age differences among the immigrants across nationalities and ports. The group with a lower incidence of servitude may have been older on average and hence farther along on the life cycle accumulation of wealth and so better able to meet the cost of passage out of accumulated savings. Although the evidence is sparse, there appears to be no substantial age differences among immigrants from the various regions of Britain in the evidence in Table 9.5.21 This suggests that the regional differences in the proportion of servants may reflect different levels of aggregate wealth, at all ages, across these ethnic groups, at least among the emigrants. This interpretation suggests that the incidence of servitude across ethnic groups is an index of their relative poverty. For example, emigration from Ulster was on the rise in the early 1770s and has been attributed to poverty occasioned by a sharp rise in land rents and a depression in the linen trade.22 These events may have been the impetus to the surge in emigration but they may not have rendered these Ulster migrants destitute since few resorted to servitude, relative to other British immigrants, in order to finance their journey to America.

170   German immigrant servitude in America, 1745–1835 Interpreting the relative wealth of British versus German immigrants based on the relative incidence of servitude is not straightforward. First of all, passage fares for Germans, who sailed primarily from Rotterdam may have been as much as 10 pounds sterling higher than for emigrants sailing from British ports.23 Second, Germans faced a long and costly overland journey down the Rhine river before reaching the ports of embarkation.24 This expense would deplete the resources of German emigrants before the passage fare was to be paid. These factors raised the total cost of transatlantic migration for Germans far above what the British experienced. Given these higher migration costs, German immigrants seem to have fared relatively well as measured by the proportion who became servants. They must have been wealthier on average than immigrants from southern Ireland and Bristol and also perhaps wealthier than English immigrants from London, since Germans who sailed from London had a lower servant percentage. This relative ranking of wealth may explain contemporary Pennsylvanian accounts which typically rate Scotch and German servants as being of a better class, generally, than Irish or English servants. The estimated difference in the incidence of servitude must be due in part to German adult migrants being on average older than British adult migrants. There was a much higher share of married migrants among the Germans.25 Thus part of the inferred wealth difference would be due to relative life cycle differences in wealth rather than absolute wealth differences across all classes.

The incidence of servitude across colonies Emigrants, whether free or servant, had many colonial destinations from which to choose voluntarily. In transplanting themselves to the New World, colonists were in general making permanent locational choices. It was not temporary work but the expected future net benefits, realized over their remaining life time, which attracted them to their chosen colonies. In this manner, the distribution of colonial destinations can be treated as the revealed preference of immigrants’ desired permanent residences. Therefore, how emigrants distributed themselves across colonies may reveal the ranking of their net expected future opportunities, the sum of both pecuniary and nonpecuniary net benefits, across colonies. By the middle of the eighteenth century the Chesapeake and Delaware regions dominated the transatlantic destinations of European emigrants. These regions, therefore, clearly represented the top colonies of opportunity in the minds of most European migrants.26 Comparing the distribution of servant versus free immigrants across colonies should reflect the difference in the net future opportunities across colonies expected by servants compared to that expected by free emigrants. Colonists who came as servants had insufficient savings even to secure passage to America. They would begin their postservant careers in the New World with little or no capital.27 Therefore, colonies with a relatively high proportion of servants among their immigrants must have been viewed as the colonies of relative opportunity for the poor compared to those who arrived with some wealth.

The incidence of servitude, 1771–1804   171 Colonies which received mostly free immigrants may have presented opportunities to succeed for those with some capital but held relatively little advantage for those arriving with no capital. Little has been done quantitatively to determine the incidence of servitude across colonies. One body of evidence, however, can be used to discover the difference in the proportion of servants among English emigrants to British America. Between 1773 and 1776 English authorities kept emigration records. Among the information recorded was the colony of destination and whether the emigrant was embarking as a servant (Table 9.6). Since these are emigration records the estimated incidence of servitude may overstate the actual absolute incidence realized after landing since some servants were able to find financing after arrival and so bought their way out of servitude. However, a reasonable idea of the variance in the incidence of servitude across colonies may be revealed by this evidence. By the American Revolution the colonies of Pennsylvania, Maryland, and Virginia clearly stood out not only for attracting the majority of European immigrants but also for having a significantly higher proportion of servants among their immigrants relative to other colonies. These regions must have been viewed as not only the general land of opportunity, but more particularly, the land of opportunity for poorer colonists in the minds of prospective English emigrants. A very distant second were the Carolinas and Georgia with under a quarter of their immigrants being servants. The colonies north of Pennsylvania, and the Caribbean colonies, had a very low incidence of servitude among their immigrants, under 9 percent on average. These latter two regions, therefore, may have been thought of as places of opportunity only for those with sufficient capital initially to set themselves up in business but of little advantage to those migrating with no capital. The prominence of the middle colonies as “The Best Poor Man’s Country” is reinforced by the fact that most German and Irish immigrant servants also chose this destination.28 The preferences revealed by servants versus free emigrants in choosing their respective destinations may be used to infer the differences in the distribution of expected opportunities for poor relative to wealthy colonists across these destinations. However, this evidence does not indicate what these differing opportunities were. There were many differences across colonies relating to nonpecuniary benefits. For example, the Caribbean colonies were relatively hotter with a more virulent disease environment. This kind of difference by itself would appear to be an unlikely explanation of the variance in the incidence of servitude across colonies since it would seem unusual to claim that wealthy immigrants were disproportionately attracted to hot, disease environments compared to poorer immigrants. A type of nonpecuniary difference which might affect free and servant immigrants differently was the working conditions faced by servants. Field labor on the sugar plantations of the Caribbean, and possibly in the rice fields of South Carolina, was viewed as strenuous relative to field labor on the grain and tobacco farms of the more northern colonies. The prospect of being engaged in harsh gang style labor possibly in conjunction with black slaves may have prompted

172   German immigrant servitude in America, 1745–1835 Table 9.6 The incidence of immigrant servitude across colonies, English emigration, 1773–6 Total emigration

Percentage listed as servants

Northern colonies Canada Nova Scotia New England New York

31 425 54 303

9.68 7.76* 1.85 11.55*

Subtotal

813

8.86†

Middle colonies Pennsylvania Maryland Virginia

859 2,217 767

78.81§ 98.33§ 90.35§

Subtotal

3,843

92.38†

Southern colonies Carolinas Georgia Florida

106 196 5

23.58** 17.86** 0.00

Subtotal

307

19.54†

Caribbean colonies Jamaica Grenada Barbados Dominica St. Kitts Antigua Tobago St. John’s Nevis St. Christopher St. Vincent Mosquito Coast Tortola Montserrat

234 71 41 38 32 21 17 13 11 10 10 7 4 1

7.69 9.86 2.44 7.89 3.13 4.76 0.00 46.15 0.00 0.00 0.00 0.00 0.00 0.00

Subtotal

510

7.25†

Source. See the source note to Table 9.3. Notes Convicts were excluded from the count. † These means are statistically different from one another above the 0.001 level, except for the Caribbean versus Northern Colony comparison which is only significant at the 0.2 level. These tests are based on one-tailed t tests. § These means are statistically different from one another above the 0.001 level. * These means are statistically different from one another above the 0.05 level. ** These means are statistically different from one another above the 0.2 level.

The incidence of servitude, 1771–1804   173 immigrant servants to disproportionately avoid colonies in the Caribbean and Carolinas relative to free immigrants. The few servants who did go to these regions were mostly skilled workers and so were not engaged in field labor. Free immigrants who went to the Caribbean would not be doing field labor and so this would not affect their choice of colonial destinations.29 This type of explanation, however, does not account for why servants also disproportionately avoided colonies north of Pennsylvania compared to free immigrants. In addition, a reluctance to work around slaves would also seem to be a weak explanation since the Chesapeake region relied far more on slave labor than on servant labor. And yet this destination was disproportionately popular among servant immigrants relative to free immigrants.30 A more comprehensive explanation of the variance in the incidence of servitude across colonies may be due to more direct differences in expected pecuniary benefits arising from differences in the crops grown and in the scale of farming across colonies. Sugar producing in the Caribbean was a very lucrative business. Caribbean sugar planters may have been the wealthiest men in English America.31 The technology of sugar production, with its relatively large fixed capital requirements for milling equipment and slaves, gave rise to large economies of scale, i.e., a declining average cost of production over a large range of output. It may have taken several thousand pounds to start a sugar plantation of even modest size.32 In addition, most of the prime land was already taken on the relatively small Caribbean islands by the mid-­eighteenth century and so good land at cheap prices was hard to find.33 Therefore, the prospects of successfully entering sugar farming for poor immigrants would appear low since the large capital requirements would be an effective barrier to entry for them. This may explain why servants disproportionately avoided the Caribbean colonies relative to free immigrants. By contrast, tobacco and grain farming in the Chesapeake and Delaware regions were both lucrative and could be practiced without significant scale economies. Because land was still relatively abundant, at least on the frontier, small farmers with little capital could still successfully begin and compete in this business. On the other hand, colonies north of Pennsylvania lacked the lucrative staple export crop that the middle and southern colonies possessed. This meant both fewer opportunities to accumulate wealth in farming and less opportunity to accumulate enough working as a hired laborer to stake oneself to a farming venture in these northern colonies relative to those farther south. The relatively low capital requirement and potential success of entering into farming in the Chesapeake and Pennsylvania regions might partly explain the high incidence of servitude among immigrants to these colonies in the second half of the eighteenth century relative to other colonies. The explanation of the variance in the incidence of servitude across colonies based on differences in the profitability of crops and farming scale economies is reinforced by evidence on the relative distribution of wealth across colonies as well as differences in the derived demand for labor across colonies. On the eve of the American Revolution the middle colonies exceeded the other mainland

174   German immigrant servitude in America, 1745–1835 colonies in terms of per capita nonhuman wealth when servants and slaves were counted in the population. In addition, per capita wealth was more evenly distributed in the middle colonies than in other mainland colonies and substantially more evenly distributed than in the Caribbean.34 These features might be attractive to those immigrants who arrived with no capital, e.g., servants, since it signaled the relative ability of poor immigrants to succeed financially in the middle colonies compared to other destinations. This might have then led to the relatively high proportion of servant versus free immigrants choosing to migrate to the middle colonies. Differences in the demand for immigrant labor across colonies is derived from difference in the gains from employing such labor, thus the relative demand for labor would signal the relative profitability of farming across colonies. Although most colonies engaged in some recruiting of European servants at various times, the Caribbean and middle colonies were by far the most active and persistant recruiters over the colonial era. The inability of northern colonies, in particular, to successfully recruit servants indicates that they could not offer servant contracts attractive enough, e.g., short enough, to compete with contracts offered by colonies farther south while still covering the cost of transporting the servant to America.35 This was due, in part, to the relatively higher productivity of labor in colonies south of New England. Thus colonies with high labor demand would tend to signal relatively high future opportunities, other things equal, and so tend to attract potential emigrant servants relative to other colonies.

Conclusions By the late eighteenth century one of the major American regions receiving European immigrants was Pennsylvania with Philadelphia possessing perhaps the single largest immigrant servant market in North America. Evidence from this market for the last quarter of the eighteenth century suggests several interesting features about the institution of immigrant servitude. The overall incidence of servitude among British subjects arriving in Philadelphia in 1773 was 29 percent and among Germans arriving between 1785 and 1804 was 45 percent. These direct estimates are lower than those of previous studies which were based on indirect estimates or emigration records. The incidence of servitude among adults shows a considerable variance which appears to be related to the relative wealth positions of individuals, in particular, to different stages of life cycle wealth accumulation. Single adult females had the highest likelihood of entering service followed by single adult males and then married persons. The older the adult, the lower was the probability of resorting to servitude. Within an immigrant group, free and servant colonists may have been as much from different stages on a common life cycle wealth path as from different social classes. For children, the probability of entering servitude was positively related to their parents’ resources and negatively related to the child’s age. The changing ability to pay cash in advance to secure ocean passage was an important determinant of the proportional magnitude of servant contracting.

The incidence of servitude, 1771–1804   175 There was considerable variance in the incidence of servitude across regions of emigration and across national groups. The negative relationship between wealth and the incidence of servitude was used to rank the relative wealth position across these various migrant groups. This inferred relative wealth ranking, from high to low, tentatively runs as follows: immigrants from Glasgow, Londonderry, the rest of Ulster, the Isle of Lewis, Cork, Germany, English from London, Irish from Waterford, Dublin, and finally English from Bristol. For British immigrants this appears to reflect absolute rather than life cycle wealth differences across groups since the age structure does not appear to vary much. Finally, for all groups there appears to be some gradual decline in the incidence of servitude over time. Finally, the incidence of servitude varied considerably across colonies with Pennsylvania and the Chesapeake having the highest incidence of servitude among their immigrants by a wide margin. This suggests that by the mid-­ eighteenth century these regions were viewed as presenting the best opportunities for poorer colonists to succeed relative to those who emigrated with some capital. This may have been due to the lucrative nature of tobacco and grain farming coupled with the lack of scale economies which kept capital requirements low in the middle colonies relative to other colonies. The author is Assistant Professor in the Department of Economics at the University of Delaware, Newark, Del. 19716. I thank Trevor Dick, David Galenson, Henry Germery, David Gerber, and Marvin McInnis for helpful comments on earlier drafts. A version of this paper was presented at the 1984 Social Science History Association meeting in Toronto and in the economic seminar at the University of Delaware. I thank the participants of these seminars for their comments.

Notes   * Reprinted from Farley Grubb, “The incidence of servitude in trans-­Atlantic migration, 1771–1804,” Explorations in Economic History, vol. 22, no. 3 (July 1985): 316–39, Copyright 1985, © 1985, with permission from Elsevier.   1 In the last quarter of the eighteenth century British passage fares to North America for adults ranged between 3.5 and 6 pounds sterling, e.g., see Campbell (1955: 17), Dickson (1966: 86–7), and Galenson (1981: 251–2). Annual per capita English income averaged between 12 and 22 pounds sterling in this period, e.g., see Deane (1965: 6). German passage fares from Holland to Philadelphia in this period ranged between 13 and 17 pounds sterling for adults, e.g., see Strassburger (1934, Vol. I: xxxvii, and Vol. III: 112–14, 131–4, and 137–8). The average annual earnings of German agricultural workers were around 11 to 12 pounds sterling and for teachers were around 16 pounds sterling in the Rhineland at the beginning of the nineteenth century, see Maynes (1977: 119–20). See also the discussions of passage fares to income ratios for various European immigrant streams in Galenson (1984: 16–22). Where values were not expressed in pounds sterling they were converted to pounds sterling following the conversions provided in McCusker (1978).   2 Smith (1947: 336). See also the estimates in Geiser (1901: 40–1) and Campbell (1955: 4–5). A lower estimate is given by Lebergott (1984: 190). He finds that indentured servants comprised about 38 percent of those immigrating to the mainland colonies in the early 1770s.

176   German immigrant servitude in America, 1745–1835   3 Since immigration records are seldom complete and colonial servant records contain many transactions that are not of new immigrant servants, this procedure used by Smith (1947) and Geiser (1901) may seriously overstate the proportion of servants among new immigrants. The estimates of Campbell (1955) are based on emigration records. Since many servants were able to secure financial help after arrival in the colonies, thus buying their way out of servitude, emigration records may also seriously overstate the proportion of immigrants who entered servitude. For example, one-­third of the Germans arriving on the Belvidere and Pennsylvania in 1803 who were in debt for their passage fares nevertheless found alternative financing to servitude, Strassburger (1934, Vol. III: 113–14, 137–8) provided the passenger lists with passage debts individually enumerated. The actual incidence of servitude was derived by cross-­matching these lists with the “Book of Redemptioners, 1785–1804.” See additional discussions and examples of paying off passage debts without entering servitude in Smith (1947: 20–2), Trautmann (1981: 40–1), Durnbaugh (1958: 231–2), Mittelberger (1960: 16), and Strassburger (1934, Vol. I: xxxvii). The estimates of Lebergott (1984) are based on the comparison of the number of servants in a sample of probate inventories with estimates of total immigration. The biases inherent in probate records make estimating the total flow of servants highly speculative, and since servants were not exclusively immigrants, Lebergott’s estimate may also overstate the incidence of servitude.   4 The estimate by Geiser (1901) forms the basis for the estimates in Herrick (1926: 177) and Smith (1947: 323). These sources in turn are those used by subsequent studies to assert the prevalence of immigrant servitude, e.g., see Galenson (1981: 3–4; 1984: 18), Heavner (1978: 702), Heilbroner and Singer (1977: 40), Lemon (1972: 10), and Salinger (1980).   5 The record of servant sales for the port of Philadelphia from 1771 to 1773 records 2835 immigrant servant sales. Germans comprised 42.3 percent, Irish 41.7 percent, Scots 4.9 percent, and English 8.3 percent of this total, see the source note to Table 9.5 for the source of this evidence. Among the Irish servants, roughly half were from Ulster and half were from southern Ireland. By the eighteenth century the colonies south of Pennsylvania had largely switched to slave labor and New England never attracted much servant labor. A significant portion of English servants still went to the Chesapeake in the eighteenth century; however, the number of German and Irish ser­ vants arriving in Pennsylvania appears to have been larger. See discussions of the transition from servants to slaves in Menard (1977), Smith (1947: 26–35), Galenson (1981: 117–68, 219–27), Dunn (1972: 47–83), Morgan (1975: 295–315), and Wood (1974: 131–66).   6 Ideally, a logit model should be performed relating whether or not an immigrant entered servitude against a host of the immigrant’s personal characteristics and other colony characteristics. However, the three bodies of evidence in this study are not compatible for merging by virtue of their construction, and since each sample only captures one different immigrant characteristic, nothing statistically can be gained beyond looking at the simple means of the distributions.   7 There were 433 contracts that appeared similar to new immigrant contracts but which could not be matched to the surviving immigrant ship lists of passengers. Some 200 of these were relatively clustered together at various points suggesting the possible absence of records on actual ship arrivals. Two ships were also included by Strassburger (1934) in the year 1800 but their arrival was undated and they also happened not to match any recorded servant contracts. These ships, numbers 439 and 440, were therefore assumed to have arrived after 1804 and so were excluded from the passenger count for the sample. If all the nonmatches were by chance actually missed matches from the surviving passenger lists in Strassburger (1934) then the servant percentage would rise to just under 50 percent. This might then be considered the upper bound to the incidence of servitude for this sample.

The incidence of servitude, 1771–1804   177   8 Geiser (1901: 41) lists 3,622 redemptioners registered between August, 1786 and December, 1804 and 5,509 foreign immigrants landing in the same period. Writing in 1901 Geiser did not have the benefits of the Strassburger (1934) compilation of ship passenger lists. The Strassburger (1934) lists uncovered many more ship arrivals than had prior investigations. Geiser therefore was using a relatively incomplete set of immigration records, which led him to understate German immigration by some 30 percent. Since the servant records were complete this led Geiser to seriously inflate the servant percentage among German immigrants.   9 In this sample almost all of the married servants had joint contracts. Only one couple formed separate servant indentures while 13 husbands had contracts which exempted their wives from service and only six wives went into service while their husbands remained free. As far as could be ascertained none of these servants were indented or brought across by fellow passengers. All the servants were sold to American residents. 10 For an empirical analysis of the life cycle wealth effect see Atack and Bateman (1981). Their regression analysis finds that age has the largest impact on explaining the variance in family wealth out of the variables of race, sex, literacy, occupation, nativity, and age. Family wealth quadrupled between age 16 and age 45. This relationship should be similar for the eighteenth century as well. See also Kearl and Pope (1983) and Main (1983). 11 This evidence is for all colonial destinations, for males above age 15 only, and taken from emigration records. Therefore, it is not strictly comparable in terms of the absolute measure of the incidence of servitude to the English reported in Table 9.5. On the use of emigration as opposed to immigration records to estimate the incidence of servitude, see footnote 3. 12 The ratio of women and children to adult men was over five times higher among German immigrants relative to English immigrants, compare Wokeck (1981: 260–1) to Campbell (1955: 4–5). 13 Mittelberger (1960: 18). 14 A sample of 528 dependent children, both servant and free, had ages recorded in the passenger lists (Strassburger 1934, Vol. III). This yielded a frequency distribution for all children of 39, 44, and 17 percent for the age categories of 0 to 6, 7 to 15, and over 15, respectively. The frequency percentage for children who entered servitude can be inferred from their contract lengths, i.e., male children were found consistently to serve until they were 19. A sample of 667 servant children from the 1785–1804 evidence reported in Table 9.1 yields a frequency distribution of ages based on contract lengths of 19, 59, and 22 percent for the age categories of 0 to 6, 7 to 15, and over 15, respectively. 15 Most British servants were indentured servants and so their contract commenced on the ship’s arrival date. This fact was crucial for matching these servants to their ships. Not all the information was available for each contract, but enough was available so that all the contracts could be matched to their respective ships. 16 Germans were exclusively redemptioners and so their contract’s labor time commenced upon sale rather than upon the ship’s arrival. Therefore, there was nothing in the Germans’ contracts that suggested an arrival date. Second, redemptioners formed their contracts on the spot market and so were their own sales agents whereas British indentured servants, having sold their contracts to the ship’s captain before departure, were actually being resold so they had a sales agent, usually the captain, recorded in the transaction in the colonies. Finally, the ship’s name was never listed for Germans in the transaction records of servant sales. On the nature of the loyalty oaths which made up the German passenger records see the discussion in Strassburger (1934, Vol. I: i–3) and Wokeck (1981). 17 This lower incidence of servitude may also be just an artifact of some indentures not  being recorded in Philadelphia. This seems unlikely since recording the contract was a safeguard both for the master and for the servant’s interests in realizing

178   German immigrant servitude in America, 1745–1835 compliance to all the contract’s terms, legally. Also recording the contract was relatively inexpensive. The low incidence of servitude found for the British may also be particular to the Pennsylvania market since the estimate is dominated by Ulster emigration. Englishmen, with a much higher incidence of servitude (50 percent), may have been relatively more important in British immigration to other colonies, i.e., see Table 9.6 and Galenson (1981: 83). Finally, it is possible that the year 1773 may have been an anomaly in the evidence. 18 This is only slightly below the estimate in Campbell (1955: 4–5) based on 1773–6 English emigration records. 19 Ulster emigration reached its eighteenth-­century peak in the 1770s (see Dickson 1966: 60–82). Therefore, the evidence here may overstate the proportional contribution by Ulster immigrants and servants in earlier periods. The difference in the incidence of servitude between northern and southern Ireland are consistent with the observations made by Dickson (1966: 89). Many studies have suggested that Irish immigrants to the middle colonies were all Scots-­Irish from Ulster and usually cite Dickson (1966) as their authority. However, Dickson (1966) states that he is only studying northern Irish emigration and has little information about the south. Dickson (1966: 67) does suggest that ships which reported southern ports as their point of departure may have really been from Ulster and only listed the southern ports as their clearance stop. The systematic difference in the incidence of servitude, as well as other contract parameter differences, strongly suggests that the two groups were fundamentally different and so probably from different regions of Ireland. 20 Cited in Salinger (1980: 110). Smith (1947: 313) suggests that Irish immigration in the 1720s was about 70 percent from Ulster. 21 The Philadelphia record of bound contracts for 1771–3 lists 1,557 British immigrant servants (see footnote 5). Among these British servants there were only 19 contracts involving married couples and only 58 contracts stipulating parental consent or education thus implying child status. Most of these families arrived as one group from the Isle of Lewis in Scotland. The general absence of family relations suggests the possible absence of both older and younger migrants from among the various British immigrant groups. See also the English servant age distribution in Table 9.3 and in Galenson (1981: 26–33). 22 See Dickson (1966: 60–82). 23 See footnote 1. This difference reflected the cost incurred by a longer voyage, extra customs stops, and other civil regulations that were imposed on German immigration relative to British immigration. During the colonial era the English navigation acts forced German immigrant vessels to pass through English customs at an English clearance port on their way from Holland to Pennsylvania. The sailing time down the English Channel was also abnormally long relative to the entire voyage, perhaps as much as 25 percent of the entire time at sea. Last, Pennsylvania levied import duties on German immigrants and required them to sign loyalty oaths. 24 See the description in Mittelberger (1960: 7–11). He states that it took him seven weeks to travel the usual immigrant route down the Rhine and that he had to go through 36 different customs stops between Heilbronn and Holland alone. He viewed this stage of the journay as both relatively time-­consuming and expensive. 25 See footnote 18. Although there were more older married men in this German sample there were also no adult females sampled. Since both samples exclude children, and since the biases in the incidence of servitude for adult females and married men work in opposite directions, the British and German samples from 1773 may be relatively comparable. 26 Shipping patterns may have had a role in giving a supply “imprint” to the direction of immigration flows. However, the western voyage has been typically considered to be the return or back-­haul portion of the round trip voyage. This suggests that there may have been some excess capacity available for prospective immigrants. Anyway, ship-

The incidence of servitude, 1771–1804   179 ping patterns should not appreciably effect the analysis of the relative distribution of free versus servant immigrants across given colonies. 27 Servants received a nonvested pension or freedom dues at the end of their contract. In Pennsylvania by the 1770s this commonly consisted of two suits of clothes, one being new, and some farming tools, or 5 to 10 Pennsylvania pounds in lieu of these goods. This sum was about equal to one-­quarter of the original purchase price for German servants. 28 For a discussion of colonial Pennsylvania as “The Best Poor Man’s Country” see Lemon (1972). For a discussion of Irish and German servant migration to Pennsylvania see Geiser (1901), Herrick (1926), Salinger (1980), and Wokeck (1981). For an account of colonial destinations of English indentured servants prior to 1775 see Galenson (1981: 220–3). 29 See Galenson (1981: 117–40) on the replacement of unskilled servants with black slaves and the shift to skilled servants. Compensating differentials in the form of contract inducements needed to encourage servant migration to the Caribbean were apparently too expensive relative to purchasing slave labor given the relative cost of transatlantic passage. 30 Menard (1977) found that the proportion of servants to slaves in Maryland was 1 to 3 as early as 1685. 31 Dunn (1973: 95). 32 See Dunn (1973: 95, 197–8, 214–15) for an example of the size and capital requirements for sugar plantations. 33 See Dunn (1973) and Wood (1974: 9). 34 See Jones (1980: 59, 68, 77, 162–3, 265–6) and Dunn (1973: 96–8). 35 For example, Galenson (1981: 110) finds that English servants were offered shorter contracts as compensation for agreeing to go to the Caribbean as opposed to the mainland colonies. For a discussion of servant recruitment by various colonies see Smith (1947: 43–66).

References “Record of Indentures of Individuals Bound Out as Apprentices, Servants, etc. and of Germans and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773.” Manuscript held at the City Archives of Philadelphia. “Book of Redemptioners, 1785–1804.” Manuscript held at the Pennsylvania Historical Society, Philadelphia. Atack, J., and Bateman, F. (1981), “The ‘Egalitarian Ideal’ and the Distribution of Wealth in the Northern Agricultural Community: A Backward Look.” Review of Economics and Statistics 63, 124–9. Campbell, M. (1955), “English Emigration on the Eve of the American Revolution.” American Historical Review 61, 1–20. Deane, P. (1965), The First Industrial Nation. Cambridge, MA: Cambridge University Press. Dickson, R. J. (1966), Ulster Emigration to Colonial America, 1718–1775. London: Routledge & Kegan Paul. Dunn, R. S. (1973), Sugar and Slaves. New York: Norton. Durnbaugh, D. F. (1958), “Two Letters from Germantown.” Pennsylvania Magazine of History and Biography 83, 316–40. Fothergill, G., trans. (1908–1911), “Emigrants from England, December 1773 to April 1776.” New England Historical and Genealogical Register v. 62–65.

180   German immigrant servitude in America, 1745–1835 Galenson, D. W. (1981), White Servitude in Colonial America. Cambridge: Cambridge University Press. Galenson, D. W. (1984), “The Rise and Fall of Indentured Servitude in the Americas: An Economic Analysis.” Journal of Economic History 44, 1–26. Geiser, K. F. (1901), Redemptioners and Indentured Servants in the Colony and Commonwealth of Pennsylvania. New Haven, CT: Tuttle, Morehouse, and Taylor. Heavner, R. O. (1978), “Indentured Servitude: The Philadelphia Market, 1771–1773.” Journal of Economic History 38, 701–13. Heilbroner, R. L. and Singer, A. (1977), The Economic Transformation of America: 1600 to Present. New York: Harcourt Brace Jovanovich. Herrick, C. A. (1926), White Servitude in Pennsylvania. Philadelphia: John Joseph McVey. Jones, A. H. (1980), Wealth of a Nation to Be. New York: Columbia University Press. Kearl, J. R. and Pope, C. L. (1983), “The Life Cycle in Economic History.” Journal of Economic History 43, 149–58. Lebergott, S. (1984), The Americans: An Economic Record. New York: Norton. Lemon, J. T. (1972), The Best Poor Man’s Country. Baltimore, MD: Johns Hopkins Press. Main, J. T. (1983), “Standards of Living and Life Cycle in Colonial Connecticut.” Journal of Economic History 43, 159–65. Maynes, M. J. (1977), “Schooling the Masses: A Comparative Social History of Education in France and Germany, 1750–1850.” Ph.D. dissertation, University of Michigan. McCusker, J. J. (1978), Money and Exchange in Europe and America, 1600–1775, A Handbook. Chapel Hill: University of North Carolina Press. Menard, R. R. (1977), “From Servants to Slaves: The Transformation of the Chesapeake Labor System.” Southern Studies 16, 355–90. Mittelberger, G. (1960), Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754. In O. Handlin and J. Clive (trans. and eds.), Cambridge, MA: Harvard University Press. Morgan, E. S. (1975), American Freedom America Slavery. New York: Norton. Salinger, S. V. (1980), “Labor and Indentured Servitude in Colonial Pennsylvania.” Ph.D. dissertation, Los Angeles: University of California. Smith, A. E. (1947), Colonists in Bondage. New York: Norton. Strassburger, R. B. (1934), Pennsylvania German Pioneers. W. J. Hinke (Ed.). Norristown: Pennsylvania German Society. Tepper, M. (Ed.) (1978), Emigrants to Pennsylvania, 1641–1819. Baltimore, Md.: Genealogical Publishing Co. Trautmann, F. (1981), “Pennsylvania through a German’s Eyes: The Travels of Ludwig Gall, 1819–1820.” Pennsylvania Magazine of History and Biography 105, 35–65. Wokeck, M. (1981), “The Flow and Composition of German Immigration to Philadelphia, 1727–1775.” Pennsylvania Magazine of History and Biography 105, 249–78. Wood, P. H. (1974), Black Majority. New York: Norton.

10 Servant auction records, 1745–1831 The proportion of females among the servants*

In 1819 the Federal Government began documenting immigration. Prior to 1819 few systematic records were kept of the number and character of immigrants. Without comprehensive records, scholars have relied on a potpourri of incomplete and indirect sources to piece together quantitative estimates of immigration for the late colonial and early Republican periods. For example, loyalty oaths, shipboard passenger lists, newspaper reports, customhouse records, naturalization documents, and shipping patterns have been used to estimate different episodes of immigration into the Delaware Valley.1 The Philadelphia records of immigrant servant sales are another of these limited sources that can be used to describe pre-­1819 immigration. Servitude was the accepted method of paying for transatlantic passage among European immigrants when their savings proved deficient. The servant agreement was a written, negotiable contract which specified a fixed amount of future labor and other conditions of employment. There were several ways immigrants could use servant contracts to directly purchase or indirectly finance the voyage to America. The servant contract was sold soon after landing in America. Records of these sales are useful because approximately half of all immigrants to Pennsylvania in the late eighteenth century came as servants. In addition, these records provide economic and demographic information on the poorer half of the immigrant stream.2 Detailed quantitative estimates of the demographic characteristics of the population are an important part of the economic analysis of past societies. Economic historians use demographic information to explain labor market behavior and labor’s impact on economic growth. Immigrant labor usually differs from domestic labor in age, skill, propensity to work, etc. These differences can be used to explain the initial migration decision and the assimilation of immigrants into the American workplace. Servant records are useful for studying immigrant labor markets because demographic information can be directly combined with contract prices and lengths. Therefore, the records can be used both to estimate the magnitude and composition of immigration and to estimate the implicit market value of individual immigrant characteristics, for example, the value of being 20 instead of 15 years of age, male instead of female, literate instead of illiterate, Irish instead of German.3

182   German immigrant servitude in America, 1745–1835

The composition of the Philadelphia servant records The Delaware Valley was the last North American region to attract substantial numbers of European servants. The areas south of Pennsylvania had substituted slaves for servants by the eighteenth century, and those north of Pennsylvania had never attracted many immigrant servants. Philadelphia was the main port of entry and major servant market for the Delaware Valley. In addition, buyers from as far away as Florida, New Hampshire, and the Missouri Territories purchased servants in Philadelphia.4 Servant sales were systematically recorded before a Philadelphia magistrate. The recording served as legal protection, both for the master and the servant, against fraud or any unilateral attempt to alter the terms of the contract after the sale. Three years of recorded sales have survived from the colonial period, 1745–6 and 1771–3, and 34 years of recorded sales have survived from the early Republican period, 1785–1804 and 1817–31 (see Table 10.1). For the 1817–31 period 96 percent of the sales were concentrated in the years 1817 through 1819. In total, the four surviving documents enumerate over 12,000 contract entries, of which over 8,000 are the initial sale of newly arrived immigrant servants. The Philadelphia servant records are unique among the states and colonies of North America, particularly in terms of number and variety of contracts, comprehensiveness of coverage, attention to contract details, and ethnic mixture.5 The documents record the full names of the servants, selling agents, and buyers, contract characteristics such as the length of service, the date when the contract began, the sale price, and special stipulations such as being taught to read, receiving training or schooling, or getting extra payments at the conclusion of the contract. They also record the buyer’s residence and occupation and the servant’s European port of origin or ship. Table 10.1 presents some general characteristics and trends in the Philadelphia servant market by contract type. In the late colonial period, 70 to 80 percent of the new servants were immigrants. Local apprentices comprised 15 to 25 percent of the total number of new servants, while local servants made up less than 5 percent. Approximately 17 percent of the servants could expect to be resold in the late colonial period. The probability of being resold declined after the Revolution, falling to around 6 percent by the 1820s. In contrast, the probability of gaining an early release from the contract rose from near zero in the early 1770s to almost 6 percent by the 1820s. Finally, the probability of being purchased by an agent for some distant master rose from around 6 percent prior to 1804 to around 19 percent by the 1820s.6 Several problems arise when using the Philadelphia servant records. The four documents do not consistently cover all the various types of servant contracts or contract transactions. Initial contracts of arriving immigrants and resales of partially completed contracts are recorded in all four documents. Non-­immigrant servants and apprentices are recorded only in the two colonial documents, while servants who were released before their contracts were completed and reassignments of servants to absentee buyers are recorded only in the latter three

The proportion of females, 1745–1931   183 Table 10.1 Composition of the four Philadelphia servant records, 1745–1831, number (percent) Type of contract in the record

October 1745– October 1771– May 1785– November 1817– October 1746 October 1773 December 1804 November 1831

New servants:a Immigrant servants Local apprentices Local servants Total new servants

617 (78.1) 138 (17.5)   35 (4.4) 790

2,933 (71.3) 2,988 (24.0) 2,192 (4.7) 4,113

4,079 1,03– 1,03– 4,079

1,033 1,03– 1,03– 1,033

Other contract entries:b Resales 131 (16.6) Releases    – Reassignments    –

2,682 (16.6)    2,2 (0.1) 2,220 (5.4)

2,536 (13.1) 2,  88 (2.2) 2,244 (6.0)

2,  60 (5.8) 2,  58 (5.6) 2,192 (18.6)

5,017

4,947

1,343

Total contract entries

921

Sources: George W. Neible, ed., “Servants and Apprentices Bound and Assigned Before James Hamilton Mayor of Philadelphia, 1745,” Pennsylvania Magazine of History and Biography 30–2 (1906–8); “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773,” (Unpub. MS., City Archives of Philadelphia); “Book of Redemptioners, Book A: 1785–1804 and Book C: 1817–1831,” (Unpub. MSS., Pennsylvania Historical Society, Philadelphia). Notes The respective percentages of total new servants are in parentheses. The 1785–1804 and 1817–31 records were created only to enumerate immigrant German servants and the subsequent resale, release, or reassignment of their contracts. a Local servants were identified by those contracts listing no port of origin, a positive sale price, and, in most cases, notation of some reason for entering servitude, such as to pay for prison fees or to acquire money “for the servant’s own use.” Local apprentices were those whose contracts listed no port of origin, no sale price, the term “apprentice,” and the consent of a relative to the contract. Immigrant servants were those whose contracts listed a port of origin, ship, ship’s captain, immigrant selling agent, a contract commencement date indicating a time of arrival, or mentioned that payment was to cover passage costs. b Resales were partially completed contracts in which the present master “assigned the remaining time” on the servant’s contract to a new master for a positive price. Releases were partially completed contracts that were canceled and the servants set free by mutual consent, or partially completed contracts purchased by the relatives of the servants. Reassignments were contracts immediately transferred to a new owner by the initial purchaser the same day as the first sale was recorded without any contract or price changes. Reassignments were not resales but purchases by proxy or by agents for masters who could not attend the servant auction in Philadelphia. There were no releases or reassignments recorded in the 1745–6 document. Whether they were supposed to be recorded, or if in fact any occurred, is not known.

documents. The variety of contracts and servants can be difficult to sort out, especially in the colonial documents. For example, immigrant servants should be distinguished from local servants and apprentices, resales from reassignments and new servants, and immigrant identured servants from immigrant redemptioners. Sometimes close scrutiny of contract characteristics and recording methods is required to distinguish the various types of servants and contracts.7 Perhaps the most serious problem with the evidence is that in three of the four documents the ethnic composition is not representative of the immigrant servant

184   German immigrant servitude in America, 1745–1835 population as a whole. The post-­colonial documents intentionally record only German servants. The number of British servants immigrating after the Revolution is not known. The 1745–6 document records an unusually large proportion of Irish servants, 89 percent of the total. However, immigration to Pennsylvania between 1730 and 1754 was predominantly German, and substantial numbers of German servants probably entered Pennsylvania in these years. The high tide of German immigration collapsed briefly in 1745 because of migration and shipping disruptions caused by the War of the Austrian Succession.8 Therefore, the 1745–6 servant document is unrepresentative of the ethnic composition of the Philadelphia servant market in the middle of the eighteenth century, because the number of Germans arriving in 1745 was atypically low. Only the 1771–3 servant document is relatively representative of typical servant markets for its period, the years between the Seven Years War and the Revolution. Extreme care should be taken when estimating long-­run demographic changes in immigration and servitude across the four servant documents because of the non-­ uniformity and unrepresentativeness of the records.

Some demographic uses of the servant records Although many studies have drawn on aspects of the Philadelphia servant records, the documents have not been fully exploited, and much work still waits to be undertaken. One reason why the servant records have not been fully exploited is the non-­uniformity, unrepresentativeness, and difficulty in distinguishing contract types in the evidence. Another reason is that usually the records must be used in conjunction with other sources. Typically the servant’s name, ship on which he arrived, or the buyer’s name has to be cross-­referenced with other quantitative Pennsylvania records to supplement the information on individuals in the servant documents. The recent development of micro-­computer software which can sort long lists of names alphabetically has made this task easier.9 There are many examples of studies which could benefit from the matching of servant records to other Pennsylvania records. The incidence of servants running away could be measured by linking the servant records to newspaper advertisements for runaways. For example, 6.7 percent of the immigrant servants recorded in the 1745–6 document were subsequently advertised as runaways in the Pennsylvania Gazette.10 Tracing buyer names to county tax lists could be used to measure the relative wealth of servant owners, or tracing servant names to later county tax lists or other documents could be used to measure the persistence and success of servants as freemen. Matching the servant records to sources on shipping routes allows for better estimation of the ethnic origin of immigrant servants (see Table 10.2). Finally, expanding the demographic information on individual servants by matching servant records with other documents can also aid in studying the auction value of servant and contract characteristics.11 Estimating the incidence of servitude by individual characteristics is particularly important for exploiting the benefits of the servant records. This estimation

The proportion of females, 1745–1931   185 Table 10.2 The sexual composition of immigrant servants to Philadelphia by nationality, 1745–1831 Nationality or port of origin

Years

Germans

1745–6 1771–3

48 1,181

43.8% 30.8

(From London)

1771–3 1785–1804a 1817–31

134 3,511 1,033

6.0 32.3 40.5

1745–6 1771–3 1745–6 1771–3 1745–6 1771–3

177 475 312 545 28 226

19.8 33.7 9.0 28.4 10.7 21.7

1745–6 1771–3 1771–3

5 36 106

20.0 25.0 39.6

English

1745–6 1771–3

7 230

14.3 10.9

(All destinations)

1654–99b

13,113

23.3

(All destinations)

1700–44b

3,494

10.7

(All destinations)

1745–75b

3,972

9.0

Irish Ulster South Ireland Unassigned Irish Scots Glasgow Isle of Lewis

Total number of immigrant servants

Percentage female

Sources: Table 10.1. When nationality or ports of departure were not recorded in the servant documents they were derived through cross-referencing the time of arrival, the time of contract commencement, and the names of servants, ships’ captains, and ships in the servant records with information in the following sources: Strassburger, German Pioneers; The Pennsylvania Gazette (1728–89; reprint, 25 vols, Philadelphia: Microsurance, 1968), vols 18–19; “Passenger List, with duties;” Michael Tepper, ed., Emigrants to Pennsylvania 1641–1819 (Baltimore: Genealogical Publishing Co., 1978), 180–1. Notes Germans embarked at Rotterdam, Amsterdam, Hamburg, and Bremen. Ulster ports included Londonderry, Belfast, and Newry. South Irish ports included Dublin, Waterford, and Cork. English ports included London, Bristol, and Liverpool. a Immigrant servants who could not be cross-referenced to the German passenger lists in Strassburger, German Pioneers, vol. 3, were excluded. b Derived from Galenson, White Servitude, 219–27. Only 1,060 of the 20,579 English servants bound for the New World had Pennsylvania as their destination, and of this total only 9.9 percent were female.

involves matching individual servant names to the names on surviving passenger ship lists. There are two periods where nearly complete ship passenger lists coincide with the servant records, 1785 to 1804 for German immigrants and September 1771 to May 1772 for British immigrants. Matching the passenger lists with   the servant records, name by name, reveals that 45 percent of German immigrants arrived as servants between 1785 and 1804, and 66.7 percent of

186   German immigrant servitude in America, 1745–1835 South Irish immigrants, 18.3 percent of Ulster immigrants, and 19.7 percent of English immigrants arrived as servants between 1771 and 1773.12 Estimates of the proportion of immigrants who entered servitude can be used to estimate both 1) the magnitude of immigration in periods where servant records exist but immigration records do not, and 2) the number of immigrant servants in periods when immigration records exist but servant records do not. For example, little information exists on the magnitude of Irish immigration to Philadelphia in the 1740s beyond the 1745–6 servant document. A rough estimate of the total number of Irish immigrants arriving from October 1745 to October 1746 can be derived by dividing the number of servants arriving from Ulster and South Ireland in 1745–6 by their respective incidences of servitude estimated for 1771–2. This procedure yields 967 immigrants from Ulster, 468 immigrants from South Ireland, and some 70 from unknown Irish ports for a total Irish immigration of roughly 1,505 for the year ending in October 1746. This total estimate implies about 60 immigrants per passenger ship from Ulster and South Ireland in 1745–6.13 The immigrant per ship ratio can then be combined with information on ship arrivals to derive rough estimates of total Irish immigration for other years. The incidence of immigrant servitude can also be used to estimate the magnitude of servant immigration to Philadelphia in the 25 years prior to the Revolution. Multiplying the incidence of servitude by recent estimates of yearly German and Irish immigration indicates that approximately 22,530 servants entered Pennsylvania between 1750 and 1775 (see Table 10.3). Abbot E. Smith suggested that it was probable that more than 50,000 servants entered Pennsylvania between 1750 and 1775. Of these, he estimated that two-­thirds were Irish and one-­third were German.14 The new estimate made here is under half of Smith’s estimate, and their ethnic composition, 78.6 percent from Germany, 11.5 percent from Ulster, and 9.9 percent from South Ireland, is substantially different from Smith’s estimate. There are two causes for the divergence in these estimates. First, the new calculation may be a lower bound estimate, because the incidence of servitude may have been higher before 1771 which would bias the estimated number of ser­ vants downward.15 Second, Smith’s calculation may be an upper bound estimate, because he used a much higher incidence of servitude and a higher estimate of Irish immigration. Smith concluded that one-­half to two-­thirds of the immigrants came as servants.16 However, Ulster immigrants fell well below that mark and German immigrants only approached the lower reaches of Smith’s estimate of the incidence of servitude by the 1770s. The upper range of Smith’s incidence of servitude, estimated at two-­thirds of the total number of immigrants, originated in an erroneous estimate made in 1901 by Frederick Geiser of the proportion of Germans who arrived as servants in Pennsylvania between 1785 and 1804. Geiser compared the total number of servants with the total number of passengers rather than matching the names in the servant records with the names on the passenger lists. Without the benefit of complete passenger lists in 1901 and without distinguishing between new and

The proportion of females, 1745–1931   187 resold contracts, Geiser systematically over-­estimated the incidence of servitude in the 1785–1804 German evidence by around 21 percentage points.17 For the last 80 years Geiser’s erroneous estimate has been extensively reported in the literature as a definitive estimate of the incidence of servitude.18 Considering the  corrections in the estimated incidence of servitude, the new estimate of 22,530 German and Irish servants immigrating into Pennsylvania between 1750 to 1775 may be an improvement over Smith’s 50,000.

The proportion of females among immigrant servants There has been an increasing interest in the role of female labor in the transformation of early America.19 The sexual composition of immigrant servitude was an important demographic category. Sharon Salinger has suggested that the institution of servitude was being slowly confined to domestic house chores or women’s work over the last 70 years of the eighteenth century. Wage laborers were replacing servants in male oriented occupations, which left housekeeping as the last bastion of servitude. Salinger estimated that the female proportion of immigrant servants rose steadily from 14 percent in 1731 to 30 percent by 1772 and then to 40 percent by 1795. The rising proportion of women was seen as a deliberate response by servant recruiters and immigrants to the declining relative demand for male servant labor. Therefore, the increasing feminization of servitude would be one of the consequences of the transformation of the American economy from traditional bound labor to an economy based on market wage labor.20 Salinger’s conclusion that the female proportion of immigrant servants rose from 14 to 40 percent over the final 70 years of the eighteenth century may be an artifact of the unrepresentativeness of her evidence. During the eighteenth century the majority of immigrant servants were Irish or German. The sexual composition of servant groups varied substantially across ethnic lines and depended primarily on the proportion of families among the various immigrant groups. The proportion of females was traditionally highest among the Germans and lowest among the English.21 Therefore, the ethnic representativeness of the servant evidence would be important for assessing the sexual composition of immigrant servitude. Salinger used portions of the first three servant documents listed in Table 10.1, and a small sample of English servants who emigrated between 1718 and 1731.22 These samples are not representative of the general German–Irish composition of Pennsylvania servitude. The samples progress chronologically from English (1718–31) to Irish (1745–6) to a mixture of Irish and German (1771–3) and finally to all Germans (1785–1804), or from groups traditionally incor­ porating a low proportion of females to groups that incorporated a high proportion of females. Therefore, comparisons across the four samples provide a false impression of a rising share of females among immigrant servants. This false impression is an example of sample selection bias caused by the historic accident of how and when servant records were kept and preserved.

1729–49a 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767

Year

189 134 135 54 80 26 80 80 0 26 26 0 0 21 75 30 60 90 75

19 13 13 5 8 3 8 8 0 3 3 0 0 8 30 12 24 36 30

34 35 0 23 35 35 12 6 35 23 18 34 34 54 60 54 40 126 54

Male

Male

Female

Ulster

South Ireland

9 9 0 6 9 9 3 1 9 6 4 17 17 27 31 27 20 65 27

Female 678 1,187 1,290 1,550 1,320 1,469 82 35 0 0 0 0 24 0 233 934 266 206 336

Male

Germany

529 925 1,006 1,208 1,029 1,144 64 27 0 0 0 0 10 0 92 370 105 81 134

Female 557 947 1,019 1,219 1,046 1,156 75 36 9 9 7 17 27 35 153 409 149 182 191

Total female servantsb

Table 10.3 Yearly estimates of the sexual composition of German and Irish servant immigration to Philadelphia, 1729–75

38.2% 41.1 44.1 42.8 42.2 43.0 30.1 22.9 20.5 15.5 13.7 33.3 31.4 29.4 29.7 28.7 28.9 30.1 29.1

Total percentage female

3,800

Totals

710

38 53 51 47 27 47 12 24 1,998

49 65 79 145 139 185 131 153 867

26 34 41 74 71 95 67 78 18,247

240 100 162 319 278 546 161 51 12,746

95 39 65 126 110 217 64 20 14,323

159 126 157 247 208 359 143 122 37.3

29.3 29.8 29.8 29.8 30.1 30.1 30.8 31.6

Notes The female servant immigration was estimated separately for each ethnic group by multiplying the yearly immigration by a constant estimated percentage of the number of immigrants who entered as servants and then multiplied by a constant estimated percentage of servants who were female. From 1729 to 1760 this calculation was (immigration * 0.654 * 0.09) for South Ireland, (immigration * 0.183 * 0.198) for Ulster, and (immigration * 0.45 * 0.438) for Germany. From 1761 to 1775 this calculation was (immigration * 0.667 * 0.284) for South Ireland, (immigration * 0.183 * 0.337) for Ulster, and (immigration * 0.45 * 0.284) for Germany. Male servant immigration was calculated by subtracting the estimated female servant immigration from the estimated total servant immigration for each ethnic group. See the text for further discussion. a Because estimates of Irish and especially Ulster immigration between 1729 and 1749 are relatively sketchy, the period was combined and yearly averages for these 11 years were not reported. b This column was the summation of German, South Irish, and Ulster female servants only.

Sources: South Irish and Ulster yearly immigration was derived from Wokeck, “A Tide of Alien Tongues,” 267–8; Dickson, Ulster Emigration, 58–69, 234–81; Audrey Lockhart, Some Aspects of Emigration from Ireland to the North American Colonies Between 1660 and 1775 (New York: Arno Press, 1976), 30–60, 175–208. Yearly German immigration was taken from Wokeck, “German Immigration,” 260–1. The proportion of German immigrants who became servants was 45 percent during 1785–1804: Grubb, “The Incidence of Servitude,” 320. The proportion of Irish immigrants who became servants was 66.7 percent for South Irish and 18.3 percent for Ulster Irish during 1771–3, derived from “Passenger List, with duties, 1768–1772”; “Record of Indentures, 1771–1773”; Grubb, “The Incidence of Servitude,” 327. The female servant proportion for each ethnic group was taken from Table 10.2.

95 132 129 117 67 99 30 60

1768 1769 1770 1771 1772 1773 1774 1775

190   German immigrant servitude in America, 1745–1835 Table 10.2 reports the share of females among immigrant servants by their region of origin for the four Philadelphia servant documents listed in Table 10.1. Germans consistently contributed the highest share of females, with 31 to 44 percent of the total being female, followed by Ulster Irish and Scots, then by the southern Irish, and finally by the English with the lowest share at around 11 to 14 percent. Conclusive evidence for the increasing feminization of immigrant servitude in the 30 years before the Revolution can be found only among the Irish.23 The share of females increased by 14 percentage points for Ulster Irish and by 17 percentage points for southern Irish between 1745 and 1772. The share of females among Irish servants had grown nearly to the level of the Germans by 1772 at just under 30 percent. This trend is consistent with the rising proportion of families among Irish immigrants in this period.24 In contrast, the share of females among German and English servants may have been declining in the 30 years before the Revolution. Among all servants leaving England, the share of females had experienced a long run decline for over a century prior to 1775. English servants arriving in Pennsylvania showed a similar trend, but the samples are too small to be reliable. The female share among German servants oscillated, declining from just over 40 percent at mid-­ century to a low of just under 30 percent in the 1770s and then slowly rising back to 40 percent by the 1820s. The declining female share prior to the Revolution is consistent with the declining ratio of women and children to adult men among all German immigrants, from 1.77 in 1727–56 to 1.35 in 1761–75, although the smallness of the 1745 sample makes this conclusion somewhat tentative.25 In the 30 years before the Revolution the share of females among the two major groups of servants arriving in the Delaware Valley was moving in opposite directions, rising among the Irish and probably declining among the Germans. For the same period the magnitude of total immigration was also moving in the opposite direction for the two groups, increasing for the Irish and decreasing for the Germans. Whether female immigrant servitude was increasing before the Revolution, either in relative share or absolute number, would depend both on the change in the proportion of females among German and Irish servants and on the change in the absolute number of servants arriving from Germany and Ireland. English and Scottish servant immigration was probably too small to influence the overall trend. Table 10.3 presents yearly estimates from 1750 to 1775 of the number of male and female servants arriving in Philadelphia from Germany, Ulster, and South Ireland.26 These estimates are based on a combination of estimates for yearly immigration, the proportion of servants among the immigrants, and the share of females among the immigrants for each ethnic group. The calculation involves several extrapolations and uses various component estimates which also rely on incomplete evidence and extrapolation. The immigration estimates were converted to servant estimates using the incidence of servitude found for the Irish in 1771–2 and for the Germans in 1785–1804. Because the incidence of servitude may have been higher earlier in the century, when immigrants may

The proportion of females, 1745–1931   191 have been poorer, the absolute number of servants may be biased downward in the earlier years.27 Therefore, the numbers presented in Table 10.3 should be regarded as crude estimates or best guesses which tentatively outline the general trends in servant shares and magnitudes. Table 10.3 shows a long run defeminization of immigrant servitude in the 30 years before the Revolution. The share of females among immigrant servants was around 40 percent prior to the Seven Years War and then fell to around 30 percent after the Seven Years War. In absolute numbers, female servitude peaked in the early 1750s at just over 1,000 a year. In contrast, female servant immigration per year after the Seven Years War fell to under one-­fifth of the 1750s peak or to roughly 200 a year. The absolute decline might be even larger given the potential downward bias in the early years.28 The magnitude and variance in German immigration was the major determinant of the trends in female servitude. German immigration experienced a relative peacetime low, both in magnitude and in the relative share of females, during the 1760s and 1770s.29 This decline swamped the trends in female servitude among other immigrant groups and most likely produced a relative and total decline in female servitude prior to the Revolution. By contrast, the female share of immigrant servants may have been rising in the 45 years after the Revolution. The evidence for this period is exclusively from German servants arriving in Philadelphia; see Table 10.2. However, German immigration probably dominated the trends in the sexual composition of immigrant servitude after the Revolution just as it did before the Revolution. The share of females increased by 12 percentage points among German servants from the 1770s to the 1820s. Although the evidence indicates a rising feminization of servitude after the Revolution, it could also be viewed as a return to the pre-­1760 female proportion. Finally, while the share of female servants among German immigrants was rising after the Revolution, the absolute magnitude was declining, from roughly 186 a year in the 1770s to 60 a year in 1785–1804 and then to near zero by the end of the 1820s. Table 10.4 decomposes female servitude into the categories of married, single adult, and children from the German evidence. This decomposition cannot be done as readily for the British. The presence of families entering servitude, Table 10.4 Age and marital status of female servants among German immigrants to Philadelphia, 1771–1831 Female servants

Number (percent) 1771–3

1785–1804

Married Dependent children Single adults

  59    –    –

(4.5%) (23.9)

Total

373

(28.4)

326 419 389 1,134

1817–31

(9.3%) (11.9) (11.1)

  94 139 185

(9.1%) (13.5) (17.9)

(32.3)

418

(40.5)

Sources: Tables 10.1 and 10.2; Grubb, “The Incidence of Servitude,” 320–4.

192   German immigrant servitude in America, 1745–1835 married women and female children, contributed substantially to the share of female servants among the Germans, accounting for 56 to 66 percent of all German female servants. If German families were removed from the estimate, the remaining German female proportion would more closely resemble the proportion among the English, at around 11 to 14 percent. The English also had few married women or female children among their servant emigrants. Therefore, the relative share of single adult females among immigrant servants may have been roughly similar across ethnic groups. This finding suggests that changes in the relative participation of immigrant families in servitude and changes in the relative magnitude of family migration may explain most of the changes in female servitude over time. For example, the increase in the proportion of females from 1771–3 to 1785–1804 in Table 10.4 can be explained completely by the relative increase in married servants. However, the subsequent increase in the female proportion from 1785–1804 to 1817–31 is primarily explained by the relative increase in single adult female servants. Whether this latter shift was caused by changes in the family composition of German immigration after the War of 1812 is not yet known. In conclusion, females were probably an important part of immigrant servitude in Pennsylvania throughout the eighteenth and early nineteenth centuries, comprising between 28 to 45 percent of all immigrant servants. The course of female immigrant servitude can be summarized as steadily declining in magnitude from 1750 to 1830, while declining in relative share from 1750 to 1775 but rising in relative share from 1785 to 1830. The underlying causes of the trends are difficult to explain. In general, the trends were dominated by the magnitude, variance, and characteristics of German immigration throughout the period, but what caused changes in German immigration is not well understood. The rise of wage labor in the Pennsylvania economy may have had little direct connection to the changing trends in female servitude, considering the opposite trends among various ethnic groups and the nonlinear overall trend. More work will have to be done on the relative causes of immigration and servitude by ethnic origin and family status before changes in the sexual composition of immigrant servitude can be better explained.

Summary The Philadelphia servant documents are unique records which can substantially add to our understanding of immigration, servitude, and labor markets in the late eighteenth and early nineteenth centuries. Careful interpretation of the servant documents and supplemental evidence from other sources are usually required to make full use of the servant evidence. The characteristics, various problems of interpretation, and potential applications of the servant documents were presented and discussed. This information may help facilitate future research. Finally, the servant evidence was used to chart the course of female servitude in the Delaware Valley from 1750 to 1830. This application demonstrated the importance of controlling the ethnic, marital, and familial composition of the

The proportion of females, 1745–1931   193 servant evidence to avoid sample selection bias. In addition, previous estimates of the magnitude, ethnic composition, and sexual composition of immigrant servitude were revised and extended.

Notes   * Originally published as Farley Grubb, “Servant auction records and immigration into the Delaware valley, 1745–1831: the proportion of females among immigrant servants,” Proceedings of the American Philosophical Society, vol. 133, no. 2 (June 1989): 154–69. Copyright 1989. © 1989 by the American Philosophical Society. Reprinted with permission.   This paper is based on “Using Servant Auction Documents to Study Immigrant Demographics and Labor Market Values in the Delaware Valley,” presented at the Symposium on the Demographic History of the Philadelphia Region, 1680–1860, Session on Immigration, American Philosophical Society, Philadelphia, May 1987. The author wishes to thank Stanley Engerman, Susan Klepp, Marianne Wokeck, and the participants in the symposium for helpful comments. Please address any correspondence to Farley Grubb, Economics Department, University of Delaware, Newark, DE 19716.   1 For example, see Bernard Bailyn, Voyagers to the West (New York: Alfred A. Knopf, 1986); Edward C. Carter II, “A ‘Wild Irishman’ Under Every Federalist’s Bed: Naturalization in Philadelphia, 1789–1806,” Pennsylvania Magazine of History and Biog­ raphy 94 (July 1970): 331–46; R.J. Dickson, Ulster Emigration to Colonial America, 1718–1775 (London: Routledge and Kegan Paul, 1966); Henry A. Gemery, “British Immigration to the U.S., 1772–1812: Evidence from the Marshals’ Returns of Enemy Aliens,” (Unpub. MS., Colby College, Waterville, Me.); Henry A. Gemery, “European Emigration to North America, 1700–1820; Numbers and Quasi-­Numbers,” Per­ spectives in American History, New Series, 1 (1984): 283–342; Ralph B. Strassburger, Pennsylvania German Pioneers 3 vols., ed. William J. Hinke (Norristown, PA: Pennsylvania German Society, 1934); Marianne S. Wokeck, “A Tide of Alien Tongues: The Flow and Ebb of German Immigration to Pennsylvania, 1683–1776,” (Ph.D. Diss., Temple University, 1983); Marianne S. Wokeck, “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography 105 (July 1981): 249–78.   2 For a general discussion of colonial servitude, see Abbot E. Smith, Colonists in Bondage (New York: W.W. Norton, 1947). For further discussion of the different types of servant contracts and the proportion of immigrants who came as servants, see Farley Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History 22 (July 1985): 316–39; Farley Grubb, “Redemptioner Immigration to Pennsylvania: Evidence on Contract Choice and Profitability,” Journal of Economic History 46 (June 1986): 407–11; Farley Grubb, “The Market for Indentured Immigrants: Evidence on the Efficiency of Forward-­Labor Contracting in Philadelphia, 1745–1773,” Journal of Economic History 45 (Dec. 1985): 855–68.   3 For example, see David W. Galenson, White Servitude in Colonial America (Cambridge, MA: Cambridge University Press, 1981); Grubb, “The Market for Indentured Immigrants,” 855–68; Farley Grubb, “Colonial Labor Markets and the Length of Indenture: Further Evidence,” Explorations in Economic History 24 (Jan. 1987): 101–6; Farley Grubb, “The Colonial Auction of Immigrant Servants: The Redemptioner Method,” (Unpub. MS., University of Delaware, Newark, DE).   4 Smith, Colonists in Bondage, 335–7; Farley Grubb, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­ Atlantic Economy,” Social Science History 9 (Summer 1985): 249–75. Baltimore also

194   German immigrant servitude in America, 1745–1835 received numbers of servants in the second half of the eighteenth century and was probably the second largest servant market in English America in this period.   5 I have not been able to find any systematic servant auction records of comparable magnitude for other locations in North America. The only other sizable set of servant records relating to English America are for English indentured servants who had their contracts recorded before embarking for America; for example, see Jack Kaminkow and Marion Kaminkow, A List of Emigrants from England to America, 1718–1759 (Baltimore: Magna Charta Book Co., 1964); Galenson, White Servitude. These records do not include information on the American auction of the servant, such as sale prices or purchaser characteristics.   6 The surviving servant documents are slices of continuous, but mostly lost, records. The resales include both resales of new servants in the surviving document itself and resales of new servants that arrived before the particular document started. New ser­ vants in a particular surviving document were also resold after that document had been ended. Therefore, the proportion of resales in a surviving document estimates the chances of being resold under the assumption that immigration and the propensity to resell servants were relatively constant for several years before and after each surviving document.   7 See the discussion in Table 10.1. For the distinction between immigrant indentured and redemptioner servants, see Grubb, “Redemptioner Immigration to Pennsylvania,” 407–11. Finally, the version of the 1771–3 servant document published in the Penn­ sylvania German Society 16 (1907) and reprinted by the American Philosophical Society is highly edited and leaves out crucial information needed to distinguish the type and source of the contracts. The original 1771–3 servant document held at the Philadelphia City Archives should be preferred over these other published versions.   8 Smith, Colonists in Bondage, 320; Wokeck, “German Immigration,” 260–1.   9 I have alphabetically sorted the 1745–6 servant record in three different ways, by servant name, buyer name, and county of residence of the buyer. Sorting the other servant documents is currently underway. Please write if you would like one of these alphabetically sorted copies of the servant records. 10 The runaway percentage should be considered lower bound. Not all masters may have advertised their runaways or have advertised in the Pennsylvania Gazette, though few advertisements for runaways were found in the German newspapers. What percentage of the runaways were caught and prosecuted is unknown. 11 For example, see Grubb, “The Market for Indentured Immigrants,” 855–68; Grubb, “Colonial Labor Markets and the Length of Indenture,” 101–6; Grubb, “The Redemptioner Method.” 12 Derived from “Record of Indentures”; “Passenger List, with duties, August 29, 1768 to August 13, 1772,” (Philadelphia Customs House Records, Cadwalader Collection, Thomas Cadwalader, Box 15 T, MS. Historical Society of Pennsylvania, Philadelphia); Grubb, “The Incidence of Servitude,” 320; Wokeck, “A Tide of Alien Tongues,” 244–311. 13 This total immigration estimate may be biased upward because the incidence of servitude may have been higher earlier in the century. As wealth increased over time the reliance on servitude to finance migration may have declined; see David Galenson, “The Rise and Fall of Indentured Servitude in the Americas: An Economic Analysis,” Journal of Economic History 44 (Mar. 1984): 1–26; Grubb, “The Incidence of Servitude,” 319–32. Perhaps the only account of the incidence of servitude from earlier in the century is for passengers and servants arriving in the year 1729/30 in the Amer­ ican Weekly Mercury. This account indicates a servant percentage of zero among German immigrants, 100 percent among Scots, 19.9 percent among Irish, and 25.5 percent among English and Welsh immigrants. However, it may not be a reliable indication of the incidence of servitude in this year, because the report is based on ship manifests and not on servant sales records. Many immigrants were beginning to use

The proportion of females, 1745–1931   195 the redemptioner method, by which the immigrant would not sign a servant contract until after arrival in America. Therefore, redemptioner servants may not have been counted as servants on ship manifests, thus understating the true incidence of servitude; see Grubb, “Redemptioner Immigration to Pennsylvania,” 407–11. 14 Smith, Colonists in Bondage, 337. Smith does not explain the procedure he used to derive his estimates of total servant immigration. He presented evidence on servitude and immigration in his appendix along with statements on the proportion of servants by contemporary observers, and then he gave his best guess on servant totals without elaborating on the method used to make the estimate. 15 The incidence of servitude was derived from evidence between 1771 and 1804. Projecting these estimates back to 1750 may systematically undercount the true number of servants, which corresponds to a given number of immigrants because the share of immigrants who used servitude to finance migration may have been higher in earlier years; see notes 13 and 27. 16 Smith, Colonists in Bondage, 320–3, 335–7. His estimate of total German immigration is very close to the estimate used here in Table 10.3. Therefore, the different magnitude and ethnic composition of servant immigration to Pennsylvania estimated by Smith must result from both a relatively higher estimate of Irish migration and a higher incidence of servitude used by Smith. 17 Karl F. Geiser, Redemptioners and Indentured Servants in the Colony and Common­ wealth of Pennsylvania (New Haven: Tuttle, Morehouse, and Taylor, 1901), 41; Grubb, “The Incidence of Servitude,” 317–20. Geiser estimated the proportion of servants among the immigrants to be 66 percent, while Grubb put it at 45 percent. 18 Geiser’s estimate forms the basis for the estimate in Cheesman A. Herrick, White Ser­ vitude in Pennsylvania (Philadelphia: John Joseph McVey, 1926), 177; Smith, Colo­ nists in Bondage, 323. These sources in turn are those used by many subsequent studies, for example, see Galenson, White Servitude, 3–4; Robert O. Heavner, “Indentured Servitude: The Philadelphia Market, 1771–1773,” Journal of Economic History 38 (Sept. 1978): 702; Robert Heilbroner and Aaron Singer, The Economic Trans­ formation of America: 1600 to Present (New York: Harcourt Brace Jovanovich, 1977), 40; James T. Lemon, The Best Poor Man’s Country (Baltimore: Johns Hopkins University Press, 1972), 10. 19 For example, see Sharon V. Salinger, “Send No More Women: Female Servants in Eighteenth-­Century Philadelphia,” Pennsylvania Magazine of History and Biography 107 (Jan. 1983): 29–48; Carole Shammas, “The Female Social Structure of Philadelphia in 1775,” Pennsylvania Magazine of History and Biography 107 (Jan. 1983): 69–83; Jean R. Soderlund, “Black Women in Colonial Pennsylvania,” Pennsylvania Magazine of History and Biography 107 (Jan. 1983): 49–68; Claudia Goldin and Kenneth Sokoloff, “Women, Children, and Industrialization in the Early Republic: Evidence from the Manufacturing Censuses,” Journal of Economic History 42 (Dec. 1982): 741–74. 20 Sharon V. Salinger, “Artisans, Journeymen, and the Transformation of Labor in Late Eighteenth-­Century Philadelphia,” William and Mary Quarterly 3rd Series 40 (Jan. 1983): 62–84. For evidence indicating that female immigrant servants were systematically employed in different occupations relative to male servants, see Grubb, “Immigrant Servant Labor,” 249–75. 21 Within a given ethnic group shifts in the share of female servants may have been largely determined by changes in the relative participation of families in servitude and by changes in the relative participation of families in migration. Single immigrants were predominantly male while immigrant families had similar numbers of males and females. 22 Salinger, “The Transformation of Labor,” 68. Her evidence on English immigrants comes from English emigration records and not Pennsylvania immigration records; see note 5.

196   German immigrant servitude in America, 1745–1835 23 There was also a small increase in the Glasgow Scottish samples, but Scottish servant immigration may be too small to be of much importance. 24 See Wokeck, “A Tide of Alien Tongues,” 276–7. In addition, the relatively high share of females among Scots from the Isle of Lewis coincides with a relatively high proportion of families among these servants. 25 Wokeck, “German Immigration,” 260–1. 26 The period from 1750 to 1775 was chosen because relatively reliable estimates of yearly immigration have been published for all three groups; see Table 10.3. Reliable estimates of Irish immigration and servitude for the two decades after the Revolution need to be made before this analysis can be extended into the 1790s. 27 See notes 13 and 15. The estimation procedure in Table 10.3 understates the true number of servants arriving in 1771–3, as found in Table 10.2, by about 20 percent for the Germans and 40 percent for the Irish. This error in Table 10.3 for the number of servants arriving in 1771–3 is a difference between the 1750–75 estimating algorithm and the true count from the 1771–3 servant document. The source of this error in the German servant estimate was probably caused by the incidence of servitude being higher in 1771–3 than in 1785–1804 where it was estimated; see Table 10.3. The estimates of total German immigration are relatively reliable. The source of this error in the Irish servant estimate was probably caused by underestimation of Irish immigration for the years 1771–3. Irish immigration estimates are based on less direct and less reliable evidence than German immigration estimates. The incidence of servitude and share of females used in the Irish servant estimate were taken directly from the 1771–2 evidence. Correcting the errors in Table 10.3 just for the years 1771–3 lowers the percentage of female servants by 1 percentage point but raises the yearly magnitude of female servants by about 100 for 1771–3. These adjustments would only slightly modify the conclusions on the general direction of the trends depicted in Table 10.3. The estimates reported in Table 10.3 were not corrected for these errors in 1771–3, because the estimates in earlier years rely on the underlying trends in the base immigration estimates which in turn rely on involved algorithms, see Wokeck, “A Tide of Alien Tongues,” 103–36, 244–311. The integrity of the immigration estimates was preserved because the trend from 1750 through 1775 was the major interest. 28 The general trends are relatively robust to alternative calculations using reasonable assumptions within the limits of the physical evidence; see note 27. For example, a “best case” for an increase in female servitude between 1750 and 1775 might be made by using the 1771–3 female share for Germans in years prior to the Seven Years War rather than the 1745 share which came from a small sample. This alternative calculation would modify the trends in Table 10.3, producing a 50 percent decline in magnitude rather than an 80 percent decline and producing almost a constant female share rather than a share that fell 10 percentage points. Therefore, even under some of the most favorable sets of reasonable assumptions, it is almost impossible to produce upward trends in the share or magnitude of female servants between 1750 and 1775. 29 Wokeck, “German Immigration,” 260–1.

11 The occupational and geographical distribution of immigrant servant labor in the Delaware valley*

Contract labor played a critical role both in financing European transatlantic migration and in providing a hirable labor force to work the estates of the New World. During the seventeenth century at least three-­quarters of the Chesapeake colonists arrived under some form of short term contract (Walsh 1977: 111). By the American Revolution, a majority of English, German, and southern Irish emigrants still used servant contracts to finance their migration to Pennsylvania (Grubb 1985). For the year 1773, 61 percent of the 387 southern Irish immigrants, 18 percent of the 1,420 Ulster immigrants, 25 percent of the 382 Scotch immigrants, and 52 percent of the 174 English immigrants to Pennsylvania entered servitude. For the years 1771–3, out of 747 German adult male immigrants to Pennsylvania 58 percent entered servitude. For 1785–1804, 45 percent of the 7,837 German immigrants to Pennsylvania entered servitude. Colonial entrepreneurs relied on the continuing arrival of bound immigrant workers as one of their principal sources of employable labor. In a frontier environment where land was plentiful and easily acquired even by those with little capital, free colonists after saving a small amount would tend to enter self-­ employment either as farmers or merchants. Under such conditions colonial planters and entrepreneurs who desired to expand their farm or business would be prompted to rely on unfree labor, i.e., immigrant servants or black slaves. In the seventeenth century, immigrant servants were mostly Englishmen and they were employed largely as field laborers on the agricultural estates of the Chesapeake and Caribbean colonies. By the beginning of the eighteenth century these colonies had shifted to reliance on slave labor for field work (Dunn, 1973; Galenson, 1981a; Galenson, 1981b; Main, 1977; Menard, 1977; Morgan, 1975; and Wood, 1974). The few immigrant servants who still migrated to plantations in these regions tended to be employed in skilled occupations rather than as field laborers (Galenson, 1981b). By the latter half of the eighteenth century, the mid-­ Atlantic was the only region among the colonies having relied at one time on imported servant labor that had not shifted to slave labor. Immigrant servants had for the most part abandoned emigration to colonies south of Pennsylvania but they continued to arrive on the Delaware throughout the eighteenth century in great numbers. This continuing supply of new immigrant servants may have been partly responsible for keeping Pennsylvania from

198   German immigrant servitude in America, 1745–1835 adopting slave labor on a large scale. The major supply of servants in this later period was from Irish and German immigrants. For the Philadelphia market, Irish immigrant servants comprised 92 percent of the immigrant servant sales in 1745 and 42 percent in 1771–3 (see City Archives of Pennsylvania and Pennsylvania Historical Society). For the 1771–3 market Germans comprised 42 percent, English 8 percent and Scotch 5 percent of the immigrant sales. The rather low percentage of German servants arriving in 1745 was due to an abnormally low German immigration that year. In the late 1730s and early 1750s German immigration to Pennsylvania was at its colonial zenith (Wokeck 1981; 260). Of some 3,719 English servants who emigrated between 1773 and 1776, 18 percent went to Pennsylvania and 77 percent went to the Chesapeake (Grubb 1985; Galenson 1981a: 222). The mid-­Atlantic economy was becoming more complex and diverse as the colonial period progressed. How were immigrant servants integrated into this economy? Where did they work and at what occupations were they employed? Were there significant regional and occupational differences in the demand for males versus females, adults versus children, and British versus German ser­ vants? Did having a choice of masters under a given contracting arrangement affect the regional distribution of servant labor? Answering these questions will help determine the composition and trends of the colonial demand for bound immigrant labor and help reveal how important immigrant servitude was to the mid-­Atlantic economy.

Sources The evidence for this study is derived from the records of servant contract sales for the Philadelphia market. Philadelphia was the main port of entry into the Delaware region and by the late eighteenth century it possessed the largest single market for bound immigrant labor in North America. Consequently, it attracted buyers from throughout the Delaware and Susquehanna valleys. Occasionally buyers from distant colonies could also be found participating in this market. Mittelberger (1960) claimed that in 1750, everyday Englishmen, Dutchmen, and High Germans came from very far away, sometimes a twenty, thirty, or forty hour journey to buy immigrant servants in Philadelphia. The Philadelphia market was also unique in that periodically the civil authorities recorded the salient features of all bound labor contract sales. These records have survived for the years 1745–6, 1771–3, and 1787–1804. For all three samples, when a servant was sold or resold the transaction was recorded before a government official who also listed the town, township, county, and colony of the buyer along with the other sale information.1 For the latter sample only, the officials also recorded the occupation of the buyer. This evidence allows for separating immigrant from nonimmigrant servants and for estimating the total regional distribution of immigrant servants both by nationality and family status.2 This is something that previous studies could not accomplish because they relied on tax and probate records, usually for small specific localities, which

The distribution in the Delaware valley   199 do not record these distinctions, e.g., see Nash (1973); Salinger (1981); Salinger (1983); and Tully (1967). Table 11.1 presents the geographic distribution of immigrant servants for the three samples combined. The city of Philadelphia was the largest single demander, purchasing 38 percent of the servants. Servitude, however, remained a strong rural phenomenon in the surrounding region. The major agricultural counties adjacent to Philadelphia, plus Lancaster county, accounted for roughly 44 percent of the servants purchased. The outlying counties of Pennsylvania, New Jersey, and the Chesapeake region only acquired 15 percent of these immigrant servants. Table 11.1 also indicates that significant geographic differences existed in the demand for labor between males, females, and married servants. The distribution for single males approximates the total distribution very closely. However, single females are far more city oriented with over half being purchased by Philadelphians. By contrast, married couples were more likely to go to outlying counties on the frontier or into the Chesapeake region. Married couples may have been viewed as more stable or settled. This quality may have been needed on the frontier in order to lower the incidence of running away due to dissatisfaction with the quality of frontier life. In addition, married couples may have been easily enticed to these regions with the promise of farm land as part of their freedom dues paid at the end of the contract.3 The relatively high demand for female servants in the city may be reflecting the relative productivity advantage of females in domestic or retail service compared to field labor. Table 11.1 The geographic distribution of immigrant servant purchasers: the Philadelphia market, 1745–1804 Residence of purchaser

Total immigrant Single males servants (N = 4,216) (N = 6,385)

Single females Married persons (N = 1,437) (N = 732)

Philadelphia City Philadelphia Co.a Lancaster Co., PA Chester Co., PA Gloucester Co., NJ Burlington Co., NJ Berks Co., PA Delaware, MD, and Virginia Bucks Co., PA

38.40% 12.01 11.45   8.08   3.85   3.18   2.77   2.76

35.77% 12.08 12.20   9.45   3.82   3.54   2.54   3.18

51.84% 12.80 11.41   6.40   3.27   2.16   2.46   1.18

27.21% 10.07   7.21   3.54   5.17   3.13   6.67   3.40

  2.27

  2.59

  1.04

  2.86

Total

84.78%

85.16%

91.58%

69.25%

b

Source: Evidence for the year 1745–6 was taken from Neible (1906–8). The evidence for the years 1771–3 was taken from the City Archives of Philadelphia, October 3, 1771 to October 5, 1773. The evidence for the years 1785–1804 was taken from the Pennsylvania Historical Society. Notes a This excludes Philadelphia City. b The absent category contains those who did not have there residence recorded, 3.02% and purchasers from outlying counties and regions, 12.20%.

200   German immigrant servitude in America, 1745–1835 The distribution of new immigrant servants may simply be the result of the distribution of the established colonial population. To check this possibility, the number of immigrant servants purchased as a percentage of the established population was estimated for the 1771–3 sample (see Table 11.2).4 In general, the geographic ranking of servant purchases shown in Table 11.1 is maintained Table 11.2 Distribution of immigrant servants and slaves by the percentage of resident population for select counties, 1771–3 County of residence

Pennsylvania (1779) Philadelphia City Chester Co. Lancaster Co. Philadelphia Co.b Bucks Co. Cumberland Co. Berks Co. Northumberland Co. Northampton Co. Bedford Co. York Co. Westmoreland Co. Washington Co. Total New Jersey (1773) Gloucester Co. Burlington Co. Salem Co. Cape May Co. Hunterdon Co. Sussex Co. Cumberland Co. Middlesex Co. Morris Co. Bergen, Sommerset, Essex, & Monmouth Total Delaware Colony (1774)

Estimated population

34,250 21,061 35,493 36,685 18,463 24,279 22,900 6,985 15,495 7,905 30,748 17,841 13,938

Percentage black

Percentage of immigrant servants purchased from 1771–3

2.94%a . . . 1.24 . . . . . . 2.67 0.72 . . . . . . 0.35 1.46 . . . . . .

2.72% 1.61 1.18 0.86 0.42 0.41 0.36 0.20 0.17 0.11 0.07 0.01 0.00

285,998 8,754 13,124 5,960 1,759 15,605 9,229 5,057 10,204 11,535 40,776

0.82% 3.61% 5.57 5.00 6.61 7.02 3.09 2.18 . . . 3.18 . . .

122,003 37,219

0.87% 0.81 0.55 0.11 0.08 0.08 0.08 0.05 0.03 0.00 0.20%

5.37%

0.13%

Sources: See Table 1 for the source of the servant distribution. The estimated population and black proportion was taken from Sutherland (1936), except for the population of Philadelphia City which was taken from Smith (1977) and Alexander (1974). Notes a Nash (1973) estimated the average number of slaves in Philadelphia City between 1772 and 1773 to be 1,007. b Philadelphia City is excluded from this estimate.

The distribution in the Delaware valley   201 in the per capita geographic ranking in Table 11.2. The city of Philadelphia easily out distanced the other counties in per capita servant purchases. The per capita distribution of new servants appears to be generally determined by the distance from Philadelphia coupled with the scale of agricultural activity. The major agricultural counties adjacent to the city of Philadelphia, i.e., Chester, Philadelphia, Gloucester, Burlington, along with Lancaster counties, were second in per capita purchases. The outlying counties lagged far behind. This suggests that immigrant servitude may have become proportionally more important to the urban economy than to the rural economy by the American Revolution. Among the counties within Pennsylvania or within New Jersey there does not appear to be much trade off between reliance on slave labor and reliance on white servant labor. For example, comparing the adjacent counties of York and Cumberland Pennsylvania reveals that while York was closer to the Philadelphia market and had fewer slaves per capita they also purchased fewer servants per capita than did Cumberland county. However, if a broad comparison is made between the colonies of Pennsylvania, New Jersey, and Delaware a case for a trade off between slave and servant labor might be made. New Jersey and Delaware maintained relatively higher per capita slave populations and also possessed relatively lower per capita purchases of new immigrant servants compared to Pennsylvania. Finally, the evidence in Table 11.2 can be used to estimate the percentage of total immigrant servants in the colonial population. The 1771–3 record of servant sales spans two years of purchases. The average contract lasted for around four  years.5 If the flow of immigrant servants immediately prior to 1771 and after 1773 was roughly similar to the 1771–3 period, as seems to be the case,  then the percentage of immigrant servants in the population can be estimated by doubling the percentage reported in Table 11.2. This procedure indicates that immigrant servants comprised over 5 percent of Philadelphia city, 2 to 3 percent of Lancaster and Chester counties, over 1 percent of Burlington, Gloucester, and Philadelphia counties, and under 1 percent of the remaining counties analyzed. This suggest that by the eve of the American Revolution, among unfree laborers, immigrant servants were still in the majority in Philadelphia city and Lancaster county but in the minority in south-­central Pennsylvania, i.e., Cumberland, York, and Bedford counties, as well as in New Jersey and Dela­ware. This raises some questions about the supposedly decisive role continued servant immigration played in satisfying the demand for unfree labor in the middle colonies. The evidence in this study yields very different estimates of the total number of servants, as well as the proportion of slaves to servants, in Pennsylvania than that found in previous studies. Salinger (1981) estimates the number of servants in Pennsylvania from tax records and uses this evidence to assert that the population of bound labor declined in the late eighteenth century. Salinger’s estimates are considerably below the number of immigrant servants sold in the Philadelphia market as recorded in the direct auction evidence. For the 1771–3 period,

202   German immigrant servitude in America, 1745–1835 Table 11.3 The geographic distribution of Irish immigrant indentured servant purchasers by county of residence: Philadelphia market, 1745–6 Residence of purchaser

Total (N = 498)

Males (N = 432)

Females (N = 66)

Philadelphia City Chester Co., PA Philadelphia Co., PA a Lancaster Co., PA Bucks Co., PA Virginia and Maryland Burlington Co., NJ Gloucester Co., NJ Salem Co., NJ New Castle Co., DE Sommerset Co., NJ Hunterdon Co., NJ Middlesex Co., NJ Monmouth Co., NJ Sussex Co., NJ

23.69% 18.27 16.06 14.66 4.42 4.22 3.82 2.81 1.41 1.20 1.00 0.60 0.60 0.40 0.20

21.99% 17.82 15.97 15.05 4.17 4.86 3.70 3.01 1.62 1.39 1.16 0.46 0.46 0.46 0.23

34.85% 21.21 16.67 12.12 6.06 0.00 4.55 1.52 0.00 0.00 0.00 1.52 1.52 0.00 0.00

Total b

93.36%

92.35%

100.00%

Sources: See Table 11.1. Notes a This category excludes residents of Philadelphia City. b The absent category is where residency was not recorded.

comparing Salinger’s estimates to those presented in this study, Tables 11.4 and 11.5, indicates that Salinger underestimates the number of servants purchased by Philadelphians by 71 percent, the number purchased by Philadelphia county residents by 56 percent, the number purchased by Chester county residents by 68 percent, the number purchased by Lancaster county residents by 52 percent, and the number purchased by Northampton residents by 54 percent.6 This large and consistent underestimation suggests that tax records may suffer a significant underreporting bias.7 The obvious desire to avoid being taxed may have been easily accomplished with regards to servant property; servants being very mobile may have been easier to hide. Therefore, tax records should be used with extreme caution when measuring the magnitude of labor property. Nash (1973: 246) and Salinger (1981) estimate the ratio of slaves to servants to be approximately 1.8 for Philadelphia city and 1.1 for Philadelphia county in 1772–3. Their estimate of servants is based on tax records and so may be too low. Using their estimate of slaves and the estimate of immigrant servants presented in Tables 11.4 and 11.5 yields a ratio of slaves to servants which is almost the inverse of that reported by Nash and Salinger, i.e., 0.54 for Philadelphia city and 0.71 for Philadelphia county.8 Besides underreporting, there may exist a wealth bias in tax records that tends to inflate the ratio of slaves to servants. The wealthy owned more slaves relative to servants than the less wealthy (Tully 1967: 295). Since the probability of showing up in tax records is positively related to wealth, tax records may over-­count slaves relative to servants in the

The distribution in the Delaware valley   203 economy. Although immigrant servitude may not have been decisive in preventing a shift to slave labor, the evidence presented here suggests that immigrant servitude may have been more important to the bound labor market than has been previously thought. The structure of the geographic distribution of immigrant servants can be further analyzed by decomposing the evidence by ethnic origins, gender, family status, and contract type for the three samples. The first phenomenon revealed by these subdivisions is the rise in the dominance of the city of Philadelphia, from absorbing 24 percent of immigrant servants in 1745 to 45 percent by the post-­ Revolutionary period (see Tables 11.3 through 11.6). However, some of this rise must be attributed to the difference in servant contracts over time. Indentured immigrants had no say in the colonial auction of their person and so would have to go with whoever bid the highest. Redemptioner immigrants, on the other hand, negotiated their own sales and so had a choice of masters and hence some influence over their geographic location. Comparing British indentured immigrants between 1745 and 1771–3 reveals very little increase in the proportion purchased by Philadelphians. Whereas comparing British indentures to British redemptioner immigrants within the 1771–3 sample, Table 11.4, shows that redemptioners had a 42 percent greater chance of being purchased by a Philadelphian. This suggests that servants had a strong revealed preference for an urban environment. The above geographic preference revealed by British redemptioners also indicates a stronger desire to stay in nearby Philadelphia county while eschewing Lancaster and Cumberland counties relative to the behavior of their indentured counterparts. Some of the rising urbanization of immigrant servitude was therefore a result of the rising use of redemptioner contracts as the century progressed.10 This allowed servants more latitude in the choice of their location of work over time. But redemptioners also showed an increasing desire to stay in town environments over time. The share of German redemptioner immigrants purchased by Philadelphians shows a 27 percent rise from before to after the Revolution. The rising urbanization of servitude can also be seen in the proportion of immigrant servants bought by residents of towns other than Philadelphia, the most important being Lancaster, Northern Liberties, and Southwark. In 1745 only 4 percent were purchased by residents from non-­Philadelphia towns. Between 1771 and 1804, regardless of ethnic origin or contract type, 13 to 16 percent were purchased by residents from other towns (see Tables 11.7 through 11.9). Thus by the 1770s nonrural servitude was clearly dominant among British redemptioners comprising 58 percent of their purchases. It also became clearly dominant among German redemptioners soon after the Revolution comprising 59 percent of their purchases. In all cases Philadelphians purchased a larger share of single immigrant female servants than of their respective male counter-­parts. They absorbed a third of the female supply of British indentured immigrants in the 1745 and 1771–3 samples and 45 percent of the female supply of redemptioners in the  1771–3 market. For redemptioners this share jumped to 65 percent in the 9

Total indentures (N = 846)

27.19% 23.17 19.15 7.45 6.74 2.48 2.48 1.65 1.65 1.54 1.18 0.95 0.71 0.47 0.35 0.35

Residence of purchaser

Philadelphia City Lancaster Co., PA Chester Co., PA Cumberland Co., PA Philadelphia Co., PAa Bucks Co., PA Burlington Co., NJ Delaware Colony Virginia Colony Gloucester Co., NJ Salem Co., NJ Northumberland Co., PA Maryland Colony Hunterdon Co., NJ Bedford Co., PA Morris Co., NJ

24.07% 21.26 21.26 8.64 5.42 3.39 3.39 1.53 1.86 2.03 1.53 0.85 1.02 0.34 0.51 0.51

Male indentures (N = 590) 23.87% 28.23 14.92 4.03 10.08 0.40 0.40 2.02 1.21 0.40 0.40 1.21 0.00 0.81 0.00 0.00

Female indentures (N = 248) 42.05% 8.59 14.02 4.55 10.35 2.27 3.54 2.53 0.76 2.15 1.26 0.88 1.01 0.51 0.00 0.00

Total redemptioner (N = 792) 42.20% 9.32 13.90 4.58 9.32 2.71 2.54 3.39 1.02 2.20 1.02 0.51 1.36 0.51 0.00 0.00

Male redemptioner (N = 590)

44.77% 7.56 13.37 5.23 13.37 1.16 4.07 0.00 0.00 2.33 2.33 1.16 0.00 0.58 0.00 0.00

Female redemptioner (N = 172)

Table 11.4 The geographic distribution of British immigrant indentured and redemptioner servant purchasers by county of residence: the Philadelphia market, 1771–3

98.71%

Total b 98.56%

0.17 0.34 0.34 0.34 0.00 0.17 0.00 0.00 0.00 0.00 0.00 98.52%

0.40 0.00 0.00 0.00 0.40 0.00 0.00 0.00 0.00 0.00 0.00

Notes The absent category in the subdivisions of status is being a married servant, i.e., 38 individuals. a This category excludes residents of Philadelphia City. b The absent category is where the residence was not recorded.

Source: See Table 11.1.

0.24 0.24 0.24 0.24 0.12 0.12 0.00 0.00 0.00 0.00 0.00

Sussex Co., NJ New York Colony York Co., PA Cumberland Co., NJ Berks Co., PA Kent Co. Westmoreland Co., PA North Carolina Colony South Carolina Colony Cape May Co., NJ Willis Co., PA 97.52%

0.13 0.38 1.64 0.00 0.13 0.00 0.25 0.13 0.13 0.13 0.13 97.13%

0.17 0.00 1.36 0.00 0.17 0.00 0.34 0.17 0.00 0.17 0.17

97.67%

0.00 0.58 0.58 0.00 0.00 0.00 0.00 0.00 0.58 0.00 0.00

32.77% 15.89 13.93 7.14 5.80 4.55 4.11 3.39 2.41 1.88 1.16 1.07 0.71 0.54 0.45 0.45 0.36 0.27 0.18 0.18 0.18 0.09 0.09 0.09

97.69%

Philadelphia City Philadelphia Co., PAa Lancaster Co., PA Berks Co., AP Chester Co., PA Burlington Co., NJ Gloucester Co., NJ Bucks Co., PA Northampton Co., PA Maryland Colony Salem Co., NJ Delaware Colony York Co., PA Bedford Co., PA Hunterdon Co., NJ Middlesex Co., NJ New York Colony Sussex Co., NJ Virginia Colony Cumberland Co., NJ Sycell Co., PA North Carolina Colony Morris Co., NJ Cape May Co., NJ

Totalsb

Notes a This category excludes residents of Philadelphia City. b The absent category is where the residence was not recorded.

Source: See Table 11.1.

Total redemptioners (N = 1,120)

Residence of purchaser

97.41%

30.69% 16.31 14.16 6.44 7.51 4.29 3.86 4.51 3.00 0.43 1.07 2.15 0.43 0.43 0.21 0.64 0.43 0.64 0.00 0.21 0.00 0.00 0.00 0.00

Single male above age 15 (N = 466)

98.50%

31.97% 14.13 12.64 6.69 7.06 7.06 5.58 2.97 3.35 1.12 0.74 0.37 1.12 0.00 0.74 0.37 0.37 0.37 0.74 0.00 0.00 0.37 0.37 0.37

Males below age 15 (N = 269)

97.42%

43.91% 18.08 15.87 3.69 2.21 2.95 4.06 2.58 0.74 1.48 0.37 0.37 0.00 0.00 0.00 0.37 0.37 0.00 0.00 0.37 0.00 0.00 0.00 0.00

Single females (N = 271)

96.49%

15.79% 14.04 10.53 19.30 3.51 3.51 1.75 1.75 3.51 10.53 5.51 0.00 1.75 3.51 1.75 0.00 0.00 0.00 0.00 0.00 1.75 0.00 0.00 0.00

Married persons (N = 114)

Table 11.5 The geographic distribution of German immigrant redemptioner servant purchasers by county of residence: Philadelphia market, 1771–3

Total servants (N = 3,129)

44.87% 11.82 7.61 4.99 4.83 3.04 2.78 2.68 2.27 1.47 1.44 1.21 0.99 0.70 0.64 0.51 0.48 0.45 0.42 0.42 0.42 0.29 0.29

Residence of purchaser

Philadelphia City Philadelphia Co., PAa Lancaster Co., PA Gloucester Co., NJ Montgomery Co., PA Berks Co., PA Chester Co., PA Burlington Co., NJ Delaware Co., PA Bucks Co., PA Northampton Co., PA Northumberland Co., PA Daulphin Co., PA Georgia State Maryland State Cumberland Co., NJ Hunterdon Co., NJ Westmoreland Co., PA Delaware State Virginia State Florida State Franklin Co., PA Salem Co., NJ

40.06% 14.03 9.77 3.84 5.30 3.35 3.63 3.35 1.88 1.54 1.61 0.77 1.26 0.28 0.42 0.77 0.84 0.28 0.28 0.84 0.42 0.35 0.28

Single adult male (N = 1,436) 65.83% 11.91 3.13 4.08 1.57 1.57 2.51 1.88 1.25 0.00 0.00 0.63 0.00 0.00 0.31 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00

Single adult female (N = 319) 28.11% 9.43 6.79 6.04 8.11 4.91 2.26 2.64 3.02 3.40 1.89 3.02 2.26 3.02 1.51 0.38 0.38 1.51 0.75 0.00 1.13 0.38 0.38

Married persons (N = 530) 50.35% 8.78 6.70 8.08 4.16 2.31 2.31 1.85 2.77 0.92 1.62 1.62 0.23 0.46 0.69 0.23 0.00 0.46 0.69 0.23 0.00 0.23 0.46

Male children (N = 433)

continued

64.27% 10.53 5.54 4.71 1.66 1.39 1.11 1.66 2.49 0.28 0.55 0.55 0.00 0.00 0.55 0.28 0.00 0.00 0.28 0.00 0.00 0.00 0.28

Female children (N = 361)

Table 11.6 The geographic distribution of German immigrant redemptioner servant purchasers by county of residence: Philadelphia market, 1787–1804

97.04%

0.00 0.28 0.21 0.00 0.00 0.14 0.07 0.07 0.14 0.07 0.14 0.21 0.14 0.00 0.14 0.14 0.00 0.00

Single adult male (N = 1,436)

94.06%

0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.31 0.31

Single adult female (N = 319)

Notes The absent family status subdivision was being a single parent, i.e., 50 servants. a This category excludes residents from Philadelphia City. b The absent category is where residence was not recorded.

Source: See Table 11.1.

96.72%

0.26 0.19 0.16 0.16 0.13 0.13 0.13 0.10 0.10 0.10 0.10 0.10 0.10 0.06 0.06 0.06 0.03 0.03

Lycomy Co., PA Cumberland Co., PA Sussex Co., NJ Ohio State Washington Co., PA Kentucky State Massachusetts State Luzerne Co., PA New York State New Hampshire State Centre Co., PA. York Co., PA Mifflin Co., PA Sommerset Co., NJ Adams Co., PA Fayette Co., PA North Carolina State South Carolina State

Totalsb

Total servants (N = 3,129)

Residence of purchaser

Table 11.6 continued

95.10%

0.75 0.00 0.38 0.75 0.38 0.38 0.38 0.38 0.00 0.00 0.00 0.00 0.00 0.38 0.00 0.00 0.00 0.00

Married persons (N = 530)

96.76%

0.46 0.23 0.00 0.00 0.46 0.00 0.23 0.00 0.00 0.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Male children (N = 433)

97.80%

0.55 0.00 0.00 0.28 0.00 0.00 0.00 0.00 0.00 0.28 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Female children (N = 361)

38.51%

42.46%

50.76%

33.87% 1.61 1.61 2.02 1.21 0.81 9.63 58.76%

42.05% 0.00 1.01 3.54 0.13 0.25 11.78 56.02%

42.20% 0.00 1.36 3.22 0.17 0.34 8.73

65.69%

44.77% 0.00 0.00 4.07 0.00 0.00 16.85

Notes a This includes 78 other towns. b The absent category is buyers who did not list their residence as a town. The absent category across the status classes is being a married servant.

Source: See Table 11.1.

Totalsb

24.07% 1.86 1.02 0.85 1.02 0.85 8.84

Philadelphia City 27.19% Nottingham, PA 1.77 Northern Liberties, PA 1.18 Southwark, PA 1.18 Alexandria, VA 1.06 West Nottingham, PA 0.83 Other townsa 9.25

Residence of purchaser Total indentures Male indentures Female indentures Total redemptioners Male redemptioners Female redemptioners (N = 846) (N = 590) (N = 248) (N = 792) (N = 590) (N = 172)

Table 11.7 The geographic distribution of British indentured and redemptioner servant purchasers by town of residence: the Philadelphia market, 1771–3

43.25%

45.68%

Totalsb

Notes a This includes 41 other towns. b The absent category are those who did not report their residence as a town.

Source: See Table 11.1.

30.69% 2.79 2.79 0.43 0.86 0.00 0.21 5.48

33.77% 2.32 1.96 0.98 0.63 0.54 0.54 5.94

Philadelphia City Lancaster Northern Liberties, PA Southwark, PA Germantown, PA Frederick Town, MD Oley, PA Other townsa

Single male above age 15 (N = 466)

Total redemptioners (N = 1,120)

Residence of purchaser

42.72%

31.97% 1.97 0.37 1.12 0.74 0.00 0.37 6.18

Males below age 15 (N = 269)

58.31%

43.91% 2.95 2.95 2.21 0.74 0.74 0.00 4.81

Single females (N = 271)

35.07%

15.79% 1.75 1.75 0.00 0.00 3.51 3.51 8.76

Married persons (N = 114)

Table 11.8 The geographic distribution of German immigrant redemptioner servant purchasers by town of residence: Philadelphia market, 1771–3

56.20%

58.74%

Totalsb

80.22%

65.83% 7.52 1.25 0.63 0.94 0.63 0.31 3.11

Single adult female (N = 319)

39.18%

28.11% 3.02 3.13 1.13 0.75 0.00 0.38 2.66

Married persons (N = 530)

62.57%

50.35% 3.23 3.46 0.46 1.62 0.69 0.46 2.30

Male children (N = 433)

Notes a This includes 47 other towns. b The absent category were those buyers who did not list a town as their residence. The absent status category is being a single parent.

Source: See Table 11.1.

40.06% 7.26 2.44 1.61 0.70 0.28 0.07 3.78

44.87% 5.85 2.40 1.21 0.80 0.38 0.26 2.97

Philadelphia City Northern Liberties, PA Lancaster, PA Southwark, PA Germantown, PA Reading, PA Frederick Town, MD Other townsa

Single adult male (N = 1,436)

Total servants (N = 3,129)

Residence of purchaser

76.60%

64.27% 6.09 2.49 1.11 0.28 0.69 0.55 1.12

Female children (N = 361)

Table 11.9 The geographic distribution of German immigrant redemptioner servant purchasers by town of residence: Philadelphia market, 1787–1804

212   German immigrant servitude in America, 1745–1835 post-­Revolutionary period. If the purchases from all towns are considered, female servitude by the 1770s, regardless of ethnic origin or contract type, was a majority urban experience. After the Revolution, some 80 percent of immigrant female servants went to nonrural owners. This represents a 38 percent growth in the share of female servants going to urban residents from before the Revolution. Only among German immigrant redemptioners can the geographic distribution be decomposed by family status, i.e., adult versus children and single versus married. For the 1771–3 sample males below age 15 appear to have approximately the same geographic distribution as males above age 15 (Table 11.5). Similarly, in the 1787–1804 sample (Table 11.6), female children have approximately the same geographic distribution as single female adults. However, in this latter sample male children have a 25 percent greater chance of staying in Philadelphia than single male adult servants. This may reflect a relatively increasing demand for apprentice labor in Philadelphia city over time. Married couples afford a sharp contrast to all other groups. The share purchased by Philadelphians was 16 percent in 1771–3 and rose to only 28 percent in the 1787–1804 sample. When all towns are considered in the post-­ Revolutionary period over 56 percent of single adult males and over 80 percent of single adult females were purchased by urban residents, whereas only 39 percent of married persons realized the same fate. Hence, only among married servants did rural demand remain dominant throughout the eighteenth century. Married servants remained relatively overrepresented in the outlying counties and colonies. For example, the share of married servants going to Bedford, Berks, and Westmoreland counties in Pennsylvania and to Maryland, Georgia, and Florida was substantially larger than the share of single adults or children. In the case of distant colonies or states it is clear that offers of cheap land as recorded in the auction evidence played an important role in attracting married servants to these places. For the 1771–3 sample ethnic comparisons of the geographic distribution between British and German immigrant servants can also be observed. Although British, who were mostly Irish, and German servants went to all areas, they appear to be attracted to regions proportionally dominated by their own ethnic group, as determined by Lemon (1972: 50) and Diffenderffer (1899: 306). This was particularly evident among servants who had some latitude in selecting their own masters, i.e., who came under redemption agreements. Thus German redemptioners tended to favor buyers from predominantly German Lancaster, Berks, York, and Northampton counties. Whereas British redemptioners tended to favor buyers from predominantly British Chester, Cumberland, and Northumberland counties as well as from the city of Philadelphia. This trend is weaker when the comparison is made with British indentured servants, most notably with Lancaster residents purchasing a significant portion of this group. Since indentured immigrants had no influence over who purchased them, this suggests that servants’ desires for a master from a region of similar ethnic background may have been stronger than employers’ desires for servants of similar ethnic background.

The distribution in the Delaware valley   213 Only for the 1787–1804 sample of German immigrant servant purchases were occupations of the buyers recorded (Table 11.10). Among the 3,126 servants purchased, buyers listed 120 different occupations or trades. Farmers and merchants were by far the most frequent buyers with 21 and 18 percent of the purchases, respectively.11 The occupational distribution supports the geographic evidence in showing that by the post-­Revolutionary era servitude had become decidedly nonagricultural. This trend can be further illustrated in Table 11.11, which compares the estimated representation of the top 20 occupations in terms of magnitude employed in the Pennsylvania economy in 1800 with the percentage of immigrant servants purchased by that category between 1787 and 1804. Farmers as a group are clearly underrepresented among servant employers. The occupations which are substantially overrepresented among servant employers are merchants and shopkeepers, gentlemen and esquires, and innkeepers and tavernkeepers. This suggests that the part of the economy proportionally more dependent on immigrant servant labor by this period was the service sector. Salinger (1981) finds a somewhat similar result when looking only at servant employers in the city of Philadelphia. The subdivision of purchasers by family status reveals that farmers purchased proportionally more married persons and single adult males than single females or children. Many of the skilled trades, e.g., miller, baker, tailor, tanner, iron master, and druggist, also show a slightly higher preference for single adult males and married servants over females and children. In marked contrast, such service occupations as merchant, innkeeper, storekeeper, grocer, and tobacconist purchased proportionally more single women and children. This general pattern seems to reflect the difference in the nature of work that might be required of the servant. Women and children possessed a relative productivity advantage in less skilled, less strenuous, customer-­oriented occupations. By contrast, single adult males and married males possessed a relative productivity advantage in skilled trades and field labor. The demand for domestic service may also explain the proportional preference by mariners for single females.

Conclusion Servitude in the first century of American colonization was a rural, agricultural institution. It provided the labor force to work the fields of the New World estates. Through the second half of the eighteenth century, immigrant servitude evolved into an urban, service sector institution, dominated by employment in Philadelphia and surrounding towns. Much of this evolution must be attributed to the revealed preference of servants. As servitude adopted the redemptioner contract form, servants had more latitude in their choice of masters and thereby tended to opt for urban employment. However, there was also some secular shift in the economy towards using servants in an urban, service sector setting by the end of the century, independent of contract type, sex, age, or ethnic background. Married persons tended to go to frontier farms while single women and children

Total redemptioners (N = 3,129)

20.84% 18.06 5.18 3.36 3.04 2.72 2.27 2.14 1.89 1.62 1.57 1.47 1.31 1.31 0.99 0.96 0.89 0.86

Occupation or title of purchaser

Farmer Merchant Innkeeper Gentleman Miller Esquire Baker Tailor Store or shop keeper Mariner Tanner Cordwainer Grocer Iron master Tobacconist Blacksmith Glazier Drugist

22.89% 15.35 6.07 2.23 3.77 2.65 3.21 3.56 1.47 0.70 2.23 1.88 0.91 1.33 0.21 1.12 0.28 1.19

Single adult male (N = 1,436) 9.09% 21.32 7.21 3.45 1.88 1.88 1.88 2.51 2.51 8.46 0.94 1.57 2.51 0.63 1.57 1.25 0.63 0.31

Single adult female (N = 319) 31.32% 15.85 1.89 7.55 3.40 3.02 0.00 0.00 1.13 0.75 0.38 0.75 0.38 3.40 1.51 0.00 0.38 0.38

Married persons (N = 530) 18.32% 21.25 3.00 2.31 2.54 3.70 2.54 0.23 1.85 0.46 1.62 0.23 1.62 0.46 3.00 0.69 1.62 0.92

Male children (N = 433)

11.08% 25.76 6.65 2.77 1.11 2.22 1.66 1.39 4.16 1.94 0.83 1.94 2.49 0.00 0.28 1.94 1.39 0.83

Female children (N = 361)

Table 11.10 The occupational distribution of German immigrant redemptioner servant purchasers: Philadelphia market, 1787–1804

90.86%

88.24%

Totalsb

89.59%

0.31 0.94 0.94 0.63 0.31 0.63 0.31 1.88 0.94 1.25 0.00 0.00 1.25 10.29

Notes a This includes 94 occupations which purchased less than 14 servants each. b The absent category was where no occupation of the purchaser was recorded.

Source: See Table 11.1.

0.84 1.05 0.91 0.49 0.77 0.42 0.91 0.49 0.28 0.28 0.70 0.56 0.07 11.69

0.86 0.83 0.74 0.58 0.54 0.51 0.48 0.48 0.45 0.45 0.45 0.45 0.45 10.20

Paper maker Brewer Victular Attorney at law Distiller Doctor Butcher House carpenter Cooper Dealer Joiner Sugar refiner Widow Other tradesa 83.06%

0.75 0.38 0.75 0.38 0.00 0.38 0.00 0.00 0.00 0.00 0.38 0.00 0.00 7.57 83.56%

1.39 0.23 0.00 1.39 0.92 0.92 0.23 0.00 0.69 0.69 0.00 0.46 0.46 9.43

88.12%

0.55 1.39 0.83 0.28 0.55 0.55 0.00 0.55 1.11 0.83 0.55 1.11 1.66 9.72

216   German immigrant servitude in America, 1745–1835 Table 11.11 Comparison of the percentage distributions of Pennsylvanian occupations with those of servant purchasers Top occupations

Pennsylvania occupations based on the 1800 censusa

Immigrant servants purchased, 1787–1804 b

Farmer or yeoman Laborer Carpenter Tailor or hatter Weaver Blacksmith Merchant or shopkeeper Widow Miller Cooper Ropemaker Gentleman or esquire Mason or bricklayer Innkeeper or tavernkeeper Shoemaker or cordwainer Metalworker Brewer or distiller Doctor or dentist Wagon-maker or wheelright

41.98% 8.87 4.44 4.13 3.71 3.41 3.01 2.87 2.41 1.59 1.59 1.51 1.46 1.40 1.38 1.08 0.92 0.81 0.75

20.84% 0.00 0.48 2.50 0.03 0.96 19.95 0.45 3.04 0.45 0.22 6.08 0.03 5.44 1.47 1.85 1.37 0.51 0.22

Total

87.32%

65.89%

Notes a Derived from Soltow and Keller (1982). b Derived from Table 11.10.

tended to work in Philadelphia. Servants also showed preference to work in areas dominated by employers of their own ethnic heritage. On the eve of the American Revolution, immigrant servants represented over 5 percent of the population of the city of Philadelphia, 2 to 3 percent of Lancaster and Chester counties, and a little over 1 percent of Burlington, Gloucester, and Philadelphia counties. Among unfree laborers in the 1770s, immigrant servants were still in the majority in Lancaster county and Philadelphia city but in a minority in New Jersey, Delaware, and in south-­central Pennsylvania. Despite the increasing urbanization of immigrant servitude, these institutions still effectively spread a large portion of the immigrating population among the established population. Enclaves or ghettos of new immigrants did not form in the same manner that they would in the nineteenth century after the ending of immigrant servitude when newcomers were free to choose their residence and occupation.

Notes   * Originally published as Farley Grubb, “Immigrant servant labor: their occupational and geographic distribution in the late eighteenth-­century mid-­Atlantic economy,” Social Science History, vol. 9, no. 3 (Summer 1985): 249–75. Copyright 1985.

The distribution in the Delaware valley   217 © 1985 by the Social Science History Association. All rights reserved. Reprinted with permission of the publisher, Duke University Press. The author is Assistant Professor in the Economics Department at the University of Delaware, Newark, de 19716. He would like to thank Stanley Engerman and David Galenson for helpful comments on earlier drafts. Social Science History 9:3 (Summer 1985). Copyright © 1985 by the Social Science History Association. ccc 0145–5532/85/$1.50.   1 For contracts that were immediately resold or reassigned the same day to a new owner at the same price as the original purchase, some 10 percent of the contracts, the residence and occupation of the final master was the one counted. This same day reassigning was largely a way of recording absentee buying or purchases made by an agent of a distant master. There would appear to be little reason to suspect reporting biases in this evidence since it was not used for any taxing purposes and provided legal safeguards in terms of contract enforcement for both parties, i.e., proof of ownership if the servant tries to run away and proof of contract termination date if the master tries not to release the servant.   2 Roughly 28 percent of new servant contracts in the Philadelphia market were formed by non-­immigrants. Most of these were apprentices but occasionally adults would enter servitude in order to secure cash advances or to service past debts. Tax and probate records do not make a distinction as to whether a servant was an immigrant or not or of what nationality he was.   3 Payments of freedom dues in land would be cheaper for a master living on the frontier than if he lived in the more settled areas. In the 1787–1804 sample 42 servants were offered landownership as part of their freedom dues and in almost all cases this involved movement to the frontier in Pennsylvania or to a distant state.   4 This is the best sample to use since the 1745 sample experienced an abnormally low number of German immigrants and there are no population estimates for that period. The 1787–1804 sample only counts immigrant German servants and so is not comprehensive in its coverage of new immigrant servants. The 1771–3 period also appears to capture a relatively normal period of servant immigration for the Germans and the Irish who largely comprised the servant population.   5 The mode contract length was four years for adults, regardless of ethnic origin, sex, or contract type. However, for children under about age 14 contract length rose at about a one-­to-one rate as age fell. Among the British there were virtually no children but among the Germans there was a sizable number, e.g., in the 1785–1804 sample 25 percent of the German servants were dependent children (Grubb 1985). Therefore, estimating the stock of servants at any point in time by taking a two-­year flow and multiplying it by two will understate the true stock of servants. Thus the estimate reported here might be a lower bound.   6 Salinger’s underestimation is probably even larger than suggested since her evidence also includes non-­immigrant servants, see note 2. Salinger’s evidence is in terms of stocks at a point in time and my evidence is in terms of a two-­year flow. To compare the two estimates Salinger’s estimate was converted to a two-­year flow by dividing what she reports for a given year by half. Since children served more than four years this procedure also results in understating the divergence between my estimates and Salinger’s estimates. In many cases taxes were not assessed on non-­adult servants and so these servants tend not to appear in tax records.   7 Other possibilities that might account for this discrepancy would be servants effectively escaping their contract obligations after landing or being quickly resold out of the colony. Both of these possibilities seem unlikely. The magnitude of running away needed to explain the divergence in estimates would be so large that there would be little reason to purchase a servant in the first place. Contemporary accounts of the trade say little about these phenomena thus their occurrence was probably minor.   8 Nash and Salinger’s estimates are in stock terms while my estimates are in terms of a

218   German immigrant servitude in America, 1745–1835 two year flow with regards to servants. Assuming servants worked on average four years, the flows can be converted to stock terms by multiplying by two. Since Nash and Salinger’s estimates include non-­immigrant servants and since children servants worked over four years, the difference between the two estimates is probably larger than reported, see note 5.   9 In Chester and Lancaster counties the top 30 percent in terms of probated wealth held from 85 to 93 percent of the slaves but only 65 to 69 percent of the servants among those listed in the probate records (Tully 1967: 295). Tully’s evidence spans 29 and 19 year intervals, respectively, and so estimating a ratio of slaves to servants is difficult. However, assuming that servants worked on average four years and slaves worked on average 20 years will yield a ratio of slaves years to servant years of 1.5 for Chester county from 1729–58 and 1.9 for Lancaster county from 1739–58. These numbers are high by the estimates reported here for the 1770s. Since the 1750s witnessed a very large influx of German servants, these estimates derived from Tully’s probate inventory evidence may be artificially high. Probate records have the usual wealth bias thus underreporting the poorer part of the population who were also the more likely owners of servants rather than slaves. 10 Among Irish immigrant servants only 6.7 percent came as redemptioners in 1745 whereas 34.4 percent came as redemptioners in 1771–3. German immigrant servants use the redemptioner form exclusively beginning in the 1720s. 11 This is the same percentage of farmers as what existed among the immigrant Germans in a sample of 579 immigrants who reported their pre-­immigration occupation for the period 1793–1807, among adult males.

References Alexander, J. K. (1974) “The Philadelphia Numbers Game: An Analysis of Philadelphia’s Eighteenth Century Population.” Pennsylvania Magazine of History and Biography 97 (July): 314–24. City Archives of Pennsylvania “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771 to October 5, 1773.” Diffenderffer, F. R. (1899) “The German Immigration into Pennsylvania Through the Port of Philadelphia, and ‘The Redemptioners’.” pt. 7 of “The German Influence in its Settlement and Development.” Pennsylvania German Society 10: 1–328. Dunn, R. S. (1973) Sugar and Slaves. New York: W. W. Norton. Galenson, D. W. (1981a) White Servitude in Colonial America. Cambridge: Cambridge University Press. Galenson, D. W. (1981b) “White Servitude and the Growth of Black Slavery in Colonial America.” Journal of Economic History 41 (March): 39–47. Grubb, F. (1985) “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804.” Explorations in Economic History (forthcoming). Lemon, J. T. (1972) The Best Poor Man’s Country. Baltimore: The Johns Hopkins Press. Main, G. L. (1977) “Maryland and the Chesapeake Economy, 1670–1720,” in A. C. Land, L. G. Carr, and E. C. Papenfuse (eds.) Law, Society, and Politics in Early Maryland. Baltimore: The Johns Hopkins Press: 134–52. Menard, R. R. (1977) “From Servants to Slaves: The Transformation of the Chesapeake Labor System.” Southern Studies 16 (Winter): 355–90. Mittelberger, G. (1960) Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754. O. Handlin and J. Clive (trans. and eds.) Cambridge: Harvard University Press.

The distribution in the Delaware valley   219 Morgan, E. S. (1975) American Slavery, American Freedom. New York: W. W. Norton. Nash, G. B. (1973) “Slaves and Slaveowners in Colonial Philadelphia.” William and Mary Quarterly 3rd Series 30 (April): 233–56. Neible, G. W. [ed.] (1906–8) “Servants and Apprentices Bound and Assigned Before James Hamilton Mayor of Philadelphia, 1745.” Pennsylvania Magazine of History and Biography: 30–2. Pennsylvania Historical Society “Book of Redemptioners, 1785–1804.” Philadelphia. Salinger, S. V. (1981) “Colonial Labor in Transition: The Decline of Indentured Servitude in Late Eighteenth-­Century Philadelphia.” Labor History 22 (Spring): 165–91. Salinger, S. V. (1983) “Artisans, Journeymen, and the Transformation of Labor in Late Eighteenth-­Century Philadelphia.” William and Mary Quarterly 3rd Series 40 (January): 63–84. Soltow, L. and Keller, K. W. (1982) “Rural Pennsylvania in 1800: A Portrait From the Septenial Census.” Pennsylvania History 49 (January): 24–47. Smith, B. G. (1977) “Death and Life in a Colonial Immigrant City: A Demographic Ana­ lysis of Philadelphia.” Journal of Economic History 36 (December): 863–89. Sutherland, S. H. (1936) Population Distribution in Colonial America. New York: Columbia University Press. Tully, A. (1967) “Patterns of Slaveholding in Colonial Pennsylvania: Chester and Lancaster Counties 1729–1758.” Journal of Social History 6 (April): 67–78. Walsh, L. S. (1977) “Servitude and Opportunity in Charles County, Maryland, 1658–1705,” in A. C. Land, L. G. Carr, and E. C. Papenfuse (eds.) Law, Society, and Politics in Early Maryland. Baltimore: The Johns Hopkins Press: 111–333. Wokeck, M. (1981) “The Flow and Composition of German Immigration to Philadelphia, 1727–1775.” Pennsylvania Magazine of History and Biography 105 (July): 249–78. Wood, P. H. (1974) Black Majority. New York: W. W. Norton.

12 Determining the method of entering servitude and modeling its performance*

Historians have identified two methods of acquiring transatlantic passage through servitude and have named them “indenture” and “redemption.” Contemporaries, however, used these names rarely. Instead, contemporaries often used the term “indenture” in its generic form, which simply meant “a contract” (Tomlins 2010: 32). They did not use this term to denote the immigrant contracting method that today we call “indentured servitude.” The two methods are distinguished by whether the immigrant entered the servant labor contract at embarkation (the indenture method) or at debarkation (the redemption method). Having entered servant contracts pre-­voyage, indentured immigrants upon arrival in America were called “servants . . . whose times are to be disposed of by the captain”. Indentured servants were sold or assigned to their first American master directly by the ship captain or his agent at debarkation. By contrast, because they had not yet signed servant labor contracts, redemptioner immigrants, upon arrival in America, were called “passengers” or “freights” who were “willing to serve a reasonable time for their passage (freight) money.”1 Redemptioner servants sold themselves into servitude at debarkation. Once a servant labor contract was entered into in America, redemptioners were thereafter also called “servants.” Distinguishing between the two methods is important because the method chosen affected the division of contracting risk between shippers and servants and the degree of flexibility in contracting between immigrant servants and American masters. For indentured servants, having fixed their servant contract lengths and other labor conditions pre-­voyage, any unexpected changes in the value of their contracts between embarkation in Europe and debarkation in America, say due to unexpected changes in American labor market conditions or unexpected changes in the health and physical prowess of the servants during the voyage, were absorbed by the shipper. Indentured servants could not be forced by their shipper to serve longer contracts once the ship had sailed—they had guaranteed fixed contracts. As such, shippers bore all the transportation risks under the indenture method. They were the residual claimants for any unexpected changes in the value of their servant cargos between embarkation and first sale in America (Galenson 1981: 13–15, 97–113; Grubb 1985: 855–68). The tradeoff for servants was that with fixed servant contracts to be sold by shippers

The method of entering servitude   221 in America, servants had no choice over who their masters would be and no ability after observing the servant market at arrival to negotiate any special or opportunistic conditions into their servant contracts. By contrast, redemptioner immigrants agreed to fixed debts with shippers at embarkation. These debts equaled their passage costs and other migration expenses that had been transferred or sold to the shipper. At debarkation, redemptioner immigrants negotiated their own servant contracts, including the contract’s length, directly with prospective American masters. The one requirement was that the contracts had to be long enough and valuable enough for American buyers to pay the shippers the redemptioners’ contracted passage debts. Under the redemption method, the shippers possessed guaranteed passage payments, and immigrants were the residual claimants for any unexpected changes in their value in the American servant market due to unexpected changes in American market conditions or in the immigrants’ health and physical prowess between embarkation in Europe and debarkation in America. Immigrants bore all the transportation risks under the redemption method. The tradeoff for these redemptioner immigrants was that, in exchange for foregoing the insurance of a fixed guaranteed servant contract at embarkation, they had at debarkation the ability to select their own American masters among those attending the servant auction. They could also negotiate special and opportunistic conditions into their servant contracts on the spot at debarkation, taking advantage of current revelations in the American servant market at the time of their landing. Once the servant labor contract was signed and sold in America the distinction between these two methods of entering servitude, indenture or redemption, ceased to matter. Servant law in America did not distinguish between and applied equally to both methods of entering servant labor contracts (Galenson 1981: 13–15; Grubb 2000a: 42–75). .

The tradeoff immigrant servants made between having contract flexibility at debarkation under the redemption method versus having a guaranteed fixed contract at embarkation under the indenture method can be illustrated by how immigrant occupational skills were valued in the transatlantic servant market. Indentured immigrants had an incentive to lie about their skills in the hope of fixing shorter contracts pre-­voyage. Under the indenture system, it was difficult for American masters to verify or discern the truthfulness of the skills claimed for a given servant pre-­contract. For example, Maryland planter William Colling lamented that his convict indentured servant John Tandy “pretends to be a wheelwright and sawyer but is master of neither” (Maryland Gazette 29 March 1770). Under the indenture system servants could not credibly reveal their skills pre-­ contract to their future American masters and so could not bargain for shorter contracts based on the true value of those skills in America. Shippers would have considered it unwise and too risky to offer shorter contracts at embarkation to servants who claimed to have skills. Shippers could not realize the higher prices that servant skills might command in the market in America unless the shippers could guarantee those skills to the America buyers, which most could not do.2

222   German immigrant servitude in America, 1745–1835 In addition, at embarkation it was difficult for shippers to determine which skills would be in demand or even whether there would be buyers for a particular skill at debarkation, given the lag in information crossing the Atlantic and the variability in voyage time (Chapter 2). Thus, under the indenture method, for the servant to take a guaranteed fixed contract at embarkation and push all the voyage and market risk onto the shipper, meant sacrificing any potential gain the servant might be able to get for his particular occupational skill. Empirically, this was the case for adult English indentured servants. Little relationship between specific skill classifications and servant contract lengths could be found in a sample of 1,790 adult indentured servants embarking between 1682 and 1759 for America (Grubb 1992: 186–92). The redemption method was a boon both to skilled servants and to American masters. Redemption gave American masters the chance to discern the skill of a servant firsthand before contracting and the chance to write skill-­contingent contracts. It allowed servants to credibly bargain for shorter contracts as compensation for their skills. For example, Adam Righart [Johann Adam Reichert], a German redemptioner who arrived in Philadelphia on the ship Union from Rotterdam 27 September 1773, signed a servant contract with Jacob Winey on 4 October 1773 for three years. This contract was a year shorter than that typically signed by adult German male redemptioners in this period. Winey paid 21.17.0 Pennsylvania pounds for Righart. In addition, the contract was registered with the following contingency clause, “Adam Righart shall be employed at the milling business in case he does not understand the business the said servant is to have the privilege to pay the said master the sum of 21.17.0 and expenses occurring this indenture to be void” (“Record of indentures;” Strassburger 1934: vol. 1, ship# 309). Voiding the contract because Righart did not understand the milling business would most likely require Righart to reenter the redemption auction. Righart would have to negotiate a new, and most likely longer, servant contract with a new American master to raise the sum to repay his old master, Jacob Winey. He owed Winey not only the 21.17.0 Pennsylvania pounds initially paid to clear Righart’s passage debt, but also an additional amount to repay Winey any post-­voyage expenses incurred on Righart’s behalf, such as for food and clothes. Under redemption, the immigrant acquired a master-­search option. Daniel Kent was a cutler and arrived in Pennsylvania from Ireland as a redemptioner in 1784 (using the hybrid “indenture of redemption” method discussed below). The redemption option in his contract gave him the chance to search for a master who was a cutler. In his particular case the search was unsuccessful and his contract reverted to the indenture option and he was sold by his shipper to a farmer. He related in a letter to his parents, “My business is of little use here. Blacksmiths, carpenters, shoemakers, bricklayers and tailors are the chief trades here” (Barnard 1904: 17). His statement reveals the tradeoffs made by shippers and servants in choosing redemption versus indenture. Chapters 13 and 14 systematically document these tradeoffs. For example, Chapter 14 shows that purchasers of German redemptioner servants in

The method of entering servitude   223 Philadelphia between 1787 and 1804 who reported that they (the purchasers) were skilled tradesmen, gave the servants they purchased 5 to 11 percent shorter contracts than did purchasers who were farmers, other things equal. This outcome is consist with redemptioners having the flexibility that indentured servants did not have to take advantage of market conditions revealed on the spot in America at the time of their debarkation to better match their particular skills to the needs of American buyers. German immigrant servants appear to have used the redemption method exclusively, whereas British and Irish immigrant servants predominantly, but not exclusively, used the indenture method. Given the differences in risk and contract flexibility between the indenture and redemption methods, determining which methods was used by different ethnic groups is important. Doing so, however, requires some detective work. Contemporaries did not often use the modern terminology of “indenture” and “redemption” when referring to immigrant servants. As such, close attention to the language they did use, eyewitness descriptions of the contracting process, and reliance on a number of different sources is required to establish this pattern.3 Surviving immigrant servant contracts and the registrations of these contracts by government officials provide some evidence as to whether immigrant servants used the indenture or redemption method. Few servant contracts have survived. The few surviving German immigrant servant contracts indicate that Germans used the redemption method. For example, the contracts of Mary Elizabeth Bauer and of Anna Margareta Sahlin, who entered servitude in Philadelphia in 1767 and 1785, respectively, were printed in Philadelphia. As such, they were unlikely to have been entered into pre-­embarkation. Their contracts are reproduced in Figures 12.1 and 12.2.4 The payments for the contracts were 27 and 20 pounds [Pennsylvania currency] “for her passage from Holland” and “her freight from Amsterdam,” respectively. Thus, even though they are using forms printed in Philadelphia, Bauer and Sahlin cannot be non-­immigrant resident servants. The ship captains were not direct parties to these servant contracts, and their names are not mentioned on the contracts. The ship captains or their agents (such as Stewart & Nesbit mentioned on Sahlin’s contract) were paid the 27 and 20 pounds by Bauer’s and Sahlin’s new America masters, Samuel Pleasants and Samuel Coats, respectively. These American masters were explicitly mentioned by name in the contracts. The payments they made cleared Bauer’s and Sahlin’s passage debts from Holland owed to their respective ship captains. Bauer’s and Sahlin’s contracts contrast sharply with the contracts used in the indenture and the “indenture of redemption” methods where the servant was initially contracted to the ship’s captain with the ship captain’s name mentioned directly on the servant contract and where the American buyer’s name (the master to whom the servant was sold to in America) does not appear directly on the servant contract (see Figure 12.3).5 The transcriptions (contract registrations) of the salient features of servant contracts and contract sales in Philadelphia have survived for only a few select

224   German immigrant servitude in America, 1745–1835

Figure 12.1 Immigrant servant contract entered into by Mary Elizabeth Bauer, 1767 (source: original contract reproduced with permission from The Historical Society of Pennsylvania, Philadelphia (society miscellaneous collection, box 9c, white servitude folder #9, indentures of apprentices). See also Klepp, Grubb, and Pfaelzer de Ortiz (2006: 16); Salinger (1987: xiv)).

years (Chapter 10). These contract registers also do not make it obvious which method, indenture or redemption, was used by a given immigrant servant whose sale was registered therein. For example, the document titles given to these contract registers were created in a later period by officials and archivists labeling materials as they saw fit. These document titles were not necessarily used by contemporaries at the time. The surviving document that registered immigrant servant sales from October 3, 1771 to October 5, 1773, the single most important surviving quantitative source of servant auction sales pre-­ revolution, appears to have been given its title in 1835 (“Record of indentures”).6 Thus, the document’s title does not necessarily establish that a particular method was in use. In this 1771–3 contract register, individual entries made by officials seldom use the words “redemption” or “indenture.” Where these words appear in the register’s entries, they typically appear together in the phrase under an indenture of redemption to [ship captain’s name] now cancelled”. This phrasing was used only on a portion of British and Irish immigrant servant sale entries and never on

The method of entering servitude   225

Figure 12.2 Immigrant servant contract entered into by Anna Margareta Sahlin, 1785 (source: original contract reproduced with permission from The Historical Society of Pennsylvania, Philadelphia (society miscellaneous collection, box 9c, white servitude folder #9, indentures of apprentices)).

German immigrant servant sale entries. It denoted a new contracting method that appeared in the early 1770s in the British and Irish immigrant servant market. This new contracting method, the “indenture of redemption,” was a hybrid contract combining elements both of the indenture and the redemption methods. The immigrant signed a fixed-­length servant contract at embarkation and so

Figure 12.3 Immigrant servant contract entered into by Daniel Kent, 1784 (source: Barnard (1904: 14–15)).

The method of entering servitude   227 looked initially like an immigrant servant using the indenture method. This indenture, however, could be “cancelled,” as opposed to “assigned” or “sold” by the shipper, at the immigrant’s discretion at debarkation if certain conditions were met. If “assigned” or “sold,” then it was the same as the indenture method. At debarkation, however, immigrants could have their indentures “cancelled” if they caused a sum, stipulated in their contracts, to be paid to their shippers. Immigrants could do this by proceeding as if using the redemption method, namely negotiating new servant contracts themselves directly with prospective American employers at debarkation, who would then pay the sums required to cancel the existing indentured servant contracts. This hybrid contracting method gave immigrants the insurance of fixed-­length indentured servant contracts before sailing but the post-­voyage bargaining flexibility of the redemption method. An actual servant contract, one of the few that have survived, consistent with this new hybrid “indenture of redemption” contracting method is for Daniel Kent who arrived in Pennsylvania from Ireland in 1784 (Barnard 1904). Kent’s contract, reproduced in Figure 12.3, shows that he was initially indentured to his ship’s captain, John Johnston. The first half of the contract follows the standard indenture method of servant contracting.7 The second paragraph, however, creates an “indenture of redemption” hybrid contract. This paragraph is a buy-­ out clause priced at 10.5 pounds sterling plus necessities payable within 20 days after arrival in America. This clause provided an option that the servant could exercise to void the initial servant contract if he could find a better deal within 20 days of landing in America. Kent failed to exercise this option and his initial three-­year contract was sold by his ship captain to a farmer in Chester County, Pennsylvania, just like that for other servants using the indenture method (Barnard 1904). Kent’s contract contrasts sharply with the contracts used by German immigrant servants (compare Figure 12.3 with Figures 12.1 and 12.2). The language in the 1771–3 contract register that indentified this hybrid method, namely “under an indenture of redemption to [ship captain’s name] now cancelled” only appears in some entries for British and Irish immigrant servants, and never in entries for German immigrant servants. Most individual servant entries in the 1771–3 contract register, as in all the other Philadelphia contract registers (Chapter 10), recorded neither the word “indenture” nor “redemption.” Identifying who used the indenture and who used the redemption method can be done by paying close attention to the language used by ship captains, servant advertisers, and contract registrars. Language, because it was imbued with legal content, was important and precisely used by these individuals. This can be seen in the language used by the officials recording the entries in the 1771–3 contract register. Indentured immigrants were already under a servant contract upon arrival, thus they could not be sold without being legally “assigned” by their old master (the ship captain) to their new (American) master. The word “assigned” was a legal term denoting the transfer or sale of an existing contract. Without the word “assigned,” the contract cannot be for an indentured immigrant.

228   German immigrant servitude in America, 1745–1835 The registration language “In consideration of [amount] paid for his/her passage” coupled with the absence of language saying “assigned by [ship captain’s name] to . . .” identified the use of the redemption method. This categorization scheme, coupled with linking servant entries to European ports of embarkation or to an arriving ship passenger manifests (in order to separate immigrant from local resident servant contracts), indicates that German immigrant servants exclusively used the redemption method, whereas the British and Irish immigrant servants exclusively used the indenture method or the new hybrid “indenture of redemption” method. This distinction between redemption and indenture methods can also be seen in the language used by shippers when advertising servant cargos for sale upon their arrival in America. Advertisements for prospective German servants placed by shippers in the Pennsylvania Gazette and the Maryland Gazette between 1730 and 1776 referred to “passengers” or “freights” who “are willing to serve a reasonable time for their passage money”—language consistent only with the use of the redemption method. These advertisements never spoke of “servants.” By contrast, every advertisement for British and for Irish immigrant servants placed by shippers never called them “passengers” or “freight” but only “servants” or “indentured servants”—language consistent only with the use of the indenture method. Precise language was so important that when Captain George Dempster of the brig Morning Star mistakenly advertised his German cargo in the newspaper as “servants,” he corrected the advertisement in the next issue by changing the word “servants” to “freights.”8 Surviving accounts written by German immigrant servants and surviving German immigrant passenger contracts both indicate an exclusive use of the redemption method by German immigrant servants. For example, Johann Carl Büttner, who arrived in Philadelphia on the ship Sally from Rotterdam on 23 August 1773 and was sold there as a servant on 26 August 1773, described the method used for his ship as being that of redemption (Klepp, Grubb, and Pfaelzer de Ortiz 2006: 177–84, 214–15, 221–2). The passenger account book, ship contract, and newspaper advertisement for the ship Britannia, which arrived in Philadelphia from Rotterdam on 18 September 1773, indicates that Germans were using the redemption method exclusively. The surviving passenger contracts for the ships Favorite and Elizabeth, which arrived in Philadelphia from Amsterdam in 1803 and 1819, respectively, also indicates that German immigrants used the redemption method exclusively.9 Finally, the few eyewitness descriptions of the sale of German immigrant servants and the advice given in letters sent back to Germany by contemporary German-­Americans are consistent with German immigrant servants using the redemption method exclusively. For example, Pastor Henry Melchoir Muhlenberg, head of the Lutheran Church in Pennsylvania, described the servitude method used by German immigrants in a letter sent to Germany in 1769 as being that of redemption. He indicated that German servant contracts were negotiated after the ships had arrival in America, the chief distinguishing characteristic of the redemption method.10

The method of entering servitude   229 In 1724, one year after migrating to Pennsylvania, Christopher Sauer sent a letter to friends in Germany specifically instructing them not to indenture themselves to the ship captain but to use what amounted to the redemption method. A similar description and recommendation can be found in a letter sent from Pennsylvania to Germany in 1728.11 Ludwig Gall, a German immigrant whose ship arrived in Philadelphia in 1819, described the method used by prospective German immigrant servants on his ship as that of redemption (Trautmann 1981: 35–65). The earliest date that the redemption method can be identified in the German transatlantic immigrant trade is 1724 (Durnbaugh 1960: 231–3). Advertisements for the sale of German immigrant servants in Philadelphia placed in the Amer­ ican Weekly Mercury indicate that the redemption system had not been adopted by German immigrants in 1722. The language in these advertisements more closely resembles that mentioned above for British and Irish indentured immigrants than for German immigrant servants after 1724. Therefore, the redemption system appears to have begun in the transatlantic German passenger trade sometime between 1722 and 1724. The descriptions across all the sources for the method used by German immigrant servants after 1724 are more-­or-less identical whether taken from the 1720s, 1750s, 1760s, 1770s, or 1810s. Apparently not much changed after 1724 regarding the method used by German immigrant servants from the mid-­1720s into the 1820s. After 1724, German immigrants used the redemption method exclusively. The origins of the redemption method, why Germans appear to have used that method exclusively after 1724, and why British and Irish immigrant servants only partially adopted that method and then only in a hybrid version (the “indenture of redemption”) from the 1770s onward, is still puzzling (Abbott 1924: 10). The answers provided here, while somewhat speculative, are the best that can be mustered after 30 years of researching this topic. Being able to negotiate at debarkation directly with perspective American employers under the redemption method may have been more important to German immigrants than to British and Irish immigrants because of the relatively high proportion of families among German immigrants (Chapters 5 and 10). The redemption method allowed German parents the flexibility to observe the servant market in America and decide after arrival how best to redistribute the family’s total passage debt among its individual members via servant contracts—something that could not be done under the indenture method. In particular, redemption allowed, at debarkation, some children potentially to take over the passage debts of a parent or of other children in the family and so serve in their place. It also allowed a parent potentially to take over the passage debts of some children, saving these children from servitude (Chapters 9 and 15). The redemption method also allowed parents, because they directly negotiated with prospective American masters at debarkation, more input and oversight regarding the servant contracts entered into by their children. For example, some parents were able to negotiate specific tailor-­made education clauses for their children (Chapters 8, 14, and 15). Finally, redemption allowed servants

230   German immigrant servitude in America, 1745–1835 opportunistically to choose specialty contract clauses, such as job-­specific work restrictions, that would have been too costly to negotiate on a contingency basis before embarkation (Chapter 13). All of these benefits were unavailable under the indenture method. Over half of German immigrant servants, however, did not migrate in family groups nor did they negotiate any specialty clauses into their contracts (Chapters 9, 13, and 14). They appear to have assumed the redemption risk of not receiving a guaranteed fixed servant contract length at embarkation for no discernible benefits. They seemingly did not intend to use the redemption method’s contract flexibility at debarkation, but nevertheless took on the risk potentially of having to enter a relatively longer contract at debarkation if the market in America just happened unexpectedly to be weak, or if they happened unexpectedly to be less healthy, during the days immediately after their landing in America. Why these German immigrants did not opt for the indentured method, with its low-­risk guaranteed fixed contract length at embarkation, is a puzzle. The answer appears to lie in the fact that most German immigrants to America, including single unskilled adults, incurred substantial pre-­embarkation migration debts. The cost of migrating to and subsisting in Dutch seaports before booking passage overseas was high, and many German migrants had to resort to borrowing money from recruiters, hostel-­keepers, and middlemen to cover these costs. This was something that their British and Irish counterparts typically did not experience. The Dutch holders of these pre-­embarkation debts wanted them liquidated before the immigrant sailed. For Dutch holders of these debts to collect them from the emigrants once the emigrants had successful debarked in America was too costly to execute and enforce. Because of this, shippers purchased (liquidated) these debts from their Dutch holders and added them to the debts owed by the emigrants for their passage fares. Because they had substantial pre-­embarkation debts to pay, many German redemptioners had to sell themselves into servitude in America for sums well in excess of the passage fare. The size of these pre-­embarkation debts varied substantially across German immigrants. They could range from zero to as much as 35 percent, and even in a few cases as high as 50 percent or more, of the initial ship-­passage fare added onto that fare (Chapter 15). The sale and liquidation of pre-­embarkation recruitment and migration debts incurred by prospective overseas migrants in the Netherlands was a long-­ standing feature of the labor recruitment process used by the Dutch East India Company (Verenigde Nederlandse Geoctroyeerde Oostindische Compaigne or VOC). The VOC had the largest recruitment operation for overseas labor in the Netherlands with extensive recruitment efforts reaching deep into Germany. The number of Germans recruited by the VOC to serve the company on the high seas and in the far east could often exceed the number of Germans migrating to America. The likelihood of running into a recruiter for the VOC in Germany was far greater than running into a recruiter for emigrant shippers headed to America. VOC recruiters were often independent agents working on commission who expected reimbursement plus a markup for fronting recruitment expenses.

The method of entering servitude   231 The extensive recruitment network that the VOC had in place that channeled Germans into Holland and then to VOC employment also served, directly or by example, to facilitate German migration to the New World. These labor recruitment networks often fronted the cost of food, lodging, and transportation. The networks consisted of a series of independent entrepreneurs. One agent might recruit Germans in Hamburg and pay to have them transported by ship to Amsterdam. Another agent might then house and feed the recruits while they waited employment interviews with the VOC. At each stage of the recruitment process, as one independent agent passed their recruits off to the next agent, each agent expected to be paid or compensated for the costs they had incurred. Thus, an immigrant so recruited might accumulate debts that would be sold from agent to agent until it was finally liquidated by VOC employment. The recruit, as a VOC employee, would have his wages garnished, or a sum deducted periodically, by the VOC. The VOC would pay that sum to the last holder of the employee’s recruitment debt in Holland, or would pay that sum to itself if it had purchased the debt outright.12 The redemption method used by German immigrants to America may have simply been an adaptation of the VOC recruitment debt-­repayment process, namely it was a method that allowed the consolidation and transfer of debts owed to different agents within the labor recruitment process by a given emigrant to some final debt holder. The various recruiters, hostel-­keepers, and middlemen that functioned as part of the VOC recruitment network, or others following their example, were willing to advance German emigrants bound for America the cost of transportation, food, lodging, and so on for the emigrants’ journeys from their homes in Germany to Dutch ports and for their stays in Dutch ports while waiting for ships bound for their desired destinations in America. The difference between the liquidation of VOC recruitment debts and the liquidation of German emigrant-­to-America pre-­voyage debts was that German emigrant-­to-America debts had to be sold outright to a shipper at the point of embarkation in Europe. There was no comparable agent in Holland, such as the VOC, to garnish wages from the emigrant-­to-America recruit and deliver it to the emigrant’s debt holder in Holland. The shipper would roll the emigrant’s pre-­voyage debts that were transferred to him into the debt the shipper charged to cover the fare for passage. Each emigrant was then required by the shipper to clear their entire debt at debarkation in America—most likely through selling themselves into servitude there. Regardless of the origins of the redemption method, the fact that passage debts (as opposed to passage fares) varied more for German immigrants than for British and Irish immigrants made using the redemption method, as opposed to the indentured method, preferable for German immigrants. Signing fixed-­length servant contracts before embarkation designed to sell for an expected sum in America, as British immigrants did under the indentured method, would have created what economists call a “lemons problem” for German immigrants (Akerlof 1970: 488–500). Because of the variable pre-­embarkation debts added to the cost of passage, German immigrant servant contract lengths could vary both because of differences in servant labor productivity and because of

232   German immigrant servitude in America, 1745–1835 differences in passage debts. If contract lengths were fixed before embarkation, buyers in America would not know which reason caused a given servant’s contract length to be shorter or longer. Risk-­adverse buyers would interpret longer contracts as a sign of low servant labor productivity (the servant was a “lemon”) rather than as a sign of a higher debt and so offer too low a price per contract length. The redemption method solved this “lemons problem” by simultaneously revealing the debt and the physical condition of the redemptioner to buyers in American when negotiating servant contract lengths. Figure 12.4 presents a basic market model of the transatlantic German redemptioner servant trade. The model is used to illustrate the expected relationships between servant contract prices, contract lengths, labor productivity, passage debts, and shipper profits earned on servant cargo. The model assumes competition among shippers in the transatlantic passenger trade from Holland to America (Chapters 2 and 3) and competition among buyers of German immigrant servants in Philadelphia (Chapters 11, 14, and 18). The model starts with the passenger contracting market in a European port, such as Rotterdam, for shipping human freight to America—the right-­hand panel in Figure 12.4. The demand by potential emigrants to secure passage to America is assumed to be a normal downward-­sloping demand curve where lower passage fares, other things equal, increase the quantity of emigrants and vice versa. Short-­term shifts in the demand curve are dominated by warfare on the migration route. Wars shift demand left. Other events could also produce short-­term shifts in the demand curve, such as the 1816–17 “year without a summer” which shifted the demand curve temporarily right (Chapters 2 and 18). Shipping contract market (Europe)

Contract price (CP)

High productivity labor

D*

CP* � D* Low productivity labor 1

2 3 4 5 6 7 8 Contract length in years (T)

Passage fare � passage debt (D)

Servant–Labor contracting market (America)

Supply

Demand Q* Quantity of human freight

Figure 12.4  The basic market model of adult German redemptioner servitude.

The method of entering servitude   233 The supply of passenger shipping services to America is assumed to be perfectly elastic (a flat supply curve). Bulky and heavy (low-­value-to-­weight) agricultural goods were shipped from America to Europe. Lighter and smaller (high-­value-to-­weight) fabricated goods were shipped from Europe to America. With trade having to be balanced value-­wise, this condition led to excess capacity on the westward leg of the transatlantic shipping circuit. In addition, most of the westbound cargo was non-­human freight. Human freight could always displace non-­human freight at a constant cost, attracting more ships into carrying passengers and increasing the passenger-­per-ton ratio, as long as passengers paid a fare that yielded a similar return to carrying non-­human cargo (Chapters 2 and 13). Both of these conditions imply that the supply of passenger shipping services should be strongly elastic, and that the risk-­adjusted passage fare should be equal to the opportunity cost of the freight space used (the value of the next best cargo that could be shipped). Thus, within the range typically observed, more emigrant cargo, i.e. shifts in the demand curve, should not bid up passage fares and vice versa. Short-­term shifts in the supply curve were dominated by warfare on the migration route. Wars shifted supply up, raising the passage cost (Chapters 2 and 3). The cost of protection via organized shipping convoys, the risks of being captured by privateers, and the increased uncertainty of the state of demand in the destination market all increased shipping costs during war. The leftward shift in demand coupled with the upward shift in supply during war combined to drastically reduce wartime immigration to America (Chapter 2). The shipping contract market for human freight in Europe, the right-­hand panel in Figure 12.4, determined the passage fare or passage debt (D*) that had to be paid to the shipper to secure passage to America. This passage fare was the same for all adults because it was based on the opportunity cost of the freight space use (the value of the next best cargo that could occupy the same freight space adjusted for the risk of cargo loss). The freight spaced used was the same across adults. Thus, it did not matter whether an adult was male, female, married, tall, short, rich, poor, skilled, unskilled, strong, weak, and so on. All faced the same passage fare. Children below age 5 paid no passage fare because they were given no freight space. They were expected to occupy their parent’s berth and eat from their parent’s food and water ration. Children between ages 5 and 12 paid half an adult passage fare. They were given half the adult freight space, i.e. half the berth space and half the food and water ration as an adult (Chapters 15 and 16). If the passage debt was to be paid at debarkation in America versus paid in cash at embarkation in Europe, an 11 to 15 percent markup was added to compensate the shipper for the risk of passenger default and the non-­use of the cash (foregone interest) between embarkation and debarkation (Chapters 13 and 16). In the absence of additional pre-­embarkation debts, this risk adjusted passage fare is the debt (D*) that adult redemptioner immigrants, by shipping contract, had to pay at debarkation in America. The left-­hand panel in Figure 12.4 shows the servant labor contracting market for redemptioners in America, those who would sell themselves into servitude to clear their passage debts. The passage debt (D*) had to be paid before

234   German immigrant servitude in America, 1745–1835 debarkation. How this debt was paid did not matter to the shipper. If relatives in America showed up at the docks and paid the immigrants’ passage debts, the shipper would release the redemptioners from the ship as if they were a free passenger. However, most redemptioners had to sell themselves into servitude to raise the required sums (Chapters 14–16). For these redemptioners, the contract prices paid by their American employers for their servant contracts had to equal this passage debt (CP* = D*). Redemptioners were heterogeneous in labor productivity. Some were tall, some short, some strong, some weak, some skilled, some unskilled, some male, some female, some educated, some illiterate, some young, some old, some experienced workmen, some inexperienced workmen, and so on. The amount of work needed to raise a given sum of money (CP* = D*) varied inversely with the redemptioner’s perceived labor productivity in America. The more productive the servant, the less work time was needed to generate a given sum of money. The fan of linear arrows in the left-­hand panel of Figure 12.4 represents redemptioner servants with different labor-­productivity profiles, from a low labor-­productivity profile on the right to a high labor-­productivity profile on the left. For a given servant, the longer the contract the higher the price paid for that contract. For a given contract length, a high-­productive servant will sell for a higher contract price than a low-­productive servant. American buyers would board the ships when they docked in Philadelphia and bargain with redemptioners over the number of years, months, and days of work (the length of the servant contract) that would be required for the buyers to pay the redemptioners’ passage debts, i.e. the T needed so that CP* = D* (Chapters 14 and 16). Competition among American buyers led to an equilibrium T that was just long enough for each servant, given that servant’s labor-­ productivity profile, to generate a payment just equal to CP* = D*. For adult redemptioners who used servitude to clear their debts and where passage fares equaled these passage debts, all would sell for the same price but would likely have different contract lengths. In the German redemptioner trade, some adult passengers had incurred substantial pre-­embarkation debts that were sold to the shipper and added to the base passage fare that these redemptioners owed the shipper at debarkation. Thus, (CP* = D*) could vary across adult redemptioners. Per given labor-­ productivity profile (per given arrow in the fan of arrows in the left-­hand panel in Figure 12.4), a higher overall migration debt required an equally higher contract price to clear that debt, which in turn required the redemptioner to accept a longer servant contract to raise that higher sum, other things equal. Conversely, on rare occasions some adult redemptioners partially pre-­paid their passage fares and owed less than the base passage fare at debarkation. In these cases, per given labor-­productivity profile, redemptioners could negotiate shorter servant contracts, other things equal. For children under age 12 who were charged less than adult passage fares, their lower productivity profiles meant longer servant contracts, other things equal, but their lower passage fares attenu-

The method of entering servitude   235 ated this effect somewhat. These children, however, faced other parental and legal debt-­shifting constraints in the market that affected their servant contracts. These constraints are discussed in Chapter 15. The model in Figure 12.4 yields several predictions about the expected relationships between adult servant contract prices (CP), contract lengths (T), labor productivity, passage debts (D), and shipper profits earned on adult servant cargo. These predictions are empirically tested in the following Chapters 13–18. The model establishes that in the German redemptioner servant auction in America, T should be the dependent variable and CP and servant fixed effects (proxies for servant labor productivity) should be independent variables (Chapter 14). The four key predictions of the model are: 1

2

T should vary inversely with labor productivity, other things equal. In other words, per given passage debt (CP = D), more productive servants should receive shorter contracts and less productive servants should receive longer contracts. Because expected labor productivity can only be measured by observable proxies related to human-­capital, this prediction implies that, per given passage debt, servants who were younger, less educated, less skilled, female, and so on should receive relatively longer contracts, other things equal. The differences in servant contract lengths measure the relative value  of these different aspects of labor productivity. In addition, servants should have to pay for any desired specialty contract clauses, such as work-­restriction clauses or master-­provided educate clauses that reduced the servant’s labor time or productivity, by agreeing to longer servant contracts, other things equal. The extra contract length measures the relative value of these specialty contract clauses. These predictions are tested in Chapter 14. CP should be largely independent of labor productivity. The variation in contract prices should simply measure the variation in passage debts incurred across redemptioners, i.e. the variation in passage fares and additional pre-­embarkation debts added to those fares. For example, the CP for German redemptioners who embarked from London were lower than for German redemptioners who embarked from Rotterdam. This simply reflected the fact that passage fares (and sailing times) were lower for sailing from London to Philadelphia compared with sailing from Rotterdam to Philadelphia. Whether additional pre-­embarkation migration debts were related to labor productivity is unclear. Young adults, who were less productive, had little ability to generate savings and so would have had to borrow more to cover food and housing in port pre-­embarkation. However, older adults, who were more productive, typically had dependents accompanying them and so they also had to borrow more for food and housing in port pre-­ embarkation.13 As such, how labor productivity affected pre-­embarkation debts is unclear. They are treated here as unrelated. This prediction is tested in Chapter 14, footnote 25.

236   German immigrant servitude in America, 1745–1835 3 4

Controlling for labor productivity, CP and T should be positively related. The higher a redemptioner’s passage debt the longer should be their servant contract length, other things equal. This prediction is tested in Chapter 14. No excess profits should be earned by shippers on servant cargos compared with what shippers could earn on comparable cargos. Shippers basically earned their risk-­adjusted freight cost D on servants which should be no more or no less than what they could earn on comparable non-­servant cargo. This prediction is tested in Chapter 13.

Figure 12.5 provides an enhanced version of the basic market model of German immigrant redemptioner servitude presented in Figure 12.4. It adds some new predictions regarding the performance of the German redemptioner servant market, particularly regarding its structural limits and the dynamic process involved in the market’s disappearance. It also incorporates some additional empirical observations taken from the market. The enhanced model is the same as the basic model with the exception that the fan of labor-­productivity profiles in the left-­hand panel is no longer assumed to be a family of straight lines. In the basic model, Figure 12.4, each labor-­ productivity profile is assumed to be proportional in CP and T, i.e. each profile has a constant CP/T ratio throughout. In other words, for each productivity profile, a 10 percent increase in passage debt (D), which is the same as a 10 percent increase in contract price (CP), yields a 10 percent increase in contract length (T), no matter what the initial starting value is for CP. By contrast, the enhanced model assumes an S-­shaped curve for each labor-­productivity profile. Servant labor in the initial years of the contract is less valuable than that in the middle or latter years of the contract. Redemptioners with Shipping contract market (Europe)

Contract price (CP)

High productivity labor

D*

CP* � D*

Low productivity labor

1

2 3 4 5 6 7 8 Contract length in years (T)

Passage fare � passage debt (D)

Servant–Labor contracting market (America)

Supply

Demand Q* Quantity of human freight

Figure 12.5  The enhanced market model of adult German redemptioner servitude.

The method of entering servitude   237 unusually high passage debts, however, are taken to be suspicious individuals and so extra-­risky bets for repayment. Such redemptioners required disproportionally more years added to their contracts to compensate for this extra risk. Several observations are consistent with, and indicate the presence of, this S-­shaped pattern for a servant’s labor-­productivity profile. First, around the neighborhood of the market’s equilibrium, where the fan of labor-­productivity profiles crosses the CP* = D* line, the estimated relationship between CP and T is not proportional, other things equal. For German adult male redemptioners landing in Philadelphia in 1771–3, a 10 percent increase in CP produces a 6 percent increase in T, other things equal (Chapter 14). For the sample of German redemptioners landing in Philadelphia in 1787–1804 where age could be explicitly controlled, a 10 percent increase in CP produces only a 1 percent increase in T, other things equal (Chapter 14). At equilibrium, the CP/T ratio is steeper than a linear ratio that starts from the origin. This result, slightly exaggerated, is depicted in Figure 12.5. Second, servant contracts involved significant fixed cost elements, namely contracting and employment transaction costs, as well as mortality, morbidity, acculturation, and acclimatization costs of adapting to the work environment in America for Germans during their initial year after landing. Given the sizable fixed-­cost element of servitude, contract lengths could not decline proportionate to the fall in passage fares as passage fares approached zero. Several factors indicate that the first few years of servitude were relatively less valuable and were heavily discounted by American masters. German immigrants suffered from a higher rate of debilitating illness and death in their first year in America compared with that experienced thereafter. This seasoning process was described and documented in Chapter 4.14 In addition, servants had to acculturate and acclimatize to their new American surrounding and work requirements, which may have been particularly trying for the many German servants purchased by non-­ German Americans (Chapter 11). This could easily take a year or two to achieve, costing labor productivity in the meantime.15 Finally, servants who tried to escape their American masters by running away overwhelmingly attempted to do so within the first two years of their servant contract. Thereafter, the attempted-­runaway rate dropped off substantially (Grubb 1999: 198; 2000a: 49; 2000b: 108–13). These observations establish and explain the initial half of the S-­shaped pattern in the servant labor-­productivity profiles in Figure 12.5. Establishing and explaining the upper portion of the S-­shaped pattern in the servant labor productivity profiles shown in Figure 12.5 is more speculative. At equilibrium, where the fan of labor-­productivity profiles crosses the CP* = D* line, a 10 percent increase in CP led to a mere 1 percent increase in T. If that rate held thereafter at higher levels of CP, then we should expect some redemptioners to take on the passage debts of other redemptioners in exchange for some side payments. In addition, we should expect one family member to take on the passage debts of all other family members. At the equilibrium tradeoff-­rate of T for CP, it always paid to consolidate passage debts across individuals and have the few serve for the many. Redemptioners as a group would always come out ahead by consolidating debts across individuals. Total servitude time for a group

238   German immigrant servitude in America, 1745–1835 of adults that consolidated debts would be less than if each proceeded individually to serve and pay for their own individual passage debts. This advantageous outcome, however, was never observed in the German redemptioner servant market. Some constraint prevented redemptioners from taking advantage of the equilibrium T for CP tradeoff when CP rose above equilibrium levels. The suggestion here is that the advantageous T for CP tradeoff at equilibrium quickly eroded as CP got unusually high. In other words, the second portion of a servant’s S-­shaped labor-­productivity profile kicked in when CP rose above equilibrium levels. Because of this, no debt consolidations across adult redemptioners that would double, triple, or more the CP for a single adult were observed in the market. The explosion in T that would accompany such a consolidation made it not worth doing. In the market, total servitude time for a group of adults would be less if they all proceeded individually to serve and pay for their own individual passage debts than if they tried to consolidate debts into the hands of a few who would serve to pay the debts of the many. The reason for this outcome was likely the risk aversion of American employers and the “lemons problem” they faced in the German redemptioner servant auction market (Akerlof 1970: 488–500). American employers would be suspicious of an adult redemptioner who showed up in the servant auction with three or four, or more, times the usual single adult passage debt. The cause of the higher debt would be unclear. Perhaps it was caused by that redemptioner’s poor, profligate, or even criminal behavior pre-­emigration, which in turn signaled to potential employers that the servant possessed a questionable work ethic (i.e. the servant was a “lemon”). Alternatively, the mere act of taking on so many debts of other redemptioners signaled to employers that this person may have some plan in hand to run off and thereby default on this servant contract. The unusually large contract payment gave the servant a magnified incentive to run away. In either case, American employers would require an extraordinary lengthening of the servant contract to compensate for these risks and suspicions. Finally, buying one servant for such a high price for a less than proportional increase in contract length was risky given the mortality and morbidity conditions in early America. Taking on the risk of putting all one’s eggs in one basket required some compensation, such as an extra-­proportional lengthening of the servant contract when CP got unusually high. The enhanced model in Figure 12.5 is useful for illustrating dynamic changes in the market. Over time, the demand by American employers for German immigrant servants rose (German immigrant servant labor productivity increased) and the supply of German immigrants offering themselves as servants decreased (Chapter 18). The combined increase in demand and decrease in supply pushed the value of German immigrant servant labor up, i.e. pushed CP/T up. In Figures 12.4 and 12.5, this effect is illustrated by a leftward rotation, anchored at the origin, in the fan of labor-­productivity profiles, such that each labor-­productivity profile crosses the CP* = D* line to the left of where it did previously. In the basic model, Figure 12.4, this process would produce a continuous decline in servant contract lengths per CP as the market approached its end. The

The method of entering servitude   239 model predicts that adult servant contract lengths should fall from an average of say four years, to three, to two, to one, to one-­half, to one-­quarter year, and so on down to perhaps only a month or less, before the market completely disappeared. The basic model also predicts that there might be some servants who were so productive that they would only have to enter a half-­year or a quarter-­year servant contract to pay their passage debts. These two predictions, however, fail to be true empirically. Almost no servant contracts under two years are entered into by German redemptioner immigrants. From 1772 to 1820, when the market effectively disappeared, average German adult male immigrant servant contract lengths fell from four years to only two years. Two-­year contracts appear to be a threshold below which the market could not go (Chapters 17 and 18; Grubb 1992: 210–25). By contrast, the enhanced model in Figure 12.5 accounts for these observations. The S-­shaped labor-­productivity profiles include an initial year or so of very low labor productivity based on factors related to health, environment, and cultural adaptation that did not change with time. It also takes into account the sizable fixed-­cost elements of servant contracting. The increases in demand for German immigrant servants by American employers coupled with decreases in the supply of German immigrant servants that occurred over time, which in turn pushed up servant labor productivity and reduced CP/T, ran into the constraint of an unchanging low immigrant labor productivity in the first year of the servant contract and the fixed costs of entering that contract in the first place. In Figure 12.5 this effect is illustrated by the fact that as the labor productivity profiles in the left-­hand panel, which are anchored at the origin, rotate to the left over time, they will pile up on the last ‘high productivity labor’ profile drawn. With the first year or so of servitude being typically useless to the master, even the most productive and highly valued servant had to serve at least a two-­ year contract to make it worthwhile for the American employer to pay the servant’s passage debt. Immigrant servant contracts under two years would not be observed. The disappearance of the market for adult European immigrant servants would not entail a continuous decline in servant contract lengths to zero, but a market collapse when the contract length was pushed to just under a threshold length of two years. The enhanced model in Figure 12.5 also solves a similar puzzle regarding the observed nonuse of immigrant servitude later in the nineteenth century in the era of mass German immigration to America. After 1830, transatlantic passage fares fell substantially which, in part, opened the floodgates to mass migration (Chapters 18 and 19; Hatton and Williamson 2005: 35–45). In both the basic and enhanced models, Figures 12.4 and 12.5, this amounts to a substantial downward shift in the supply of shipping services in the right-­hand panel of the model. This, in turn, substantially lowered the equilibrium CP* = D* line in the left-­ hand panel of the model where servant contract lengths are determined. Even at these low passage fares there would still be some emigrants who did not have the resources to pay their migration costs upfront and who would be willing to use the redemption method to finance these migration costs. They

240   German immigrant servitude in America, 1745–1835 would be willing to sell themselves as servants in America to clear their migration debts. The basic model, Figure 12.4, because of its linear labor-­productivity profiles and proportional declines in T with declines in CP, predicts that such should be observed. Yet, an active market for European immigrant servant contracts in America is not observed after 1820 (Chapters 17 and 18). The era of mass migration of Europeans to America from 1835 to 1914 is also the era of free passengers. The enhanced model in Figure 12.5 solves this puzzle by showing that as passage fares fall substantially below their pre-­1830 levels and the equilibrium CP* = D* line moves closer to zero, the T per CP tradeoff at these low CP levels becomes very unattractive. Given the sizable fixed-­cost element of servitude, contract lengths could not decline proportionate to the fall in passage fares. A couple years of servitude were still required to liquidate very minor passage debts. Under such conditions, it would make more sense for a potential emigrant to save enough income pre-­emigration to purchase the passage fare with cash upfront at embarkation and then debark in America as a free worker (Chapter 19). European immigrant servitude in America as a method for financing transatlantic passage was simply not economically viable when passage fares became relatively low. This market model of German redemptioner servitude, especially the enhanced version in Figure 12.5, forms the basis for the research that follows in Chapters 13–19. Therein, the predictions and empirical elements of the market, as viewed through the lens of this model, are systematically worked out, estimated, tested, and investigated in detail.

Notes   * Adapted and updated from Farley Grubb, “Labor, markets, and opportunity: indentured servitude in early America, a rejoinder to Salinger,” Labor History, vol. 39, no. 2 (May 1998): 235–41, copyright 1998, © 1998 by Taylor & Francis Publishers and the editors of Labor History, used with permission; appendix B of Farley Grubb, “Immigration and servitude in the colony and commonwealth of Pennsylvania: a quantitative and economic analysis,” PhD Thesis, University of Chicago, 1984 [previously unpublished].   1 See the language used in shipping advertisements for immigrant servants in the Penn­ sylvania Gazette and the Maryland Gazette.   2 Shippers seldom did repeat business or remained in port long enough to effectively provide a credible guarantee to American masters that the servants they sold had the skills claimed for them. A rare exception can be seen in the convict indentured servant market among a few shippers who did extensive repeat business and so could offer credible guarantees regarding the skills of their convict servants and, therefore, could command a higher price for said servants. Even here, it was done only for a small number of convict servants among the many who claimed to be skilled (Grubb 2001: 302).   3 Failure to distinguish between historic and modern usage of language led Salinger (1997: 331–3) to claim, erroneously, that Germans used the indenture and not the redemption method. See Grubb (1998: 235–41). Regarding Irish immigrants using

The method of entering servitude   241 both the redemption and indenture methods in the 1770s and 1780s, see the 1789 letter from Phineas Bond, British Consul at Philadelphia, to the British Foreign Office reprinted in Abbott (1924: 10).   4 To view a similar contract of a German immigrant entering servitude in America to pay for his passage from Rotterdam, see Mayer, H. (1738) “Servant Contract for Henry Mayer, 1738.” Online. Available HTTP: http://en.wikipedia.org/wiki/ File:Indenturecertificate.jpg (accessed 3 September 2010). See also the servant contract for Susanna Herbster transcribed in chapter 16.   5 See also the three English immigrant indentured servant contracts from 1684 reproduced in Galenson (1981: 41–3).   6 The titling page is written in a different hand than the rest of the document entries and includes the phrase “American Philosophical Society Presented by Thomas P. Roberts Philadelphia March 20, 1835.”   7 See note 5.   8 For examples, see the Pennsylvania Gazette 3 October and 21 November 1771; 23 January, 30 April, 9 July, 21 and 28 October 1772; 13 and 20 January, 5 and 12 May, 14 July, 1, 15, 22, and 29 September, and 6 and 20 October 1773; and the Maryland Gazette 22 October 1772 and 28 October 1773. See also the advertisement in the Pennsylvania Messenger reproduced in Herrick (1926: 274). The same language can be seen in the advertisement used to describe potential German immigrant servants on board the brig Bubona which arrived in Philadelphia in 1817. The advertisement is quoted in chapter 16.   9 See chapter 15; the passenger contract for the ships Favorite and Elizabeth in chapter 16; “Munstering [mustering] book”; Tepper (1975: 236, 257–8). 10 The full quotation is provided in chapter 14. It is taken from a letter quoted in Strassburger (1934: vol. 1, xxxvii); Diffenderffer (1899: 192). Mittelberger (1960: 16–18) gives a similar description in 1754 of German immigrants using the redemption method. 11 The full quotation is provided in chapter 14. It is taken from a letter quoted in Durnbaugh (1960: 231–3). Sauer subsequently became a respected Pennsylvania-­German printer and immigrant advocate. For the 1728 letter, see Sachse (1909: 5–25). 12 See Bruijn, Gaastra, and Schoffer (1979–87); Fertig (1994: 203); Lucassen (1994: 153–91). 13 See the cases of John Frederick Whitehead and Johann Carl Büttner, who were young single adults German redemptioners, in Klepp et al. (2006), and the cases of older adult German redemptioners on the ship Britannia in Chapter 15. 14 For a firsthand account of seasoning illness that dramatically reduced the value of a servant’s first year of labor, see the autobiography of John Frederick Whitehead in Klepp et al. (2006). 15 For example, see the firsthand accounts of John Frederick Whitehead and Johann Carl Büttner in Klepp et al. (2006).

References Abbott, E. (1924) Immigration, Chicago: University of Chicago Press. Akerlof, G. (1970) “The market for lemons,” Quarterly Journal of Economics, 84: 488–500. Barnard, E. K. (1904) Letters and other papers of Daniel Kent: emigrant and redemp­ tioner, Lancaster, PA: New Era Printing. Bruijn, J. R., Gaastra, F. S., and Schoffer, I. (1979–87) Dutch-­Asiatic shipping in the 17th and 18th centuries, vols. 1 and 2, The Hague: Nijhoff. Diffenderffer, F. R. (1899) “The German immigration into Pennsylvania through the port of Philadelphia, and ‘the redemptioners’,” Pennsylvania German Society, 10: 1–328.

242   German immigrant servitude in America, 1745–1835 Durnbaugh, D. F. (1960) “Two early letters from Germantown,” Pennsylvania Magazine of History and Biography, 84: 219–33. Fertig, G. (1994) “Transatlantic migration from German-­speaking parts of central Europe, 1600–1800: proportions, structures, and explanations,” in N. Canny (ed.) Europeans on the move: studies on European migration, 1500–1800, New York: Oxford University Press. Galenson, D. W. (1981) White servitude in colonial America, New York: Cambridge University Press. Grubb, F. (1985) “The market for indentured immigrants: evidence on the efficiency of forward‑labor contracting in Philadelphia, 1745–1773,” Journal of Economic History, 45: 855–68. Grubb, F. (1992) “The long-­run trend in the value of European immigrant servants, 1654–1831: new measurements and interpretations,” Research in Economic History, 14: 167–240. Grubb, F. (1998) “Labor, markets, and opportunity: indentured servitude in early America, a rejoinder to Salinger,” Labor History, 39: 235–41. Grubb, F. (2000a) “The statutory regulation of colonial servitude: an incomplete-­contract approach,” Explorations in Economic History, 37: 42–75. Grubb, F. (2000b) “The trans-­Atlantic market for British convict labor,” Journal of Economy History, 60: 94–122. Grubb, F. (2001) “The market evaluation of criminality: evidence from the auction of British convict labor in America, 1767–1775,” American Economic Review, 91: 295–304. Hatton T. J. and Williamson, J. G. (2005) Global migration and the world economy, Cambridge, MA: MIT Press. Herrick, C. (1926) White servitude in Pennsylvania, Philadelphia: J. J. McVey. Klepp, S. E., Grubb, F., and Pfaelzer de Ortiz, A., eds. (2006) Souls for sale: two German redemptioners come to revolutionary America: the life stories of John Frederick White­ head and Johann Carl Büttner, University Park, PA: Pennsylvania State University Press. Lucassen, J. (1994) “The Netherlands, the Dutch, and long-­distance migration in the late sixteenth to early nineteenth centuries,” in N. Canny (ed.) Europeans on the move: studies on European migration, 1500–1800, New York: Oxford University Press. Mayer, H. (1738) “Servant contract for Henry Mayer, 1738.” Online. Available HTTP: http://en.wikipedia.org/wiki/File:Indenturecertificate.jpg (accessed 3 September 2010). Mittelberger, G. (1960) Journey to Pennsylvania in the year 1750 and return to Germany in the year 1754, Cambridge, MA: Harvard University Press. “Munstering [mustering] book for the Britannia Capt. James Peters from Rotterdam 3rd July 1773,” Harrisburg, PA: Unpublished manuscript, Historical Society of Pennsylvania, AM 209. “Record of indentures of individuals bound out as apprentices, servants, etc. and of German and other redemptioners in the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773,” Philadelphia: Unpublished manuscript, Philadelphia City Archives. Sachse, J. F. (1909) “A missive from Pennsylvania in the year of grace 1728,” Pennsylva­ nia German Society, 28: 5–25. Salinger, S. V. (1987) “To serve well and faithfully” labor and indentured servants in Pennsylvania, 1682–1800, New York: Cambridge University Press. Salinger, S. V. (1997) “Labor, markets, and opportunity: indentured servitude in early America,” Labor History, 38: 311–38.

The method of entering servitude   243 Strassburger, R. B. (1934) Pennsylvania German pioneers, 3 Vols., Norristown, PA: Pennsylvania German Society. Tepper, M., ed. (1975) Emigrants to Pennsylvania, 1641–1819, Baltimore: Genealogical Publishing Company. Tomlins, C. (2010) Freedom bound: law, labor, and civic identity in colonizing English America, 1580–1865, New York: Cambridge University Press. Trautmann, F. (1981) “Pennsylvania through a German’s eyes: the travels of Ludwig Gall, 1819–1820,” Pennsylvania Magazine of History and Biography, 105: 35–65.

13 Servant contract choice and shipper profits*

Transporting servants from Europe to North America was an important and profitable business.1 Immigrants were the most convenient return cargo for the westbound voyage. However, the cost of passage exceeded most immigrants’ resources. Therefore, shippers financed the voyage in exchange for immigrant pledges to enter servitude in America to repay the loans. At the end of the eighteenth century approximately half of the immigrants arriving in Philadelphia entered into servitude.2 The most important finance methods were redemptioner and indentured servitude.3 The indentured method developed first, shortly after the establishment of Jamestown, and remained quantitatively important well into the eighteenth century.4 The redemptioner method developed in the Philadelphia market in the 1720s and quickly became universal in the German trade.5 The British also adopted it but at a slower pace (see Table 13.1). It has been suggested that shippers exploited servants by selling them for more than the cost of passage, thereby gaining excess profits, and by using the redemptioner instead of the indentured method which shifted repayment risks from shipper to servant.6 However, the servant trade was also competitive. Thus excess profits should not have been realized, and cost differences between contract types should have been fully compensated.7 Little has been done to investigate the evolution from indentured to redemptioner servitude, the determinants of contract choice, and the profitability of the trade.8

Indentured versus redemptioner servitude: contract insurance versus contract flexibility Indentured immigration was a barter transaction whereby shippers traded freight space and provisions for forward-­labor contracts.9 Contracts were fixed  before leaving Europe, commenced upon arrival in America, and were sold  at the shippers’ discretion. By taking possession of contracts before sailing,  shippers were guaranteeing the contracts’ terms. Any unexpected changes in labor value after the voyage had begun only altered the shippers’ realized sale price. Thus the risks in shipping indentured servants fell on the shippers.

Servant contract choice and shipper profits   245 Table 13.1 Indentured versus redemptioner servitude in the Philadelphia immigrant market, 1745–73 Total immigrant servants 1745–6 Irish German English Scottish 1771–3 Northern Irish Southern Irish German English from London English from Bristol Scottish

Percentage redemptioner

534 36 7 5

6.74%* 100.00 85.71 0.00

491 519 1,199 146 78 140

48.07%* 19.65* 100.00 71.92* 19.23* 100.00

Sources: George W. Neible, ed., “Servants and Apprentices Bound and Assigned Before James Hamilton Mayor of Philadelphia, 1745,” Pennsylvania Magazine of History and Biography, 30–2 (1906–8); “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia, October, 3, 1771 to October 5, 1773,” (unpublished manuscript, City Archives of Philadelphia); Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, 1934), vol. 1; and Michael Tepper, ed., Emigrants to Pennsylvania 1641–1819 (Baltimore, 1978): 180–1. Notes Northern Irish were from Newry, Belfast, and Londonderry. Southern Irish were from Cork, Waterford, and Dublin. Indentured and redemptioner contracts can be distinguished in the evidence as follows: indentured servants had their contracts transferred or “assigned” to the buyer by a third party other than the servant and the contract’s labor time commenced on the day of arrival in port, whereas redemptioners assigned themselves to their masters and their contracts commenced when the sale was made. * Means are statistically different from one another above the 0.05 level based on a one-tailed t-test.

The redemptioner method was a loan transaction secured by human-­capital collateral.10 Immigrants borrowed the price of passage along with money for extra provisions and port expenses from shippers before leaving Europe. They had to repay the debts within a few weeks of arrival in America where most negotiated a labor contract to sell for the amount they owed. This was a spot market in which the labor contract commenced only after sale. Except for death or escape, shippers experienced little risk with redemptioners because debts were fixed prior to sailing. Any change in labor value after the voyage had begun would alter the amount of labor redemptioners had to offer in America to repay their debts. Thus most of the risk in using the redemptioner method fell on the immigrant. The evolution from indentured to redemptioner contracting in a competitive market implies the existence of positive benefits in using the redemptioner method which compensated the immigrant for assuming the risks borne by the shipper under the indentured method. One benefit was a choice of masters. Shippers chose buyers for indentured immigrants, but redemptioners negotiated their own labor contracts on the spot in America and so had some choice of masters.

246   German immigrant servitude in America, 1745–1835 The choice was important and influenced the buying patterns of masters. For example, between 1771 and 1773 redemptioners had a 42 percent greater chance of having a Philadelphia master relative to similar indentured immigrants.11 This suggests that some servants preferred urban employment. The ability to choose masters was also important among families. Redemptioners found masters who would purchase the whole family, or who were located nearby, thus keeping the family intact. Parents could also supervise the contract negotiations of their children. Married servants with a joint contract, children with parental consent to contract, and other minors systematically chose redemptioner over indentured servitude (see Table 13.2). Evidence on contract sales also indicates that most young children went to the same master as their parents, and older children were sold before their parents suggesting some parental supervision of the sale.12 Families could also avoid some of the extra risks in being redemptioners by using the flexibility of the method to pool risks among family members (see the discussion in the next section). Another benefit the redemptioner method afforded was greater flexibility in designing the labor contract thus achieving a more advantageous match between servants and masters. Ideally, indentured immigrants would prefer forward contracts specifying a continuous schedule of contract lengths contingent on other nonpecuniary circumstances of servitude, such as the type, pleasantness, and location of the work, whether training or education was provided, the size of freedom dues, and so forth. Given this schedule, shippers would sell to the masters with the highest valuation among contract contingencies thus achieving an effecient match between masters and servants. In the absence of such a set of contract contingencies, shippers had an incentive to ignore nonpecuniary preferences of servants and masters and to sell the fixed dimension contracts to the highest bidder. Table 13.2 Distribution of contract characteristics by contract type in the Philadelphia immigrant servant market, 1745–73 (in percentages)

Occupational training provided Extra freedom dues required No freedom dues required Parental consent given Education provided Must be employed at a skilled trade Unmarried female Married with a joint contract German males not signing the loyalty oath (under age 16) Number of contracts

Indentures

Redemptioners

1.26% 0.59 0.07 0.00 0.00 0.37 23.33 0.59 0.00 1,350

4.62% 6.08 0.55 3.47 11.15 1.91 23.30 7.23 9.79 1,991

Source: See Table 13.1. Note The percentage differences between indentured and redemptioner contract characteristic are all statistically significant above the 0.001 level, except for the unmarried female category.

Servant contract choice and shipper profits   247 Indentured contracts with a matrix of contingency clauses were never observed in the market, so their benefits must have been negated by the cost of negotiating and forecasting the value of such contracts. Shippers had to accurately forecast the future American spot price when competing for indentured contracts in Europe. Forecasting the value of a large set of contingent sale conditions could be costly for shippers.13 Longer contracts would be required to compensate for the forecasting costs and for guaranteeing contingencies in contracts. Immigrant servants with a preference for restricted employment or specialized contract conditions could secure them at a cheaper price using the redemptioner method. Redemptioners could observe the revealed cost of contract contingencies in America before entering servitude. If special conditions, such as being trained to be a carpenter or receiving extra dues at the end of the contract, were more expensive than expected in terms of extra service time, the redemptioner could simply drop these stipulations from his set of contract demands. There were no incentives for both parties to renegotiate an indentured contract once the ship left Europe regardless of what information was revealed upon landing in America. But the redemptioner could, in effect, renegotiate his contract with himself when all the information was revealed after the voyage. This flexibility allowed redemptioners to have a higher frequency of special contract stipulations (see Table 13.2). By choosing the redemptioner method over the indentured method the immigrant was trading contract insurance, having the contract guaranteed before leaving Europe, for the flexibility of choosing his master and tailoring his labor contract to meet the circumstances at the time of arrival in America.

Risk and the rate of return to shipping redemptioner servants It is unlikely that redemptioners were exploited by shippers earning excess profits on financing their migration. Economic profits are measured by comparing returns to the relevant opportunity cost, adjusted for risk and uncertainty. Previous studies compared an average range of transportation costs with an average range of servant prices taken from literary sources. For example, Abbot Smith concluded: “Thus a merchant who spent from four to ten pounds on getting a servant to America could count on selling him for from six to perhaps thirty pounds. This was a comfortable profit, despite the large risks that sickness or death might scale down the value of the cargo”.14 His indirect estimates range too widely to support firm conclusions. They also do not account for the opportunity cost of freight space and provisions used, nor do they address the magnitude of risk compensation. Direct estimates of the returns to shipping servants can be made for German redemptioner immigrants. German immigrants had the option, specified contractually before sailing, of paying their passage before embarkation or upon arrival in America. An example of such a contract reads:

248   German immigrant servitude in America, 1745–1835 we agree to pay our passage with the following stipulations: Those who are in position to settle in Amsterdam, to pay for one person whether man or woman, (Children under 4 years old being free), from 4 to 14 years six and one-­half guineas, from 14 years and older thirteen guineas. Those who cannot settle here but will settle in America, to pay, (Children under 4 years being free), from 4 to under 14 years seven and one-­half guineas, from 14 years and older fifteen guineas. Those who pay their passage in America shall be bound to produce it within 10 days. No passenger shall be allowed to leave the ship in America without knowledge of the Captain, especially such as have not yet settled for their passage. If one of the passengers dies on the voyage, the family of such a person shall be obligated to settle for his passage15 A rate of return was calculated by comparing the cash price for free immigrants with the credit price for redemptioner servants on four German immigrant voyages (see Table 13.3). This measure accounts for the opportunity cost of the freight space and provisions employed because servants and free immigrants were close substitutes.16 The average return on shipping redemptioners based on the quoted fares ranged between 11 and 25 percent per voyage above the return to shipping prepaying immigrants. However, quoted fares differed from actual recorded debts because shippers gave extra cash advances, discounts off the quoted fare, and accepted partial prepayments.17 With adjustments the average actual return was 11 to 15 percent per voyage above the return to shipping free immigrants.18 This return represented the interest on the implicit passage loan plus compensation for the trouble of collecting on the loan in the colonies.19 A per voyage return of about 15 percent would be substantial given the average two-­month voyage. Annual interest rates on secure investments were around 5 percent.20 Either shipping redemptioners was profitable or the risks involved in financing redemptioner migration were high. The high return being compensation for undertaking the risk. Ignoring compensation to uncertainty per se, a servant default rate of approximately 12 to 13 percent per voyage would leave merchants with returns equal to those on safe investments.21 Passage mortality on completed voyages in the German trade averaged about 4 percent. In addition, the probability of shipwreck was about 1.5 percent.22 Therefore, defaults from expected voyage fatalities might explain about half the premium for shipping servants. Loan defaults also occurred upon arrival when servants escaped or refused to pay their debts.23 In addition, some servants arrived too sick to be sold or died soon after landing. Debarkation morbidity averaged 3.5 percent, and post-­voyage mortality in the first few months after arrival averaged 2.6 percent for German immigrants in the middle of the eighteenth century.24 Expected defaults from these sources might account for another 20 to 50 percent of the premium charged redemptioners. Therefore, at least half and maybe close to all of the premium charged redemptioners might be explained as recovering expected losses from defaults.25

Servant contract choice and shipper profits   249 Table 13.3 Average rates of return per voyage for financing German redemptioner immigration, 1802–19 (in percentages) Ships Belvidere (1802) Average rate of return from quoted fares Average actual rate of return Average actual rate of return on redemptioners migrating alone Average actual rate of return on redemptioners migrating in family groups Family members among all redemptioners

25.0%

Total passengers aboard

82

Commerce Pennsylvania Elizabeth (1803) (1803) (1819) 15.4%

20.5%

11.1%

10.7 (2.6) 13.3a (2.6)

14.6 (2.1) 16.2b (1.8)

11.5 (0.9) 12.2c (0.8)

5.4a (2.0)

11.5b (2.6)

7.1c (0.1)

32.9

33.3 202

24

13.4 120

Sources: Strassburger, Pioneers, vol. 3: 112–14, 131–4, 137–8; and “Passenger List of the Ship ‘Elizabeth’, Which Arrived at Philadelphia in 1819,” Pennsylvania Magazine of History and Biography, 25 (1901): 225–58. Notes Standard errors are in parentheses. The rate of return was calculated by subtracting the fare quoted those who paid prior to sailing from the fare quoted those who paid after the voyage and then dividing by the fare quoted those who paid prior to sailing. The average actual rate of return was similarly calculated except the amount of interest free loans given to servants and the discounts given free passengers were subtracted from their fare. Actual fare payments were listed by family on the Elizabeth and individually on the other vessels except for the Commerce which did not list passenger debts individually. Families were identified by their last name and by being listed together on the passenger lists. If any spurious family associations were made it would bias the estimated rates for single versus family redemptioners closer together than they actually might have been. All four ships departed Amsterdam and were bound for Philadelphia. The prepaid fare for the Pennsylvania was taken from the passenger contract of the Commerce, with appropriate currency conversions taken from John J. McCusker, Money and Exchange in Europe and America, 1600–1775, A Handbook (Chapel Hill, 1978): 9, 11. a Means are statistically different above the 0.001 level. b Means are statistically different above the 0.1 level. c Means are statistically different above the 0.001 level.

If risk considerations were important, then risk pooling and spreading should be observed. If redemptioners pooled their debts so that survivors paid debts of those who died, escaped, or were too sick to work, as well as their own debts, then loan defaults and consequently the risk premium charged would be reduced. However, the strategy might lead to a “moral hazard” problem. A servant might feign illness hoping to pass his debt on to a stranger, or the incentive to escape might be greater if one was forced to serve in another’s place. One congregation of individuals that might overcome the moral hazard problem was the family. If the welfare of spouses and children enters each other’s utility functions or represents assets in each other’s portfolios, then the incentive to cheat is lessened. Pooling debts among family members implies less

250   German immigrant servitude in America, 1745–1835 risk than doing so with strangers. Pooling debts was routinely written into passenger contracts for families. Contemporary accounts of the trade also reported family interdependence of passage debts.26 The existence of debt pooling suggests that the average rate of return for family members should be lower than for redemptioners migrating alone because the expected loan default was contractually reduced for families relative to single servants. Redemptioners were separated into family and nonfamily groups, and the rate of return was recalculated at the more disaggregate level (see Table 13.3). On each ship the average return to redemptioners in family groups was relatively lower. The difference averaged around one-­third to one-­half of the total premium to shipping redemptioners. This was approximately equal to the portion of the premium attributable to recovering defaults from expected mortality on completed voyages—exactly what the pooled debt contract was designed to achieve. Merchants competed for families by selectively giving extra cash advances or fare discounts. This lowered the rate of return accruing to families to the degree that the pooled debt contract reduced the expected voyage mortality risk to the merchant. Besides default recovery, some of the premium to shipping servants may have been compensation to uncertainty because low-­variance income streams are preferred to high-­variance streams which have the same expected average. The causes of mortality and morbidity which lead to defaults on passage loans, such as bad weather, exhaustion of provisions, contagious diseases, and arrival during periods of slack demand or yellow fever epidemics, were highly variable and voyage specific.27 Therefore, loan defaults covaried positively within a given shipload of servants. This increased the uncertainty cost on the per-voyage level. German immigrant shipping was consistent with risk spreading across voyages. Merchants did not specialize in servants on any given voyage and carried servants infrequently, which was a possible response to the voyage-­ specific positive covariance in risks. Their behavior suggests that uncertainty was an additional cost, though it cannot be directly calculated from the evidence. Subtracting expected defaults from the premium charged redemptioners would probably leave less than a quarter of the premium as residual compensation attributable to uncertainty. Passenger-­per-ton ratios were calculated to investigate the degree of specialization in human cargo on German immigrant voyages (see Table 13.4). The meaning of the absolute magnitude of the ratios can be illustrated in relation to slave-­per-ton ratios, a possible upper bound, which were over a third larger. Within the German trade the average passenger-­per-ton ratio was 23 percent higher on ships owned by the 20 largest shippers in Philadelphia, those with interest in 11 or more vessels. This would be consistent with risk spreading behavior because merchants with few vessels would have to carry a higher proportion of merchandise to human freight to achieve the same degree of risk spreading relative to merchants with many vessels.28 Shippers combined three different assets on each voyage—servants, free passengers, and other merchandise—which all had different risks and rates of return. Their overall risk position

Servant contract choice and shipper profits   251 Table 13.4 Comparison of passenger-per-ton ratios Number of voyages sampled German immigration to Philadelphia (1730–74) 66 Large merchantsa 34 Small merchants 32 Transatlantic slave trade Virginia (1727–69) 162 Jamaica (1782–6) 106

Average passengers per tonb 1.040 (0.650) 1.143* (0.650) 0.927* (0.640) 1.442 1.941

Sources: Herbert S. Klein, The Middle Passage (Princeton, 1978), pp. 124, 126, 142; Strassburger, Pioneers, vol. 1; and “Ships Registered for the Port of Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, 23–8 (1899–1904). Notes Standard errors are in parentheses. * Means are statistically different above the 0.1 level. a Large merchants were the top twenty or those with interest in eleven or more ships registered in Philadelphia, Simeon J. Crowther, “The Shipbuilding Output of the Delaware Valley, 1722–1776,” Proceedings of the American Philosophical Society, 117 (Apr. 1973): 103. The small merchants were all others. b All tonnage was converted into 1695 measured tons to allow better comparison of relative cargo capacity as follows: registered tonnage multipled by 1.971 for ships registered between 0 and 50 tons, 1.518 for ships between 51 and 100 tons, 1.422 for ships between 101 and 150 tons, and 1.338 for ships between 151 and 200 tons. The conversions were taken from a combination of John J. McCusker, “Colonial Tonnage Measurements: Five Philadelphia Merchant Ships as a Sample,” Journal of Economic History, 27 (Mar. 1967), pp.  82–91; Gary M. Walton, “Colonial Tonnage Measurement: A Comment,” Journal of Economic History, 27 (Sept. 1967), pp. 392–8; Christopher J. French, “Eighteenth-Century Shipping Tonnage Measurements,” Journal of Economic History, 33 (June 1973), pp. 434–43; and Charles Garland and Herbert S. Klein, “The Allotment of Space for Slaves Aboard Eighteenth-Century British Slave Ships,” William and Mary Quarterly, 3rd series, 42 (Apr. 1985), pp. 238–48. Where German immigrant vessels listed only the total number of freights or adult men, they were converted to total passengers for the purpose of calculating passenger per ton ratios following Wokeck, “German Immigration to Philadelphia,” p. 265. For German immigrant vessels, only ships on the Netherlands to Philadelphia route were considered, and only ships where the ship’s name and captain could be matched between the port register and the Strassburger ship lists were considered.

per voyage was thus improved because the positive risk covariance was lower among servants and other prepaid freight than just among servants. Among all German immigrant voyages studied, 75 percent were made by captains who participated only once.29 This figure is consistent with a high degree of competition, no excessive profit on average, and substantial risk which was gambled on infrequently. However, among captains who made over five German immigrant voyages, at least half were on ships owned by the top 20 shipowners in Philadelphia. Thus it appears that merchants with many vessels participated more frequently in the trade and carried relatively more passengers per ton. Top shipowners were better able to spread the risk of handling redemptioners across their business than merchants with few vessels.30

252   German immigrant servitude in America, 1745–1835

Conclusions In the early eighteenth century redemptioner servitude developed as an alternative to the common indentured method of financing passage to America. The innovation and evolution to the redemptioner method was not a product of shipper exploitation of poor immigrants, but rather market adaptation to the increasingly complex desires of some immigrants. It represents a shift from forward-­labor contracts to labor contracts formed on the American spot market. The transition involved trading a standardized guaranteed contract for a contract which allowed the flexibility to choose masters and special contract restrictions and to adapt to new information revealed after the voyage. The redemptioner method achieved a more efficient match between servants and masters. The complex contract needs of families quickly led them to rely on the redemptioner method. Finally, shippers did not realize excess profits transporting redemptioners. The premium charged redemptioners was a compensation for uncertainty and for recovery of expected defaults. The risk to shipping servants was reduced by risk pooling for families and risk spreading across voyages. The development of the redemptioner method in the transatlantic servant trade is suggestive of a vital and innovative market which was responsive to human desires, physical constraints, and economic costs.

Notes   * Originally published as Farley Grubb, “Redemptioner immigration to Pennsylvania: evidence on contract choice and profitability,” Journal of Economic History, vol. 46, no. 2 (June 1986): 407–418. Copyright 1986. © 1986 by Cambridge University Press and the editors of The Journal of Economic History. Reprinted with permission. Journal of Economic History, Vol. XLVI, No. 2 (June 1986). © The Economic History Association. All rights reserved. ISSN 0022–0507. The author is Assistant Professor of Economics at the University of Delaware, Newark, Delaware 19716. He wishes to thank Gary Becker, David Galenson, Henry Gemery, Claudia Goldin, Gary Libecap, Robert Margo, Jacob Price, Ralph Shlomowitz, and the members of the University of Pennsylvania Economic History Seminar for helpful comments on earlier drafts.   1 See K. G. Davies, The North Atlantic World in the Seventeenth Century (Minneapolis, 1974): 96–7; R. J. Dickson, Ulster Emigration to Colonial America 1718–1775 (London, 1966): 95; David W. Galenson, White Servitude in Colonial America (Cambridge, MA, 1981): 97–8; Karl F. Geiser, Redemptioners and Indentured Servants in the Colony and Commonwealth of Pennsylvania (New Haven, 1901): 47; Cheesman A. Herrick, White Servitude in Pennsylvania (Philadelphia, 1926): 175, 184–7; Richard Hofstadter, America at 1750 (New York, 1973): 33; J.R.T. Hughes, Social Control in the Colonial Economy (Charlottesville, 1976): 108; Stanley Lebergott, The Americans: An Economic Record (New York, 1984): 185; and Abbot E. Smith, Colonists in Bondage (New York, 1947): 35–42, 209.   2 Farley Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, 22 (July 1985): 316–39.   3 In America there were no legal or practical distinctions in the sale or operation of the two contract types. The key difference was in how they financed the voyage. Finance methods of lesser importance were serving according to the “customs of the country” and convict transportation.

Servant contract choice and shipper profits   253   4 See David Galenson, “The Rise and Fall of Indentured Servitude in the Americas: An Economic Analysis,” Journal of Economic History, 44 (Mar. 1984): 1–8.   5 Smith, Bondage: 21.   6 For example see Galenson, White Servitude: 14–15; Richard B. Morris, Government and Labor in Early America (New York, 1946): 319–22; and Smith, Bondage: 39–41.   7 For a discussion of competition in the trade see Farley Grubb, “Was the Shipping of German Immigrants to Colonial Pennsylvania Monopolized?” (unpublished manuscript, University of Delaware, 1985).   8 See the assessment in John J. McCusker and Russel R. Menard, The Economy of British America, 1607–1789 (London, 1985): 243.   9 See Galenson, White Servitude: 8–15, 97–113; and Farley Grubb, “The Market for Indentured Immigrants: Evidence on the Efficiency of Forward-­Labor Contracting in Philadelphia, 1745–1773,” Journal of Economic History, 45 (Dec. 1985): 855–68. 10 See Geiser, Redemptioners and Indentured Servants: 59–76; Morris, Government and Labor: 315–22; and Smith, Bondage: 20–5. 11 See Farley Grubb, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­Atlantic Economy,” Social Science History, 9 (Summer 1985): 256. 12 Derived from “Book of Redemptioners, 1785–1804” (unpublished manuscript, Pennsylvania Historical Society, Philadelphia). Galenson, White Servitude: 14; and Smith, Bondage: 22 also note the preference by families for the redemptioner method. However, the reasons given for the preference, that families wanted to partially prepay the passage, have contracts where the whole family was freed at the same time, or have children serve for their parents, are not supported by the contract evidence, see also fn. 17. Only husbands and wives had joint contracts. 13 Shippers carrying indentured contracts with infrequently used employment or sale restricts failed to correctly forecast the colonial value of the contracts, Grubb, “Market for Indentured Immigrants,” 867. 14 Smith, Bondage: 39. Dickson, Ulster Emigration: 95, used the same method to conclude that merchants received twice the money on servants relative to shipping free immigrants. These estimates do not match a given individual’s contract price and passage cost, making the measure very speculative. The reliance on servant prices as the proper comparison to passage costs may also be misleading. Servant prices contained returns other than transporting servants, such as implicit interest for purchasers who bought on credit and in the case of indentured servants returns to guarantee the contract terms before leaving Europe. 15 Strassburger, Pioneers, vol. 3: 131–4. 16 Most German immigrant ships carried both types of passengers who shared the same quarters and provisions during the voyage. 17 Extra cash advances were interest free loans added to the passage loan. This increased the total debt but decreased the implicit interest on the investment. It was the most popular way of adjusting passage fares; 80 percent received them on the Belvidere. Few partially prepaid, the largest was 8.3 percent on the Elizabeth. Discounts off the quoted fare were also used, the largest was 35 percent of the free passengers on the Elizabeth. See also the descriptions in Gottlieb Mittelberger, Journey to Pennsylvania in the Year 1750 and Return in the Year 1754 (Cambridge, MA, 1960): 16–18; and Strassburger, Pioneers, vol. 1: xxxvii. 18 The estimated return is similar to what John Dick, a redemptioner merchant, reported in 1750. He claimed it was customary to add about 15 percent to the passage fare “as an Indemnity for the Charges and laying out of the money,” Smith, Bondage: 40. 19 The redemptioner method provides the best evidence for measuring the returns to shipping servants because it was a pure loan agreement. Returns to shipping indentured servants included interest on the implicit passage loan, a return to risk, and also

254   German immigrant servitude in America, 1745–1835 a return to guaranteeing the contract’s labor terms before sailing or speculating in forward-­labor contracts. 20 See Sidney Homer, A History of Interest Rates (New Brunswick, 1963): 165; and Lawrence H. Officer, “Dollar-­Sterling Mint Parity and Exchange Rates, 1791–1834,” Journal of Economic History, 43 (Sept. 1983): 599–600. Given a two-­month voyage and a month or two for transaction time, the annualized return to shipping redemptioners would be around 45 to 60 percent. This is not to say that merchants could invest the same resources at this rate annually. At best two round-­trip voyages could be made a year. Annualizing this rate simply aids in comparing it to the returns of other investments which are expressed as annual interest rates. 21 This calculation used the credit and cash prices from the Commerce (see Table 13.3), and assumed an opportunity cost of capital ranging between a 3 to 5 percent annual interest rate on a two to four month investment. The formula used was: [(1 – default rate)(credit price) – cash price]/cash price = opportunity cost of capital per voyage. 22 See Farley Grubb, “Morbidity and Mortality on the North Atlantic Passage: Evidence from Eighteenth-­Century German Immigration to Pennsylvania,” Journal of Interdisciplinary History (forthcoming). Others have suggested a voyage mortality of 10 percent or more which would explain all of the premium charged redemptioners, see Henry A. Gemery, “Emigration From the British Isles to the New World, 1600–1700: Inferences From Colonial Populations,” Research in Economic History, 5 (1980): 186–7. The shipwreck rate was derived from Strassburger, Pioneers, vol. 1; and Marianne Wokeck, “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, 105 (July 1981): 254–5 for the German passenger trade to Philadelphia from 1727 to 1775. 23 For example, Ludwig Gall brought 13 redemptioners to Philadelphia in 1819. Two refused to repay their passage loan. Gall had them thrown into prison and recovered only part of his loans in court, see Frederic Trautmann, “Pennsylvania Through a German’s Eyes: The Travels of Ludwig Gall, 1819–1820,” Pennsylvania Magazine of History and Biography, 105 (Jan. 1981): 40–41. However, the frequency of such events is unknown. 24 See Frank R. Diffenderffer, “The German Immigration into Pennsylvania Through the Port of Philadelphia, and ‘The Redemptioners’,” Pennsylvania German Society, 10 (1899): 192; Mittelberger, Journey: 17–20; and Grubb, “Morbidity and Mortality.” 25 Periods of slack demand also slowed the recovery of passage loans. For example, Muhlenberg commented in 1764: “Four shiploads of German immigrants have arrived again this fall, but since money is scarce, the people are not being sold very fast,” Henry Melchior Muhlenberg, The Journals of Henry Melchior Muhlenberg (Philadelphia, 1942), vol. 2: 121. 26 See Mittelberger, Journey: 16–20; and Diffenderffer, “The German Immigration into Pennsylvania, 188–93. 27 The standard deviations in passage mortality and debarkation morbidity per voyage in the German immigrant trade were as large as their expected mean values. And the yearly standard deviation in post-­voyage mortality was about 77 percent of its yearly average, Grubb, “Morbidity and Mortality.” 28 The servant-­per-ton ratio was not known. Because servants were the risky freight and free passengers were relatively safe, comparing passenger-­per-ton ratios within the trade or with the slave trade, understates the relative diversification of risky versus less risky freight. 29 Derived from 324 German immigrant voyages to Pennsylvania between 1727 and 1775, Strassburger, Pioneers, vol. 1; and the sources cited in Table 13.4. 30 Merchants with many vessels may have also obtained some scale economies in information and reputation thus becoming specialized in the business.

14 The auction of German immigrant servants in Philadelphia, 1771–1804*

The redemption system dominated the final century of European servitude in America. It began replacing the older indenture system in the 1720s and became the prominent method of financing migration through the sale of long-­term labor contracts after mid-­century. Immigrant servitude continued under the redemption system after the American Revolution. For example, 45 percent of the 7,837 German immigrants arriving in Philadelphia between 1785 and 1804 negotiated servants contracts after landing and used the proceeds to redeem their transportation debts. The organized transportation of redemptioner servants ended in the 1820s.1 The redemption system transformed the auction of immigrant servants in America. Under the older indenture system, immigrants negotiated their servant contracts with shippers before embarkation. They could decline to migrate if satisfactory contracts were not found. Shippers recovered their freight costs by selling the indentures in America to the highest bidders.2 By contrast, redemptioners agreed to negotiate servant contracts of sufficient value to repay their transportation debts only after arrival in America. Because redemptioners could not return to Europe once the ship had landed, they may have been trapped in poor bargaining positions and forced to accept whatever servant contracts were offered.3 While narrative accounts indicate that the redemption system changed the American servant auction, little is known about the actual performance of the redemption auction from quantitative evidence.4 How did the auction develop, and how did it influence bargaining behavior, contract prices, and contract lengths? Were redemptioners forced to sell themselves for less than their opportunity cost, or did competition among buyers protect them from unfair contracts? Did the redemption auction limit the range of contract negotiations, or could redemptioners parlay productivity advantages into better contracts? Evidence from 4,455 British and German redemptioner servant contracts sold in Philadelphia between 1771 and 1804 are used here to study the performance of the redemption auction.

256   German immigrant servitude in America, 1745–1835

Development of the redemption auction in America Origins of the redemption system are obscure. Most descriptions involve the German passenger trade from Rotterdam to Philadelphia, where the system is believed to have begun. The first known reference to the redemption system is in a letter Christopher Sauer sent to friends in Germany one year after he migrated to Philadelphia in 1724. No one should indenture himself to the captain, but rather promise to pay on this side if he has no money or little money. One can borrow the money here and repay in cash. Many of the Palatines paid the captain nothing, although they had money. Instead, they bought knives, combs of ivory and horn, steel and nails, and only paid after they had come here. Everyone should act according to his own best judgement. The children of this world are very clever in their kind.5 In 1724 Sauer thought that redeeming transportation debts after migrating was a new and clever way of securing passage to America. It offered the possibility of avoiding servitude by borrowing money from family or friends already in America or by selling personal possessions upon arrival. While Sauer did not mention what would happen if funds could not be raised, the default option was servitude. The redemption system just postponed signing the servant contract from embarkation to debarkation. Initially, redemptioners were allowed to leave the ship and canvass Philadelphia and the nearby countryside for friends and relatives who might pay their transportation debts. If unsuccessful, redemptioners could still search for the best servant contract among employers of the region. Unlike indentured servants, redemptioners could bypass the shipper and negotiate directly with American buyers. They could bargain for their desired master, location, and work conditions. The unrestricted search for alternative options in America would produce the shortest contract needed to repay the transportation debt while maximizing the chance of avoiding servitude. By 1728 the practice of allowing redemptioners to leave ship before paying their debts had become problematic. Once off the ship, many redemptioners never returned. Shippers threatened legal action against redemptioners who disappeared without paying off their debts.6 They eventually solved the problem by keeping redemptioners on board until payment was rendered and by requiring payment within 10 to 14 days of arrival. By the 1740s the practice had become universal and was later incorporated as a standard clause in redemption transportation agreements. After mid-­century, eyewitnesses describe arriving ships as the marketplace where redemptioners waited for buyers to come and offer servant contracts or for friends to come and pay their freight. In 1765 this market practice was legally regulated to allow redemptioners 30 days to make payment after arrival with no additional charges for provisions while living on the ship during that period.7

Auction of German immigrant servants, 1771–1804   257 Search restrictions placed the redemptioner at a disadvantage. There was little the redemptioner could do if he were dissatisfied with the contracts offered by employers who happened to board the ship within the 30 days allotted. He could not return to Europe or freely search for better contracts. Debtors’ prison and arbitrary sale into servitude by the shipper were the only alternatives. In practice, most redemptioners took contracts offered within the restricted search period. From 1771 to 1804, 70 percent of redemptioners entered their contracts within 14 days and 90 percent within 30 days of arrival.8 The redemptioners’ precarious bargaining position has made the system appear exploitive. Redemptioners had little opportunity to find friends and relatives who would pay their passage debts. They may have been forced to take unattractive servant contracts, required to work more years than needed to repay their passage debts, and found that individual productivity advantages could not be parlayed into better contracts.9 Thirty years after initially recommending the redemption system, Christopher Sauer’s opinion had changed. In 1755 Sauer sent a letter to the Governor of Pennsylvania describing the plight of redemptioners as their ships arrived in Philadelphia. They are anxious to come on shore to satisfy hunger—they pay what is demanded—some sighing, some cursing; some believe their case differs little from such as fall into the hands of a highwayman, who presents a pistol and demands according to his own terms.10

Competition and average redemptioner servant compensation Competition should have protected the average redemptioner from any bargaining disadvantages created by search restrictions. By confining redemptioners to their ships, the burden of the search was shifted from redemptioners to their American employers. Competition would lead employers to search across redemptioner ships for servants who cost less than comparable resident labor. Competition among the many buyers boarding redemptioner ships would reveal the market rate of compensation, and, at the margin, redemptioners would receive contracts that produced compensation equal to comparable resident labor income. Because immigrants would not know the market rate of compensation or who was hiring labor when their ships landed in America, organizing the market with employers searching for servants rather than servants searching for employers may have been more efficient.11 American employers experienced few informational or market barriers in competing for redemptioner servants. The Pennsylvania Gazette reported the weekly arrival of ships in Philadelphia, including ship names, captains, and European ports of origin. These reports informed employers of the arrival of redemptioners because most ships from Holland and Ireland carried redemptioners, particularly in the fall. Many shippers also placed advertisements in weekly newspapers stating the number of redemptioners available. Because redemptioners negotiated

258   German immigrant servitude in America, 1745–1835 their servant contracts directly with American buyers, shippers did not control the sale of redemptioners and did not restrict access to the auction. While shippers occasionally sold large blocks of indentured servants to one buyer, redemptioners seldom sold themselves in blocks. Most buyers acquired only one redemptioner or one redemptioner family.12 Diverse locations, occupations, and ethnic backgrounds were represented among the buyers. In 1750 Mittelberger observed that, “Every day Englishmen, Dutchmen, and High Germans come from Philadelphia and other places, some of them very far away, sometimes twenty or thirty or forty hour’s journey, and go on board the newly arrived vessel that has brought people from Europe and offers them for sale.”13 Salient features of each servant sale were recorded before a Philadelphia magistrate, including the contract price, length, and other special contract stipulations, the European port of origin, the servant’s name, and the buyer’s name, occupation, and geographic location. The record legally protected both servant and buyer against attempts to change the servant contract. Sale records of 2,135 adult male redemptioners for the years 1771–3 and 1785–1804 are used here to estimate average redemptioner compensation (see Table 14.1). To compare relatively similar groups of immigrants and residents only adult males are used, and British and Germans redemptioners are considered separately. The majority of redemptioners were adult males.14 In exchange for several years of labor, every redemptioner received three forms of compensation: the contract price, freedom dues, and maintenance. Passage debt represented the bundle of transportation goods the redemptioner had already consumed. The servant contract was negotiated to sell for a price equal to the passage debt, which was paid directly to the shipper. Freedom dues consisted of a bundle of money and goods paid at the termination of the contract like a lump-­sum, non-­vested pension. The legally mandated “customary” dues were two suits of clothes, one of which was to be new, or its monetary equivalent. Many redemptioners contracted for extra freedom dues which were added to the customary dues. All servants were maintained by their masters, which included the provision of food, shelter, and clothing. While the average contract price and the value of freedom dues are taken directly from sale records, direct evidence on individual servant maintenance does not exist. I use Billy Smith’s estimate of minimal yearly maintenance budgets for poor Philadelphians to measure the opportunity cost of yearly maintenance compensation received by redemptioners. The average contract price and value of freedom dues are divided by the average number of years of the contract and added to the yearly maintenance compensation to derive the redemptioner’s average yearly compensation. The secular decline in servant contract lengths under the indenture system continued under the redemption system.15 While transportation and maintenance costs stayed constant or even rose, the average labor cost of transatlantic passage declined from four to three years for adult males between 1771 and 1804. The successful negotiation of shorter contracts over time would be consistent with a secular rise in servant productivity.16

Auction of German immigrant servants, 1771–1804   259 Passage debt averaged 13 Pennsylvania pounds for British redemptioners and 22 Pennsylvania pounds (or 75 U.S. dollars) for German redemptioners. The belief that redemptioners commonly prepaid part of their passage fare in Europe is not supported by the evidence. The average transportation debt, especially for Germans, is higher than the freight charge for passage. Most redemptioners borrowed sums in excess of the freight charge to pay for extra voyage provisions and other pre-­embarkation expenses. Total transportation debt averaged less than half a year’s income for a resident Philadelphia worker, but adult male redemptioners were required to work three to four years to repay the debt. While this comparison suggests that immigrant servants were exploited, the comparison ignores servant compensation other than contract price.17 Approximately 78 to 82 percent of the redemptioner’s yearly labor value went to pay for maintenance costs. Freedom dues represented another 6 to 7 percent of the redemptioner’s yearly compensation, which may have been a substantial savings rate for poor Americans.18 The contract price represented only 10 to 16 percent of the redemptioner’s yearly compensation. After deducting the cost of maintenance and freedom dues, little of the year’s labor time was left to contribute toward repaying the passage debt. Therefore, several years of servitude were required to repay a debt that equaled less than half a year’s income. Total yearly compensation for adult male redemptioners, adjusted for expected defaults and discounted to the point of sale, averaged 34 to 36 Pennsylvania pounds in the early 1770s, 51 Pennsylvania pounds from 1787 to 1795, and 204 U.S. dollars from 1796 to 1804. This compensation equaled the yearly income of free resident Philadelphia laborers who worked between 208 and 260 days a year, but was below the yearly income of resident artisans.19 Whether redemptioners were paid their opportunity cost would depend on the proportion of laborers to skilled artisans among redemptioners. While the exact composition of redemptioner skills is unknown, unskilled laborers and low-­ skilled artisans probably predominated. For the few redemptioners who were artisans, their skills may have been below that of resident journeyman artisans because adult redemptioners were typically quite young, in their early twenties. Therefore, redemptioner incomes should probably be compared with resident laborer rather than resident artisan incomes in measuring the opportunity cost of redemptioner labor.20 While the estimates in Table 14.1 are too crude to support a definitive conclusion, redemptioners were not grievously exploited by their bargaining position in the servant auction. Their compensation equaled the income of resident Philadelphia laborers, which approximated the opportunity cost of their labor. Competition among buyers protected redemptioners from exploitation in the servant auction. The importance of buyer competition may explain why redemptioners only accepted transportation to large American ports, such as Philadelphia and Baltimore, where the pool of potential buyers was known to be large.

82.4 100.0%

Estimated minimal maintenance per year Food 9.61 Rent 11.24 Fuel 4.44 Clothing 2.40 Subtotal 27.69 Average yearly compensation 33.61

31.23

51.13

40.11

3.74 (2.09)

208 work days per year

Artisans 260 work days per year

208 work days per year

Average contract length (years)

Free Philadelphia residentse Laborers 260 work days per year 39.04

Comparisons of estimated yearly income Redemptioner immigrant 33.51 servantsd

0.3

10.4% 6.9

Average servant compensation Contract price (passage debt)a 13.07 (3.84) Value of customary freedom 8.67 (2.06) duesb Value of extra freedom duesc 0.41

1771–3 British (N = 583) (in Pennsylvania pounds)

40.11 4.09 (1.31)

51.13

31.28

39.04

36.31

9.61 11.24 4.44 2.40 27.69 35.55

0.69

22.79 (5.19) 8.67 (2.06)

77.9 100.0%

0.4

15.7% 6.0

1771–3 German (N = 510) (in Pennsylvania pounds)

94.62 47.78 75.70 38.23 3.09 (1.28)

51.84 57.35 41.47 45.88

51.06

13.84 16.63 6.07 3.28 39.82 50.60

1.20

22.03 (3.76) 10.08 (1.46)

78.7 100.0%

0.8

14.1% 6.4

1787–95 German (N = 450) (in Pennsylvania pounds)

Table 14.1 Redemptioner servant compensation for single adult male immigrants: Philadelphia, 1771–1804

416.00 183.37 332.80 146.70 3.18 (1.28)

260.00 282.21 208.00 225.77

204.16

63.37 58.60 33.97 12.15 168.09 203.65

9.28

74.97 (10.30) 28.83 (12.25)

82.5 100.0%

1.4

11.6% 4.5

1796–1804 German (N = 592) (in U.S. dollars)

Notes Standard errors are in parentheses. Only immigrant servant redemptioners arriving in Philadelphia who were unattached adult males and who had their contract price denominated either in Pennsylvania pounds or U.S. dollars were considered. Where necessary, values were converted using dollars/Pennsylvania pounds = 3.16, which was derived by comparing average market prices for similar groups of labor contracts. a The third column under each heading is the percentage distribution of the components of average yearly servant compensation. b A monetary sum “in lieu of customary freedom dues” was written into 652 contracts, often with the servant given the option of the money or the clothes. The average monetary equivalent to customary dues in these contracts is the number reported. c Between 1771 and 1773, 6.2 percent of the British and 12.4 percent of the German redemptioners in the table contracted for extra freedom dues, and from 1787 to 1804 about 25 percent of the redemptioners contracted for extra dues. The average amount of money expected in a typical contract is the number reported. Extra dues in goods or land, 0.7 percent and 0.6 percent of the contracts in 1787 to 1804, were not included. d [(contract price + customary freedom dues + extra freedom dues)/contract length] + per year maintenance, discounted and adjusted for expected defaults. The average contract length was adjusted for expected contract default through escape, 5 percent, and death, 6.14 percent in the first year and 3.7 percent in subsequent years, see Bernard Bailyn, The Peopling of British North America (New York, 1986): 64; Farley Grubb, “Morbidity and Mortality on the North Atlantic Passage: Eighteenth-Century German Immigration to Pennsylvania,” Journal of Interdisciplinary History, 17 (Winter 1987): 577–84; Richard B. Morris, Government and Labor in Early America (New York, 1946):  443–9; Sharon V. Salinger, “To Serve Well and Faithfully:” Labor and Indentured Servants in Pennsylvania, 1862–1800 (New York, 1987): 133; Billy G. Smith, “Death and Life in a Colonial City: A Demographic Analysis of Philadelphia,” Journal of Economic History, 36 (Dec. 1977): 871. The resulting number of expected years of service was discounted to the point of sale at an interest rate of 10 percent, see Sidney Homer, A History of Interest Rates (2nd edn., New Brunswick, 1963): 279. Freedom dues were adjusted for the probability that the servant would not be present at the end of the contract to collect and discounted to the point of sale at an annual rate of 10 percent. The contract price and adjusted freedom dues were divided by the expected years of present value in the average contract. Maintenance per year was adjusted for the probability of default within a year and added to the sum to derive the expected labor income per year at the point of sale. No adjustments were made for expected morbidity or disability. e The first income estimates listed in the 1787 to 1795 and 1796 to 1804 columns were derived from Donald R. Adams, Jr., “Wage Rates in the Early National Period: Philadelphia, 1785–1830,” Journal of Economic History, 28 (Sept. 1968): 406. All other free Philadelphian resident income estimates were derived from Billy G. Smith, “Material Lives of Laboring Philadelphians, 1750–1800,” William and Mary Quarterly, 38 (Apr. 1981): 181–200, with the artisan incomes being for journeyman tailors only. The ethnic composition of these estimates is not known.

Sources: “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773” (unpublished manuscript, City Archives of Philadelphia); “Book A of Redemptioners, 1785–1804” (unpublished manuscript, Pennsylvania Historical Society, Philadelphia); Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, 1934), vols. 1–3. Maintenance estimates were adapted from B. G. Smith, “Material Lives,” 163–202.

262   German immigrant servitude in America, 1745–1835

The redemption auction Besides producing average compensation equal to the average opportunity cost of labor, competition among American employers should have produced variance in redemptioner compensation commensurate with the variance in redemptioner productivity. Redemptioners should have been able to bargain productivity advantages into better contracts and purchase desirable work conditions with contract concessions. While existing eyewitness descriptions of the redemption auction lack detail, they offer a consistent outline of the bargaining process. For example, Pastor Muhlenberg described the auction in a letter sent to Germany in 1769: The ship becomes the market-­place. The buyers make their choice among the arrivals and bargain with them for a certain number of years and days. They then take them to the merchant, pay their passage and their other debts and receive from the government authorities a written document, which makes the newcomer their property for a definite period. Young and unmarried persons of both sexes are sold first . . . Married people, widows, and the infirmed are dull sale.21 These descriptions indicate that redemptioners negotiated directly with individual buyers, that the bargaining parameter was the contract length in daily increments, and that the contract price equaled the redemptioner’s transportation debt which was fixed before the auction. Many accounts also reported the average contract length and indicated that individual contract lengths depended on age, ability, health, and marital status. The redemption system altered how contract prices and lengths cleared the servant auction. Under the older indenture system, contract lengths were contractually fixed in Europe, long before the American auction. Bargaining over contract prices cleared the American servant auction. The auction of indentured servants in America was the end-­sale in a forward market for long-­term labor contracts. The variance in contract prices measures the degree of merchant error in forecasting servant values. The variance in contract lengths measures the merchant’s forecast before leaving Europe of the expected value of the servant in America.22 By contrast, under the redemption system contract prices were contractually fixed in Europe, and bargaining over contract lengths cleared the American servant auction. The redemption auction was a spot market in long-­term labor contract. The variance in contract lengths measures the variance in servant skills and work amenities at the point of final sale in America. The variance in redemptioner contract prices represents only variance in transportation debts. The redemption auction can therefore be modeled as an asset-­pricing market for long-­term labor contracts. The transportation debt is equivalent to the contract price. Because the debt was determined by contract in Europe months before the auction, it was an independent variable in the servant auction. The redemptioner was constrained to negotiate a contract of sufficient value to repay

Auction of German immigrant servants, 1771–1804   263 this transportation debt. In a competitive market the value of this servant contract would equal the discounted summation of the value of the redemptioner’s labor productivity net of maintenance over the life of the contract, equation 1: e–rtdt= transportation debt = CP.

(1)

The net value of labor productivity is modeled as a function of fixed immigrant traits observed upon arrival in America and work conditions desired by the contracting parties, (VNP) = f (Z). These characteristics are modeled in the auction as a vector of independent variables compensated by or purchased through adjustments in contract length.23 The function estimated is a double-­log version of equation 1 where Z is a multiplicative exponential vector, r is a constant, and the contract length is the dependent variable:

(2)

Where: CP = contract price; T = contract length; VNP = value of labor productivity net of maintenance; r = the discount rate; Z = a vector of servant, sale, and contract characteristics at the time of the auction, excluding contract length and price; and a = estimating coefficients. The model was estimated using 561 German and 785 British redemptioner contracts from 1771 to 1773 (see Table 14.2) and 3,109 German redemptioner contracts from 1787 to 1804 (see Table 14.3).24 Overall, the independent vari­ ables explain a significant portion of the variance in contract length and their pattern is consistent with a labor productivity explanation of the redemption auction.25 Less productive redemptioners had to work longer to repay a given debt. A wide variety of contract conditions and work amenities were purchased through adjustments in contract length. The number of variables systematically related to contract length indicates that search restrictions placed on redemptioners did not severely reduce their ability to negotiate individual productivity advantages into shorter contracts or to tailor work conditions to meet individual needs. Competition among American employers led to more detailed and complex redemptioner contracts than observed under the indenture system.26 Many servant characteristics observed by the buyer were related to productivity, such as sex, height, weight, strength, age, health, ethnic origin, demonstrable skills, and literacy. While productivity can be modeled as a vector of these observed traits, not all observable characteristics were recorded, for example height, weight, strength, and health were not. Age is available only for a subset of 430 contracts (14 percent) in the 1787–1804 sample, which was estimated separately in Table 14.3. Many of the independent variables should be interpreted as proxies for unavailable measures of productivity. The impact of large numbers of children among German redemptioners is controlled by limiting the 1771–3 German sample in Table 14.2 to males above

R2 F-statistic Durbin–Watson statistic

Port of departure London Ulster South Ireland Isle of Lewis, Scotland Liverpool and Bristol

Constant Year of arrival Natural log of contract price Unmarried females Married with a joint contract Illiterate Parental consent to contract Education provided Occupational training provided Days between arrival and contract sale × 10–2 (Days)2 × 10–3

Explanatory variables

0.29*** (0.05) 0.20*** (0.02) 0.28* (0.11) –0.03* (0.02)

7 6 561 561

0.51 20.84*** 1.99

–0.08** (0.03)

0.10*** (0.02) 0.05*** (0.01)

51 149

89

18.82 (23.07) –0.01 (0.01) 0.34*** (0.03)

561 561 561

–12.5

97.7 50.3 2.8

22.0 8.7

6.0%

95.0 75.0 46.0 2.0

5.3%a 25.2

0.48 18.21*** 1.84

–0.07* (0.05) –11.2 –0.17*** (0.04) –23.0 –0.11* (0.05) –16.5 0.07* (0.04) 12.9 –0.06 (0.08)

0.33** (0.13) 0.29*** (0.04) 0.21*** (0.04) 0.11 (0.07)

28 86 62 512

105 236 71 104 15

0.34*** (0.03) 0.13* (0.02) 0.06 (0.05)

–0.04 (0.10)

785 172 30

785

Coefficient

Number of cases

Number of cases

Coefficient

British redemptioners

German adult male redemptioners

Table 14.2 The auction of British and German redemptioner immigrants: Philadelphia, 1771–3

Notes Standard errors are in parentheses. The method of estimation is ordinary least squares. The dependent variable is the natural log of contract length in years continuously discounted at 25 percent, see text and Farley Grubb, “The Market for Indentured Immigrants: Evidence on the Efficiency of Forward-Labor Contracting in Philadelphia, 1745–1773,” Journal of Economic History, 45 (Dec. 1985):  860–64. The results are relatively invariant to discount rates between 5 and 50 percent. The independent variables are constructed as follows. The natural log of contract price and extra freedom dues are expressed in decimal equivalent Pennsylvania pounds. Unless otherwise indicated, all variables are dummy variables and equal one if the listed condition was present and zero otherwise. Illiterates were servants who were unable to make a signature. The reference category for the German regression is unmarried males above age 15 embarking at Rotterdam. The reference category for the British regression is unmarried adult males embarking at Dublin. Included in the regression, but not printed, were dummy variables representing the month of arrival, adjustments to customary freedom dues, purchases by agents for absentee buyers, and residence of the purchaser by county and city. * Significant above the 0.1 level. ** Significant above the 0.01 level. *** Significant above the 0.001 level. a These columns report the percentage deviation from the reference contract length for each significant independent variable, evaluated at sample means for continuous variables. The deviation from average contract price was evaluated at a 10 percent increase in contract price. The deviation for days between arrival and contract sale was evaluated at twice the average of 7.5 days for the Germans including (Days)2, and at twice the average of 11.3 days for the British. The British Days coefficient is marginally significant at the 0.13 level.

Sources: “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773” (unpublished manuscript, City Archives of Philadelphia); Michael Tepper, ed., Emigrants to Pennsylvania 1681–1819 (Baltimore, 1978): 180–81; Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, 1934), vols. 1–2.

Occupation of purchaser Farmer Baker Cordwainer Esquire or Gentleman Storekeeper

Constant × 102 Year of arrival (Year)2 Natural log of contract price Unmarried adult female Married with joint contract Single parent Dependent child Parental consent to contract Education provided Occupational training provided Age (Age)2 (Age)3 × 10–2 Number of siblings plus one Number of children per married couple Number of children per single parent Number of children remaining free with servant parentsb Days between arrival and contract sale × 10–1 (Days)2 × 10–3

Explanatory variables

–5.19** (2.13) –7.51*** (2.56) –3.10** (1.49) –3.22* (2.24)

0.38 (0.35) 0.07 (0.49) –1.86** (0.78) 1.15 (1.15) 0.06 (0.27) 0.04 (0.13)

794 530 50 273 3,109 3,109

71 46 190 59

5,259.18*** (539.13) –585.52*** (60.02) 0.16*** (0.01) 0.91* (0.58) 2.19*** (0.78) 1.96* (1.23) 6.73*** (2.56) 13.13*** (1.82) 17.61*** (1.29) 21.44*** (1.12) 7.51*** (2.61)

3,109 3,109 3,109 3,088 319 530 50 794 536 481 43

–7.7 –10.9 –4.7 –4.9

–2.9

139

–3.07** (1.61)

–5.95*** (0.60) 0.15*** (0.02) –0.10*** (0.02) 0.41 (0.68) 0.19 (0.97) 3.77** (2.03) 2.06 (1.95) –0.70 (0.60) –0.30 (0.30)

–4.1

5.4

14.00*** (2.33) 22.0

59 430 430 430 117 98 32 98 430 430

6.47*** (1.32) 0.9 0.23 (1.59) 0.58 (2.29) –2.35 (5.40) 6.82** (3.83) 10.0

430 113 98 32 117

0.1 3.5 3.2 11.5 24.5 35.0 45.3 13.0

–2.9%a

8.96* (5.52) –0.43* (0.31)

430 6.8%a 430

Coefficient × 10–2

Number of cases

Number of cases

Coefficient × 10–2

Age subsample

Full sample

Table 14.3 Auction of German redemptioner immigrants: Philadelphia, 1787–1804

67 55 145 246 68 191 565 0.55 57.09***

–3.29* (2.20) –4.95** (2.41) –2.30* (1.58) –4.73*** (1.37) –3.42* (2.19) –0.69 (1.44) 3.00*** (1.15) 4.9

–5.0 –7.4 –3.5 –7.1 –5.2

0.67 41.32***

Notes See notes to Table 14.2 and text. The natural log of contract price is expressed in U.S. dollars. Where prices were expressed in pounds they were converted to dollars, see notes to Table 14.1. Included in the regression, but not printed, are dummy variables representing the month of arrival, adjustments to customary freedom dues, port of embarkation, and residence of the purchaser by county and city. The reference category for the occupation of the purchaser is farmer for the full sample and being other than a farmer for the age subsample. The reference category for the rest of the regression is unmarried adult males. The method of estimation is ordinary least squares corrected for first-order autocorrelation. The autocorrelation coefficient is –0.21 with a 0.02 standard error for the full sample and –0.25 with a 0.05 standard error for the age subsample. The uncorrected least squares regression had a Durbin-Watson statistic of 1.56 and 1.53 for the full sample and age subsample indicating positive serial correlation in the residuals. While uncorrected positive serial correlation will not affect the bias or consistency of the estimates, it will understate the standard errors of the coefficients causing the statistical significance of the regressors and the R2 to be inflated. Typically, the source of positive serial correlation is difficult to identify. A possible cause here is the sequential sale among family members. Because the observations were coded in order of sale, a nonrandom sales sequence among family members may have caused some positive correlation in the residuals. Sale sequence for each family member could not be controlled in the regression. * Significant above the 0.2 level. ** Significant above the 0.1 level. *** Significant above the 0.01 level. a These columns report the percentage deviation from the reference contract length for significant independent variable, evaluated at sample means for continuous variables. The deviation for average contract price was evaluated at a 10 percent increase in contract price. The deviation for year of arrival was evaluated at a fiveyear increase from the average sample year of 1795, including (Year)2. For the full sample, the time trend was a parabola which reached a minimum around 1795 with an average 6.7 percent increase in contract length when moving to 1790 or 1800. The secular pattern in the full sample was probably caused by shifting age and skill patterns among the immigrants over the period. See text for the impact of age on contract length. b Many parents signed contracts that allowed some of their children to continue to live with them while they were servants. The children were exempt from servitude and maintained at the expense of their master.

Sources: “Book A of Redemptioners, 1785–1804” (unpublished manuscript, Pennsylvania Historical Society, Philadelphia); Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, 1934), vol. 3.

Adjusted R F-statistic

2

Tailor Tanner or skinsdresser Other retail trade Other manufacturing trade Other construction trade Other service trade Merchant

268   German immigrant servitude in America, 1745–1835 age 16.27 British redemptioners were mostly adults with few family groups, and the complete British sample was used in Table 14.2. Because family membership and dependency could be identified and controlled in the 1787–1804 evidence, the complete sample from 1787 to 1804 was used in Table 14.3. Variables related to lower productivity were systematically associated with longer contracts. Children were less productive than adults and, therefore, had to accept longer contracts. Dependent children served 95 percent longer in the British sample and 24 percent longer in the German sample. Controlling for age, the contracts of children migrating as dependents of family groups were still 10 percent longer. Independent of age-­related productivity differences, dependency was associated with longer contracts and thus even lower productivity. Successively longer contracts were associated with illiteracy, the provision of occupational training, and the provision of education. Illiterate adults served 9 percent longer. Contracts with occupational training were 46 to 50 percent longer in 1771–3 and 13 percent longer in 1787–1804. Contracts providing education were 75 to 98 percent longer in 1771–3 and 45 percent longer in 1787–1804. Because learning to read and write usually preceded learning a trade, longer contracts for education compared with training may have been caused by age-­ related productivity difference within the controlled categories. Controlling for age reduces the coefficient on education by a half, to a 22 percent longer contract. This reduction indicates that the cost of acquiring education, independent of productivity differences relating to age, was a 22 percent increase in service time.28 Age was an important signal of relative productivity among children and teenagers and was closely related to contract length. Studies of age and contract length among indentured servants have only explored a relatively narrow range, ages 15 to 20.29 Among German redemptioners ages ranged from 3 to 69, and children comprised 25 percent of the 1787–1804 sample. Contract length declined in a one-­to-one relationship with the rise in age from ages 3 to 12. This decline was slightly faster among females. Males under age 12 served until about age 19. Females under age 12 served until about age 17. After age 12 the decline in service time tapered off as age increased, leveling off for females around age 18 and leveling off for males in their early twenties. Age-­related differences in contract length were relatively insignificant after age 21. Age-­contract-length profiles look similar to age-­earnings profiles.30 Females had relatively shorter contracts among children and young teenagers, but relatively longer contracts among adults, similar to patterns found in slave and indentured servant markets.31 The market found females to be relatively more productive as children and relatively less productive as adults. When only child status was controlled, adult females served 25 percent longer among the 1771–3 British redemptioners, and 3.5 percent longer among the 1787–1804 German redemptioners. Longer contracts for adult females were greater for British redemptioners than for British indentured servants in the same market. Because the redemption estimate is a spot market value and the indenture estimate is a forecasted value, the difference in the gender premium between the two

Auction of German immigrant servants, 1771–1804   269 types of contracts indicates that merchants mistakenly overvalued female indentured servants and lost money on shipping them.32 Married servants were sold under one contract. They could not be involuntarily separated, and both spouses received the same contract length. Married persons took several days longer to find a buyer than single redemptioners, and they had to offer longer contracts to entice employers to purchase their restrictive contracts. Marital tie-­in contracts were 22 percent longer in 1771–3 and 3 percent longer in 1787–1804. However, other differences in family structure, such as the number of siblings, number of children per couple, and number of children allowed as free-­riders on their parents’ contract, had little impact on contract length. Even within the relatively small pool of buyers who would accept a family tie-­in contract, competition prevented exploitation. Redemptioner families were able to package group sales and joint contracts at relatively little cost in terms of extra work, especially after 1787, which may explain why families were particularly attracted to the redemption system. The order of sale and the resulting sequence of auction prices were different between the redemption auction and the auction of slaves or indentured servants. Slaves and indentured servants were the property of the ship captain or merchant who transported them. The shipper organized the auction, usually selling laborers in sequence from most valuable to least valuable. Organizing the order of sale by descending productivity yielded a declining price of slaves and indentures over the course of the auction.33 By contrast, the auction on board redemptioner ships was not organized because redemptioners were not the property of the shipper. Each redemptioner negotiated his own contract and could freely search among buyers who boarded their ship. German redemptioners took an average of 7.5 days, and British redemptioners took an average of 11.3 days to form a contract after arrival. More search time should have yielded shorter contracts for redemptioners. However, increased search time was associated with slightly longer contracts in 1771–3, and search time was unrelated to contract length in 1787–1804. The search restrictions placed on redemptioners may explain this relationship. Redemptioners only had 30 days to search and could only search across buyers who boarded their ship. On any given ship the reservoir of available servants declined as servants were sold. Thus the pool of prospective buyers dwindled over the course of the auction. The insignificant relationship between contract length and search time may have been caused by redemptioners balancing the benefits of continued search across successive buyers against the costs associated with a declining pool of buyers.34 The occupation of the purchaser was systematically related to contract length and influenced contract negotiations in the redemption auction. Redemptioners purchased by masters in the manufacturing, retail, and construction trades received shorter contracts than servants purchased by farmers. The more productive redemptioners, those who could bargain for shorter contracts because they were older or more skilled, were systematically purchased by tradesmen, manufacturers, and businessmen in the community. Farmers may have had a relative

270   German immigrant servitude in America, 1745–1835 advantage in purchasing young or unskilled servants when the work was largely farm labor. Controlling for age, the estimated relationship between contract length and employer occupation was reversed. Per age or experience level, farmers had to offer shorter contracts to entice redemptioners into their employment. Therefore, farm work may have been least preferred by redemptioners.35 Redemption auction data are used to estimate the elasticity of contract length with respect to contract price (the marginal value of additional work at the end of the contract), a value which cannot be estimated for indenture auctions. While redemptioners had to work longer to repay larger debts, the relationship is inelastic, for reasons that are unclear. The result would be consistent with increasing productivity over the life of the contract caused by the servant’s growing older, gaining experience, and adapting to the American language, culture, and work environment. In addition, expected net earnings during the early part of the contract may have been low because of a higher probability of escape and death. The redemption market also measures the relative productivity of immigrant servants from various regions in Europe. The servant auction produced systematic relationships between contract lengths and European ports of embarkation. This relationship can be interpreted as an inverse measure of productivity among immigrants from various European ports. The productivity ranking of redemptioners from British ports, from high to low, was Ulster Irish, Southern Irish and Londoners, Dubliners and English from Bristol and Liverpool, and finally Scotsmen from the Isle of Lewis. The relative productivity of servants from Northern and Southern Ireland, as measured by the market, is more flattering than narrative accounts of their character and desirability.36

Conclusions The redemption system prospered for a hundred years despite published condemnation on both sides of the Atlantic. Some redemptioners were deluded by unscrupulous recruiters, some had their inflated expectations dashed in the redemption auction, and individual cases of fraud did occur. But immigrants continued to use the redemption system in large numbers. Letters from trusted Americans, circulated and published in Europe, provided them with reasonably accurate accounts of American wages, work, and environmental conditions, and descriptions of how the redemption system worked. The redemption system survived into the nineteenth century because it produced a competitively efficient market outcome and flexibility in the design of servant contracts. The disappearance of servitude among Europeans migrating to the United States after the 1820s cannot be explained by desires to avoid an exploitative system. Search restrictions and the precarious bargaining position redemptioners experienced upon arrival in America were overcome by competition among American employers. Quantitative evidence indicates that redemptioners were paid their opportunity cost and that servant productivity was priced through compensating adjustments in contract length. The systematic relationship between contract length and various productivity signals, contract restrictions,

Auction of German immigrant servants, 1771–1804   271 and market conditions indicates that redemptioners had considerable latitude in negotiating their contracts. Redemptioners even contracted around the auction search restrictions which reduced their opportunity to find friends or relatives who would pay their passage debts. Many negotiated a contingency buy-­out clause into their contracts, which stated that if the servant paid a specific sum within a stated time period, the contract would be void. With the exception of slavery and convict labor, the potential and opportunity for exploitation in an American labor market may never have been greater than in the redemption system. Therefore, the competitive efficiency, lack of exploitation, and increasing complexity of labor contracts found in the redemption trade may provide a measure of the general character of American labor markets. In labor markets, experimentation and innovation in institutional arrangements that survived competitive alternatives, such as the redemption system, may have been generally beneficial rather than detrimental to the worker.

Notes   * Originally published as Farley Grubb, “The auction of redemptioner servants, Philadelphia, 1771–1804: an economic analysis,” Journal of Economic History, vol. 48, no. 3 (September 1988): 583–603. Copyright 1988. © 1988 by Cambridge University Press and the editors of The Journal of Economic History. Reprinted with permission. The Journal of Economic History, Vol. XLVIII, No. 3 (Sept. 1988). © The Economic History Association. All rights reserved. ISSN 0022–0507. The author is Assistant Professor of Economics, University of Delaware, Newark, DE 19716. Earlier drafts were presented at the Economic History Seminar, University of Pennsylvania, 1985; the Economics Seminar, University of Delaware, 1986; the Econometric Society Meetings, New Orleans, 1986; the Economics Seminar, University of North Carolina-­Chapel Hill, 1987. I wish to thank the participants of these seminars and Gary Becker, Stanley Engerman, Robert Fogel, David Galenson, Robert Gallman, Henry Gemery, Claudia Goldin, Saul Hoffman, Ken Koford, Ken Lewis, Robert Margo, Thomas Mroz, Richard Steckel, the editor, and the many anonymous referees for helpful comments on earlier drafts. I also wish to thank the City Archives of Philadelphia and the Pennsylvania Historical Society for providing microfilm of manuscript contract evidence.   1 See Farley Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, 22 (July 1985): 316–39. Farley Grubb; “Redemptioner Immigration to Pennsylvania: Evidence on Contract Choice and Profitability,” Journal of Economic History, 46 (June 1986): 407–18; Marcus L’ee Hansen, The Atlantic Migration 1607–1860 (Cambridge, MA, 1940): 102–6; Cheesman A. Herrick, White Servitude in Pennsylvania (Philadelphia, 1926): 254–66.   2 For recent studies of the market for indentured servants, see David W. Galenson, White Servitude in Colonial America (Cambridge, MA, 1981); Farley Grubb, “The Market for Indentured Immigrants: Evidence on the Efficiency of Forward-­Labor Contracting in Philadelphia, 1745–1773,” Journal of Economic History, 45 (Dec. 1985): 855–68; Farley Grubb, “Colonial Labor Markets and the Length of Indenture: Further Evidence,” Explorations in Economic History, 24 (Jan. 1987): 101–6.   3 For descriptions of the redemptioner system see Galenson, White Servitude: 13–15; Karl F. Geiser, Redemptioners and Indentured Servants in the Colony and Commonwelath of Pennsylvania (New Haven, 1901): 5–76; Grubb, “Redemptioner Immigration,” 407–18; Abbot E. Smith, Colonists in Bondage (New York, 1947): 3–42. The two most cited eyewitness accounts of the redemption auction are in Gottlieb

272   German immigrant servitude in America, 1745–1835 ­ ittelberger, Journey to Pennsylvania in the Year 1750 and Return to Germany in the M Year 1754, Oscar Handlin and John Clive, trans. and eds. (Cambridge, MA, 1960): 16–18, and in letters written by Henry Melchior Muhlenberg reprinted in Frank R. Diffenderffer, “The German Immigration into Pennsylvania Through the Port of Philadelphia, and ‘The Redemptioners’,” Pennsylvania German Society, 10 (1899): 189–93.   4 See the assessment in John J. McCusker and Russell R. Menard, The Economy of British America, 1607–1789 (Chapel Hill, 1985): 242–4.   5 Quoted in Donald F. Durnbaugh, “Two Early Letters from Germantown,” Pennsylvania Magazine of History and Biography, 84 (Apr. 1960): 231–3. Sauer became a respected Pennsylvania-­German printer. Previously, scholars have traced the redemption system only as far back as 1728, see A. E. Smith, Colonists in Bondage: 21; Diffenderffer, “German Immigration into Pennsylvania,” 172–3, 200–1. Advertisements for the sale of German immigrant servants in the American Weekly Mercury in Philadelphia indicate that the redemption system had not been adopted by German immigrants in 1722. Therefore, the redemption system appears to have begun sometime between 1722 and 1724.   6 See newspaper accounts reproduced in Diffenderffer, “German Immigration into Pennsylvania,” 200–1.   7 See the discussion in Diffenderffer, “German Immigration into Pennsylvania,” 191–4; Geiser, Redemptioners and Indentured Servants: 64–8. For an example of restrictive search clauses written into redemptioner transportation agreements see the passage contract for the ship Commerce reproduced in Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, 1934), vol. 3: 131–4.   8 See the discussions in William Eddis, Letters from America (Cambridge, MA, 1969): 39–40; Frederic Trautmann, “Pennsylvania Through a German’s Eyes: The Travels of Ludwig Gall, 1819–1820,” Pennsylvania Magazine of History and Biography, 105 (Jan. 1981): 40–1. The speed of sale was derived from the evidence cited in Tables 14.1–14.3.   9 See also the discussions in Diffenderffer, “German Immigration into Pennsylvania,” 170–216, 238–55; Galenson, White Servitude: 15; Geiser, Redemptioners and Indentured Servants: 56–70; A. E. Smith, Colonists in Bondage: 20–2, 39–41. 10 Reprinted in I. D. Rupp, History and Topography of Northumberland, Huntington, Mifflin, Centre, Union, Columbia, Juniata and Clinton Counties, Pa. (Lancaster, 1847): 57. For examples of similar denunciations in later years of the redemption trade see Henry Melchior Muhlenberg, The Journals of Henry Melchior Muhlenberg, Theodore G. Tappert and John W. Doberstein, trans. (Philadelphia, 1942), vol. 2: 157; Eddis, Letters from America: 38–40; Henry Bradshaw Fearon, Sketches of America (3rd edn., London, 1819): 148–51. 11 In addition, competition among shippers would cause passage debts to equal the cost of transportation, see Farley Grubb, “The Market Structure of Shipping German Immigrants to Colonial America,” Pennsylvania Magazine of History and Biography, 111 (Jan. 1987): 27–48; Grubb, “Redemptioner Immigration,” 411–18. 12 Derived from the sources cited in Tables 14.1–14.3 and The Pennsylvania Gazette 1728–1789, 25 vols. (reprinted, Philadelphia, 1968). 13 Mittelberger, Journey to Pennsylvania: 16–18. For the geographic distribution of servant purchasers in the Delaware Valley, see Farley Grubb, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­ Century Mid-­Atlantic Economy,” Social Science History, 9 (Summer 1985): 249–75. 14 The 1785 to 1804 servant document was designed only to record German redemptioner contracts. The 1771 to 1773 servant document recorded indentured, redemptioner, and resident servants, see Farley Grubb, “Servant Auction Records and Immigration into the Delaware Valley, 1745–1831: The Proportion of Females Among Immigrant Servants,” Proceedings of the American Philosophical Society,

Auction of German immigrant servants, 1771–1804   273 132 (Sept. 1988), forthcoming. Redemptioners were distinguished from other servants in the 1771 to 1773 records by their immigrant status and by the distinction of not being “assigned” to a master by a current owner. Unlike all other servants, redemptioners were not the property of anyone when they entered the servant auction, and they assigned themselves to their chosen master. 15 See Grubb, “Colonial Labor Markets,” 102, for a discussion of the secular decline in contract lengths for indentured servants. 16 Within the 1787 to 1804 sample, controlling for age, the secular rate of decline in contract length was 3 percent for every five years (see Table 14.3). This estimate, however, is only marginally significant. 17 See the discussions in Diffenderffer, “German Immigration into Pennsylvania,” 141–315; Galenson, White Servitude: 13; Mittelberger, Journey to Pennsylvania: 16–18; A. E. Smith, Colonists in Bondage: 20–2. 18 Accumulation of wealth may not have been significantly reduced by entering America as a servant. Average personal savings rates were around 15 percent in the early nineteenth century. The savings rate for the poorest segment of society was probably lower. See Donald R. Adams, Jr., “Earnings and Savings in the Early 19th Century,” Explorations in Economic History, 17 (Apr. 1980): 118–34; Donald R. Adams, Jr., “The Standard of Living during American Industrialization: Evidence from the Brandywine Region, 1800–1860,” Journal of Economic History, 42 (Dec. 1982): 914–15. 19 The yearly income of resident workers was estimated by multiplying the average daily wage rate by an estimated number of workdays per year. The average number of days a fully employed laborer could expect to work per year is not known. B. G. Smith, “Material Lives,” 188, suggests that “five of every six work days or ten of every twelve months” would be a reasonable estimate. The estimates of resident artisan incomes by B. G. Smith and Adams were substantially different. Because these different estimates could not be reconciled, both were reported in Table 14.1. Only resident wage rates that excluded room and board were considered. 20 Between 1793 and 1807, a few ship manifests recorded the occupations of German immigrants arriving in Philadelphia (see Strassburger, German Pioneers, vol. 3). Among 579 immigrants with recorded occupations, 58 percent were artisans or professionals, 26 percent were farmers, and 11 percent reported no trade. Because only the poorest 45 percent of the immigrants entered servitude, farmers and those without a trade may have predominated among redemptioners. The contract lengths and prices of 66 immigrants, who reported artisan trades and were traced in the servant records, were significantly lower than for the average servant reported in Table 14.1 (t-­statistic of 1.66 on the difference in average contract length, and t-­statistic of 2.63 on the difference in average contract price). The comparison indicates that the average redemptioner was probably not a skilled artisan. Differences in productivity and work amenities between immigrant servants and free resident laborers that could not be directly measured may explain any residual income differences. For example, free labor may have received a premium for uncertainty in the duration and variance of employment. By contrast, servants may have received a premium for surrendering to the employer the right to inflict moderate corporal punishment and the right to make labor-­leisure decisions at the margin. The net effect on relative compensation of these differing work amenities is uncertain. In addition, language and custom differences may have reduced immigrant productivity and thus lowered redemptioner relative compensation per reported skill level. 21 Quoted in Strassburger, German Pioneers, vol. 1: xxxvii; Diffenderffer, “German Immigration into Pennsylvania,” 192. Mittelberger, Journey to Pennsylvania: 16–18, gives a similar description of the redemption auction in 1750. 22 See Grubb, “The Market for Indentured Immigrants,” 855–68. 23 For a discussion of hedonic indices see, Sherwin Rosen, “Hedonic Prices and Implicit Markets: Product Differentiation in Pure Competition,” Journal of Political Economy,

274   German immigrant servitude in America, 1745–1835 82 (Jan. 1974): 34–55; Zvi Griliches, ed., Price Indexes and Quality Change (Cambridge, MA, 1971). 24 The evidence was constructed by matching the names of redemptioners in the servant sale records to the names of immigrants, ships, and captains in ship passenger lists and custom house records. Merging these data sources was necessary to separate immigrant servants from resident servants and to expand the information about each servant, such as determining the servant’s age, family structure, port of origin, literacy, and the time taken to form a contract after arrival. The matching procedure differed for the British and German samples and for the two sample periods, 1771 to 1773 and 1787 to 1804. Therefore, the model was estimated separately for the two time periods and the two ethnic groups. 25 The sample of German redemptioner servants from 1771 to 1773 used in Table 14.2 was constructed in the same manner as the evidence used in Robert O. Heavner, “Indentured Servitude: The Philadelphia Market, 1771–1773,” Journal of Economic History, 38 (Sept. 1978): 701–13; Robert O. Heavner, Economic Aspects of Indentured Servitude in Colonial Pennsylvania (New York, 1978): 48–101. Although constructed in the same fashion, the sample in Table 14.2 is much larger, includes more independent vari­ables, and differs in the number of contracts with education and training than the sample reported by Heavner. Heavner mistakenly identified the contracts in the sample as indentured servant rather than redemptioner. This mistake led Heavner to apply the wrong model to the evidence. He used contract price as the dependent variable and all other information as independent variables. Heavner interpreted this regression as measuring differences in servant human capital through systematic changes in contract prices. This error explains numerous puzzling results, such as insignificant coefficients on literacy, skill, and training, but a large price reduction for education, and lower compensation for Germans migrating from London compared with Rotterdam. Correctly interpreted, Heavner’s regression explains the variance in transportation debts rather than servant productivity. There is little reason to expect passage debts among adults to vary with literacy, skill, or training. The 43 percent lower passage debt for redemptioners asking for education may simply reflect the half-­fare charged children under age 16. Finally, lower passage debts for Germans sailing from London would be expected because passage fares from London were lower than from Rotterdam. 26 Indentured servant contracts were relatively uniform. Contract lengths were typically fixed in whole years with almost no variance in contract lengths among adults. Few indentured servants migrated in family units, and few indentured contracts contained stipulations other than customary freedom dues. 27 The surviving German passenger ship lists from 1771 to 1773 only consist of loyalty oath signatures. Loyalty oaths were only required of German immigrant males above age 16, see Strassburger, German Pioneers, vols. 1–2. Matching the names of German redemptioners with the names on the loyalty oaths produced a German servant sample of adult men, exclusively. This sample was used in Table 14.2. 28 Longer contracts associated with education and training may have been caused either by the lower productivity of these servants or by the cost of acquiring these marketable skills. The cost of education and training to the master was both out-­of-pocket expenses and the lost work time both of the servant and master. Education was stipulated in contracts either as a stated number of months of schooling or as being taught to read and write. 29 For example see, Galenson, White Servitude: 97–113. 30 Mittelberger, Journey to Pennsylvania: 18, stated, “The very young, between the ages of 10 and 15, have to serve until they are 21.” He also stated that children under age 5 “must stay in service until they are 21 years of age.” Table 14.3 indicates that Mittelberger’s claim was more of a rule of thumb for providing information to prospective emigrants than a mandatory or customary rule for assigning contract lengths to children.

Auction of German immigrant servants, 1771–1804   275 31 For example see, Galenson, White Servitude: 105–7; David Galenson, Traders, Planters, and Slaves (Cambridge, Mass., 1986): 53–114; Robert W. Fogel and Stanley L. Engerman, Time on the Cross (Boston, 1974): 75–7. 32 For a discussion of forecast error in the indentured servant market see, Grubb, “The Market for Indentured Servants,” 866–7; Grubb, “Colonial Labor Markets,” 102–3. 33 See, for example, Galenson, Traders, Planters, and Slaves: 71–92. 34 An alternative explanation would be that the sick postponed their sale, hoping to recover sufficiently from the arduous voyage to raise their labor value and thus shorten their contracts. This behavior would cause shorter contracts to occur early in the auction because healthy servants who could bargain for shorter contracts would choose to sell themselves first. See the description in Mittelberger, Journey to Pennsylvania: 16. Sickness, however, was not very important. Debarkation morbidity was under 3.5 percent among German passengers, see Grubb, “Morbidity and Mortality,” 573–7. 35 Purchasers who were listed as merchants were the exception to the general occupation-­contract length pattern. The merchant category was relatively broad and the kind of work expected of the servant is unclear. Merchants may have been wealthy enough to purchase household servants to do domestic chores. Thus merchants may have purchased younger, inexperienced servants who could not negotiate shorter contracts because of their lower productivity. 36 The productivity difference could be considerable. For example, Ulster Irish were able to negotiate 23 percent shorter contracts than Irish redemptioners from Dublin, and German redemptioners departing from London were able to negotiate 12 percent shorter contracts than Germans leaving through Rotterdam. All British ports were significantly different from one another, except for London versus South Irish ports. The partial F-­test was 2.49 for Ulster versus South Ireland (significant at the 0.11 level), 4.08 for Ulster versus London (significant at the 0.04 level), and 0.66 for South Ireland versus London (insignificant). Among indentured servants sold in Philadelphia, Ulster servants were also found to be the most productive, see Grubb, “Colonial Labor Markets,” 105.

15 Debt shifting within German immigrant families*

European migration to colonial and early republican America was expensive, easily exceeding half of a year’s, and frequently a whole year’s, income for many adult migrants. Unattached individuals had to bear the entire burden of financing their own migration. During the late eighteenth and early nineteenth centuries, roughly half of them entered servant labor contracts to do so—selling their future labor in America to finance their passage expense. Individuals migrating as part of a family, however, could shift the cost of their own migration onto other family members. Some family members could enter servant contracts that sold for more than that member’s share of the family’s passage debt. For example, parents could sell their children into bondage for a long enough term of service that would not only cover the children’s passage debts but also that of the parents—thus freeing the parents from any obligation to enter servitude themselves. Alternatively, a parent could take the debts of other family members upon him or herself, freeing these other members from servitude obligations.1 German parents have been singled out in the modern literature for financing their passage to colonial America by selling their children into servitude. One implication, often left unspoken and not directly addressed in the literature, is that these German parents were cruel and selfish, regarding their children as purely a salable capital asset with no consumption-­good utility or emotional familial value. An alternative implication, more often mentioned but not fully explored in the literature, is that these immigrants were at the mercy of callous, profit-­seeking, and perhaps even deceptive shippers, who trapped them in a predicament en route that forced them to sell their children into service to extricate themselves. Strangely, however, the benefits of this measure accrued to the parents and not to the shippers.2 In any case, neither the kinds of debt-­shifting strategies adopted within families to finance migration nor the reasons for their use are well documented or well explained in the literature. This chapter presents and evaluates the evidence for and against intra-­family debt shifting among German immigrants in early America, ultimately establishing its existence only on a limited basis. Shippers did not deceive parents into selling their children for the shippers’ benefit, and German parents were not abusive exploiters of their children. Intra-­family debt

Debt shifting within German immigrant families   277 shifting emerged from the relationship between profit-­oriented market competition and colonial welfare law. Debt shifting may well have been welfare-­ enhancing compared with the relevant alternatives.

German migration to america, 1720–1820 Germans comprised the largest non-­British European migration to British colonial America and the early United States. From 1720 to 1775, more than 108,000 Germans migrated to British America; 80,000 of them landed in Phil­adelphia. Between 1785 and 1820, another 21,000 to 25,000 Germans landed there. The key demographic characteristic distinguishing the German from the British migrants was the relatively high proportion of families among the Germans. For example, under 10 percent of English migrants to Pennsylvania between 1774 and 1776—the only years when comprehensive records were collected on British emigration to eighteenth-­century America—were either married persons or dependent children, whereas between 52 and 72 percent of German migrants to Pennsylvania between 1727 and 1820 were either married persons or dependent children. Because these German immigrants were also relatively literate—80 percent of the adult males able to sign their names—they were probably as well informed about their choices as any eighteenth-­century transatlantic migrants could be.3 This demographic difference finds confirmation in the 25,000 immigrant ser­ vants reported in the surviving emigrant-­servant registers for British ports— Bristol (1654–86), Middlesex (1683–84), Liverpool (1686, 1697–1707), London (1682–86, 1718–59), and all British ports (1773–5)—and from those in Philadelphia (1679–87, 1746, 1771–3, 1784–1804, and 1817–31). For example, between 1785 and 1804, when records were complete enough to identify the family status of all German immigrants who entered servitude, among the 7,837 German immigrants arriving in Pennsylvania, 35 percent of married people and 40 percent of dependent children entered servitude to pay their passage debts. Parents and dependent children comprised 51 percent of all German immigrants and 44 percent of German immigrants sold as servants. Out of 927 married couples, more than one-­third had three or more children. It is difficult, if not impossible, to find anything comparable in the British and Irish migration stream to America during this period. Hence, family-­migration strategies that potentially entailed intra-­family debt shifting via servitude was almost exclusively a German, rather than a British or Irish, phenomenon, and considerable scope existed for intra-family debt shifting given the number and size of German families in the migration stream.4 The German passenger trade from Holland to America used a three-­tier fare structure. Surviving ship contracts indicate that adults, those older than 14, or 12 on some ships, were charged a full “freight.” Children older than 4 but under 13 or 15 on some ships, were charged half-­freight, and those younger than 5 incurred zero freight. The term freight corresponded to the allotment of space and rations. Children charged half-­freight received half the space and rations as adults, and

278   German immigrant servitude in America, 1745–1835 children charged zero freight had to share their parents’ berth and their parents’ ration. To this freight charge, or base passage fare, immigrants faced other expenses—various fees for taxes, baggage, extra provisions, registration and inspection charges, and monies with interest advanced by the shipper to liquidate debts accumulated in the embarkation port before entering the shipping contract.5 From 1720 to 1820, German migrants used the redemption method—a promise to pay for passages shortly after arrival—to finance migration for those who did not have the cash yet. This method entailed the same three-­tiered passage fare structure described above but with a mark-­up of approximately 15 percent to cover interest and default costs. Emigrants could decide how to pay this debt, but were not allowed to disembark until they had done so. Most of them negotiated servant contracts while docked at their New World destination to repay their migration debt.6 The redemption method allowed German parents the flexibility to observe the servant market in America and decide after arrival how best to redistribute the family’s total passage debt among its individual members. Signing fixed-­length servant contracts designed to sell for an expected sum before embarkation, as British migrants typically did, would have been difficult for German families because of what economists call the “lemons” problem. Because Germans typically incurred a range of debts for food and shelter before embarkation that shippers added to the cost of passage, the length of their servant contracts could vary. But they could also vary because of differences in productivity. If contract lengths were fixed before embarkation, buyers in America would not know the reason for a given servant’s contract length. Risk-­adverse buyers would interpret longer contracts as a sign of low servant productivity rather than a sign of higher debt and so offer too low a price per contract length. The redemption method helped solved this problem by revealing the debt and the physical condition of servants to buyers at the American auction. But did German immigrant families actually take advantage of this opportunity for intra-­family debt shifting provided by the market?7

Evidence that German parents sold their children “as if they were cattle” The main evidence for the indictment of German parents as unfeeling child exploiters consists of four seemingly independent literary sources. Elder Johannes Naas, shortly after arriving in Pennsylvania in 1733, wrote a letter to his son in Switzerland describing how German immigrants used servitude to pay for their passage: “What I heard concerning the people who do not have money for the passage, surprised me greatly, . . . the child takes two freights (fare for two) upon itself, its own and that of the father or of the mother. . . . Small children often pay one freight and a half until they are twenty-­one years old.”8 In 1756, Gottlieb Mittelberger wrote a polemic against German emigration in which he described how German emigrants used servitude to pay for their passage:

Debt shifting within German immigrant families   279 Many parents in order to pay their fares . . . and get off the ship must barter and sell their children as if they were cattle; for if their children take the debt upon themselves, the parents can leave the ship free and unrestrained (italics added) . . . People who arrive without the funds to pay their way and who have children under the age of five, cannot settle their debts by selling them. They must give away these children for nothing to be brought up by strangers; and in return these children must stay in service until they are twenty-­one years old. Children between five and ten who owe half-­fare . . . must also go into service in return until they are twenty-­one years old, and can neither set free their parents nor take their debts upon themselves. On the other hand, the sale of children older than ten can help to settle a part of their parents’ passage charges.9 In 1756, Benjamin Franklin wrote a letter to Sir Everard Fawkner in which he presented reasons why a recent act of Parliament would ruin the servant trade. This act authorized military officers to enlist servants in exchange for paying the prime cost of a servant minus a sum equal to the proportion of time already served. Franklin argued that this price was too low, in part because “many . . . Servants [were] purchased young of their Parents, who coming with large Families, [bound] some of their Children to Tradesmen and Farmers, in order to raise a Sum to pay the Freights of the whole, and keep themselves free.”10 Similarly, Henry Melchoir Mühlenberg, head of the Lutheran Church in Pennsylvania, in a letter sent to Germany in 1763 that described how German immigrants used servitude to pay for their passage to America, wrote, Children under six are gratuitously disposed of . . . [Other] . . . children . . . are sold, and the parents’ fare charged to the children’s account, and the children are consequently obliged to serve a longer time. Children are in this way not infrequently separated forever from their parents. . . . By having their children sold, parents are allowed to leave the ship. . . . Little did the parents anticipate such things. . . . The children of poor parents, if kept in hardship, learning that because of the non-­sale of father or mother they have to serve the longer, often became incensed, yea even embittered against their own parents.11 Six stylized facts can be taken from this literary evidence: (1) German immigrant parents shifted their passage debts onto their children (Naas, Mittelberger, Franklin, and Mühlenberg); (2) children younger than 15 had to serve until age 21 (Naas, Mittelberger); (3) children younger than 5 were given away (sold for a zero price) into servitude (Mittelberger, Mühlenberg); (4) older children, those charged full freight, were sold for two full freights, thus freeing a parent (Naas, Mittelberger); (5) children between the ages of 5 and 10, who were charged half-­ freight, could not be used to help pay their parents’ passage debts because they could not be sold for more than their freight (Mittelberger); (6) children charged half-­freight and sold for one and a half freights helped to pay their parents’ passage debts (Naas). Stylized fact 1, though true, is far more limited in scope

280   German immigrant servitude in America, 1745–1835 than these writers imply. Stylized fact 2 is roughly correct, but stylized facts 3, 4, 5, and 6 are erroneous.12 There are reasons for questioning these writers’ veracity. Mittelberger’s commentary was openly intended as a polemic to dissuade Germans from migrating to America and thereby to ingratiate Mittelberger to the Duke of Wurttemberg. Franklin’s account was intended as rhetorical persuasion in order to justify altering an act of Parliament. With regard to motive, the accounts of Naas and Mühlenberg appear less impeachable. They were written in personal letters with no apparent ulterior motive other than to inform friends and relatives in Germany of their observations in America, though Mühlenberg may have been angling for funds to build an orphanage. None of those writers directly participated in the German servant trade or appear to have directly observed the auctions of German servants for any extended period of time or in any detail. Naas even opened his description with the words, “What I have heard.” To what extent their accounts are based on hearsay rather than observation is unknown.13

Evidence to the contrary The only two surviving (published) German eyewitnesses to German immigrant servant sales who themselves were sold into servitude were Johann Carl Büttner and John Frederick Whitehead [Johann Friedrich Kukuck] who arrived at Philadelphia in 1773. Neither mention the selling of children by their parents. Christopher Sauer, a German printer who migrated to Pennsylvania in 1724 and served as an advocate for German immigrants regarding troublesome aspects of the trade, does not say anything about it either. That these writers would fail to comment about this kind of debt shifting within families is curious.14 Of greater relevance is the quantitative analysis of four bodies of servant auction records on German immigrant arrivals in Pennsylvania for the years 1746, 1771–3, 1785–1804, and 1817–31. They contain, in total, individual sale prices for 5,907 German servants who arrived on well over 200 vessels carrying German passengers. The numeric ages of the servants were seldom recorded in these records. Only for a small sample in the 1787–1804 record can a numeric age be matched to servant sales from the ages listed on a few ship manifests. Overall, roughly 25 percent of the 5,907 total were dependent children of adult migrants. These servant auction records indicate that no German servant was sold for a zero price (that is, was given away gratis); almost no German servant sold for what can clearly be considered two full freights (two typical adult passage debts); and very few German servants sold for what can clearly be considered one-­and-a-­half full freights (1.5 times the typical adult passage debt). As such, this evidence contradicts stylized facts 3, 4, and 6 above, made by Naas, Mittelberger, and Mühlenberg, and also casts doubt on stylized fact 1, their core claim that German parents shifted their passage debts onto their children who were traveling with them.15 The absence in the auction records of German servant contracts with a zero price or no price recorded is not due to magistrates failing to record zero-­price contracts. Over 1,000 contracts with no (zero) price were recorded in the 1746

Debt shifting within German immigrant families   281 and 1771–3 records. These contracts were largely for non-­immigrant apprentices and servants. Likewise, that the absence of any such German servant contracts in these records was due to a lack of German immigrant children younger than 5 is difficult to believe given the preponderance of dependent children among German immigrant servants. Children under age 5 were charged a zero freight price for passage. Either these children did not enter servitude or, if they did enter it, they must have sold for a positive price and entered into contracts with a length of 17 or more years, bringing them to the age of 21. Yet, very few contracts of that length exist. Of the 48 contracts for German servants in the 1746 record, none has that length. Only three of the 1,315 contracts for Germans in the servant-­auction records of 1771–3 are that long, and only another three that ran for one year less. Similar results hold for the 3,511 Germans in the auctions from 1785–1804 and the 1,033 Germans in them from 1817 to 1831.16 What happened to all of the children under age 5? First, many German parents entering servitude negotiated contracts whereby children younger than 5, who were charged zero freight cost, were allowed to stay with them and become free at the same time. For example, the servant contract for Johan Georg and Regina Geiser included the clause, “their infant child to be fed & clothed gratis & to be free with the parents.” Second, some of these children may have avoided servitude by living with American relatives outside the formal servant-­auction market—in which case their placement went unrecorded by the magistrates. In any event, the servant-­auction evidence from 1746, 1771–3, 1785–1804, and 1817–31 indicates that children under 5 were not given away to strangers. Either the servant market had radically changed after Mittelberger and Mühlenberg made their claims in 1756 and 1763, respectively—without notice or comment from elsewhere—or Mittelberger and Mühlenberg were wrong. They may have simply confused the zero freight charge with a zero servant price for these children.17 The four bodies of servant-­auction records reveal that only one German servant sold for approximately two full adult freights, 19 pistoles, in 1746. In other words, stylized fact 4 above is erroneous. The sole German servant out of the 48 in the record of 1746 who sold for close to two full freights does not appear to have been a dependent child, since no “consent” to contract was recorded, as it was for the other children in this record. In the 1771–3 servant-­ auction record, the typical German adult passage debt (sale price) was around 20 to 25 Pennsylvania pounds, though it could range as high as 30 or more. Büttner (age 19) and Whitehead (age 16) wrote in their memoirs that they sailed to Philadelphia from Rotterdam in 1773 with no other family members or accompanying friends (apparently they were not friends either), and that neither had relatives nor personal connections already in America. Büttner’s individual accumulated passage debt upon arrival, and thus his servant contract price in Phil­ adelphia, was 30 pounds in Pennsylvania currency; Whitehead’s was 28 pounds, 13 shillings, and six pence. The typical price paid for a married German servant in this record was 20 to 25 Pennsylvania pounds, and the average passage debt for passengers charged a full freight on the Britannia (Tables 15.1 and 15.2),

16.50 16.50 16.50 8.25 8.25 8.25 8.25 82.50 16.50 16.50 16.50 8.25 57.75 16.50 16.50 16.50 49.50 16.50 16.50 16.50 8.25 8.25 8.25 74.25

Jung family Parents: Wilhem Anna Maria Son: Nicholas Adam (estimated age = 16–17) Son: Anthony (estimated age = 12–13) Son: John Hendrick (estimated age = 10–11) Son: Carel Daughter: Maria Family total

Schott family Parents: Lawrence Anna Son: Hans Peter (estimated age = 14–15) Son: Jacob (known age = 11) Family total

Linck family Parent: Johan Simon Son: Johan Jurgen (estimated age = 18–19) Son: Johan Phillip (estimated age = 20+) Family total

Reinhard family Parents: Hans Salomea Daughter: Salomea (estimated age = 14–15) Son: Hans Jurgen (estimated age = 11) Son: Jocchem (estimated age = 10) Son: Hans Family total 17.68 17.68 19.86 8.93 8.93 8.93 82.01

21.31 21.34 21.34 63.99

19.49 17.68 17.68 9.43 64.28

17.68 17.68 17.68 8.93 8.93 8.93 8.93 88.76

12.20 0.00 29.03 18.50 24.00 0.00 83.73

22.10 22.10 22.10 66.30

18.15 0.00 26.85 21.00 66.00

“paid by children” “paid by children” 20.00 20.00 25.00 25.00 0.00 90.00

Passage debt Total debt Sale price

“paid by himself” “paid by children” 9/21 9/21 9/23 “paid by other children”

5.50 8.50 11.00

3.00 3.75 3.00

5.50 10.00

9/21 9/20

9/21 9/20 9/21

3.00

4.00 6.00 9.00  ?

Contract length (years)

9/22

9/24 9/23 9/20 ?

Sale date

Table 15.1 Seven families on the ship Britannia arriving from Rotterdam at Philadelphia, September 18, 1773

16.50 16.50 16.50 16.50 8.25 74.25 (blank) 16.50 16.50 49.50

Wanner family Parents: Jacob Maria Magdalena Son: Johan Georg (estimated age = ?) Daughter: Maria Catharina (estimated age = 16) Daughter: Eva Salomee (estimated age = 14) Family total

Beuning family Parents: Philip Anna Maria Daughter: Maria Christiana (estimated age = 17) Family total 14.77 17.65 17.65 50.07

17.90 17.65 17.65 17.65 8.90 82.20

22.98 23.15 20.96 67.09

14.77 17.90 17.90 50.57

4.30 (blank) 26.85a 20.35 26.85 78.35

24.00 24.00 21.00 69.00

“received in Money” (purchased by husband) 9/20

9/21 9/20 9/28

“received of himself in past”

9/22 9/22 9/24

4.00

8.00 5.00 4.50

4.50 4.50 4.50

Notes The difference between sale price and total debt for the family totals represents either an exchange-rate error adjustment or an additional fee not listed. A question mark means that the data were not recorded in the “Record of Indentures.” All prices were converted to Pennsylvania pounds following the conversion rates in John J. McCusker, Money and Exchange in Europe and America, 1600–1775 (Chapel Hill, 1978), 9, 186. average excess gained by parents on children sold (sale price – total debt) Children charged full-freight (N = 9): (23.13 – 19.09) = 4.04 Pennsylvania pounds Children charged half-freight (N = 7): (22.91 – 9.00) = 13.91 Pennsylvania pounds average percentage mark-up gained by parents on children sold [(sale price – total debt)/total debt] Children charged full-freight (N = 9): (23.13 – 19.09)/19.09 = 21.2% Children charged half-freight (N = 7): (22.91 – 9.00)/9.00 = 154.6% a Johan Georg also contracted to receive 8 pounds cash at the end of his servitude. Because of this extra payment, contract length cannot be used reliably to estimate his age.

Sources: “Munstering Book” for families, passage debt, and total debt; “Record of Indentures” for sale price, sale date, and contract length; Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, Pa., 1934), I, 749–51, for families and estimated age; Grubb, “The Long-Run Trend in the Value of European Immigrant Servants, 1654–1831: New Measurements and Interpretations,” Research in Economic History, XIV (1992),” 171, for estimated age based on contract length.

16.50 16.50 16.50 49.50

Kunkele family Parents: Philip Barabara Son: Michael (estimated age = 16–17) Family total

Passage debt 16.50 16.50 16.50 8.25 8.25 8.25 74.25 16.50 16.50 8.25 8.25 49.50 16.50 16.50 8.25 8.25 8.25 57.75 16.50 16.50 8.25 41.25

Families

Knor family Parents: Christopher Anna Dorothea Daughter: Maria Catharina Daughter: Anna Dorothea Daughter: Margareta Barbara Son: Johan Coenaad Family total

Beuning second family Parent: Anna Barbara Son: Johan Diederick (estimated age = 14–15) Daughter: Charlotta Margareta (estimated age = 11) Daughter: Susanna Catharina Family total

Martz family Parents: Simon Anna Maria Son: Johan Philip (estimated age = 11) Daughter: Elizabeth Daughter: Anna Margareta Family total

Hees family Parents: Augistusus Maria Elizabeth Daughter: Anna Margareta Family total

22.80 22.70 10.95 56.45

20.40 17.65 8.90 8.90 8.90 64.75

24.20 22.35 12.95 14.60 74.10

19.60 19.10a 25.10b 12.15b 12.15b 12.15b 81.15

Total debt

21.45 18.95 19.20 59.60

  2.40 “15.64 received of him in the past” (blank) 22.00 (contract length = 8.50 years) 15.50 (blank) 55.54

24.45 27.00 (contract length = 6.00 years) 20.00 (contract length = 10.17 years)   3.75 75.20

“22.33 paid by his children”a “drowned at sea” 25.25 12.40 22.33 22.33 82.31

Stated amount received

Table 15.2 Six more families on the ship Britannia arriving from Rotterdam at Philadelphia, September 18, 1773

16.50 16.50 8.25 8.25 16.50 66.00

Deegen family Parents: Jacob Anna Elizabeth Son: Daniel Son: Philip Jacob Daughter: Catharina Elizabeth Family total 18.30 17.65 8.90 8.90 17.65 71.40

22.80 17.65 8.90 8.90 58.25

8.35 “cash received, the sum charged his children” 28.30 18.10 17.90 “paid by her brothers” 72.65

44.45 (blank) 18.00 (blank) 62.45

notes The difference between sale price and stated amount received for the family totals represents either an exchange-rate error adjustment or an additional fee or payment not listed. All prices were converted to Pennsylvania pounds following the conversion rates in John J. McCusker, Money and Exchange in Europe and America, 1600–1775 (Chapel Hill, 1978), 9, 186. average excess gained by parents on children sold (sale price – stated amount received) Children charged full-freight (N = 2): (26.13 – 23.73) = 2.40 Pennsylvania pounds Children charged half-freight (N = 10): (18.78 – 10.49) = 8.29 Pennsylvania pounds average percentage mark-up gained by parents on children sold [(sale price – stated amount received)/stated amount received] Children charged full-freight (N = 2): (26.13 – 23.73)/23.73 = 10.1% Children charged half-freight (N = 10): (18.78 – 10.49)/10.49 = 79.0% a This value is not included in the family total. b This value includes an item called “paid by cash for debt of Elders her/his proportion.”

Sources: “Munstering Book” for families, passage debt, total debt, and stated amount received; “Record of Indentures” for contract length; Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, Pa., 1934), I, 749–51, for families and estimated age; Grubb, “The Long-Run Trend in the Value of European Immigrant Servants, 1654–1831: New Measurements and Interpretations,” Research in Economic History, XIV (1992),” 171, for estimated age based on contract length.

16.50 16.50 8.25 8.25 49.50

Kunkele second family Parents: Johan Georg Anna Margareta Son: Johan Philip Daughter: Catharina Family total

286   German immigrant servitude in America, 1745–1835 which arrived in Philadelphia in 1773, was 20 Pennsylvania pounds. No German servant of the 1,315 total in the servant-­auction record of 1771–3 sold for more  than 44 Pennsylvania pounds; only two sold for 40 to 44 Pennsylvania pounds.18 In the 1785–1804 servant-­auction record, adult German males migrating alone sold for $81.00 on average. No German servant among the 3,511 in this record sold for twice that amount. In addition, the average prices of dependent children are equal to, or less than, this amount. The same pattern is found in the 1817–31 servant-­auction record. Thus does the quantitative evidence indicate that almost no family member took two full adult passage debts onto him/herself—their own and another family member’s full freight debt. Again, unless the servant market changed drastically after 1733 and 1756, respectively, the claims of Naas and Mittelberger were wrong.19 Few German servants in these auctions sold for one-­and-a-­half full freights, leaving stylized fact 6 in serious doubt. In the 1746 servant-­auction record, two German servants out of 48 sold for around 1.5 times the typical adult price of 10 to 12 pistoles—15 pistoles. In the 1771–3 servant-­auction record, only 7 of the 1,315 German servants were sold for between 35 and 40 Pennsylvania pounds and only another 36 sold for over 30 but under 35 Pennsylvania pounds,  about 1.5 times the typical individual adult German passage debt of 20 to 30 Pennsylvania pounds. Similar results are found in the 1785–1804 and the 1817–31 servant-­auction records. Barring another undocumented change in the market, Naas’ claim of 1733 was at worst false or at best greatly exaggerated.20 Finally, although evidence in the servant-­auction records suggests that at least some children charged half-­freight were sold for that amount and not more as Mittelberger claimed (stylized fact 5), the numbers so sold seem few relative to the roughly 25 percent of German immigrant servants who were dependent children. For example, in the 1746 servant-­auction record, one German servant out of 48 sold for 4 pistoles, about half of the typical adult price of 10 to 12 pistoles. In the 1771–3 record, 64 of the 1,315 German servants sold for what would appear to be half the typical adult passage debt, 12 Pennsylvania pounds or less. The records for 1785–1804 and 1817–31 show similar proportions. Hence, Mittelberger’s claim in 1756 has no better validity than the others investigated above. Overall, intra-­family debt shifting along the lines described by Naas, Mittelberger, Franklin, and Mühlenberg cannot be found in the 1746 to 1831 servant-­auction records; German parents do not appear to have sold their children “as if they were cattle” for personal gain.21 In fact, the quantitative evidence shows that German parents, far from being callous or selfish, were concerned about their children’s welfare. The timing of servant sales within an auction indicates that children were sold before their parents. This pattern is consistent with parents trying to sell their children for sums above their children’s passage debts in hopes of raising enough cash to defray some of their own (the parents’) passage debt. However, the prorated sale prices of parents are not below the range of prices for adults who migrated alone.

Debt shifting within German immigrant families   287 This timing of child sales appears, instead, to have been designed to maximize parents’ supervision of their children’s servant contracts. For example, parents negotiated complex education clauses into them.22 The auction evidence for 1785–1804 can be organized by family structure. Figure 15.1 shows the probability that a dependent child would enter servitude conditional on the number of siblings migrating with the family, and on whether the child’s parents entered servitude or remained free. Although children clearly went into servitude while their parents remained free, the likelihood was relatively low—an only child having a 10 percent chance and a child with three siblings a 39 percent chance. The probability of entering servitude increased with the number of siblings, suggesting that large families did not have the resources to pay for the passage of all members. Some members had to pay their own passage through entering servitude—typically the oldest children, who had already reached the age of entering apprenticeship. This evidence leaves little room for a systematic, large-­scale purchasing of parental freedom via the selling of their children into servitude.23 100 90

Percentage who entered servitude

80 70

All two-parent households where the parents entered servitude Number of children = 639

60

All two-parent households Number of children = 2,185

50 40 30

All two-parent households where the parents remained free Number of children = 1,546

20 10 0

Probability of entering servitude conditional on a child’s number of siblings 0

1

2

3

4

5

6+

Number of siblings

Figure 15.1 Servitude among German immigrant children, 1785–1804 (sources: derived from “Book of Redemptioners”; Grubb, “The Incidence of Servitude in Trans-Atlantic Migration, 1771–1804” Explorations in Economic History, XXII (1985), 324; Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, Pa., 1934), III).

288   German immigrant servitude in America, 1745–1835 Children were more likely to enter servitude if their parents did, but the probability of them doing so was far from 100 percent. An only child had a 42 percent chance, whereas a child with three siblings had a 63 percent chance. One reason for this gap was that many parents entering servitude negotiated contracts that allowed children under age 5, who were charged zero freight cost, to stay with them and share their eventual freedom. This is hardly callous, selfish behavior. Even though the versions of debt shifting described by Naas, Mittelberger, Franklin, and Mühlenberg find little support in the servant-­auction records, another avenue for intra-­family debt shifting not mentioned by these writers remains possible. The servant-­auction evidence rules out any substantial parental debt shifting onto older children who were charged full freight for passage or onto very young children who were charged zero freight for passage. Children under age 13 (or 15 on some ships) but older than age 4 were charged half-­ freight, though a seemingly small number of German servants were sold for half the typical adult passage debt. In addition, since very few German servants were sold for 1.5 times the typical adult debt, those children charged half-­freight were not sold for their own plus an additional whole adult’s passage debt. Thus, by logical deduction, a substantial number who were charged half-­freight were sold for roughly twice their own passage debt, or the equivalent of approximately one full adult passage debt. Because age and freight charges were not recorded in the servant auction records, children charged half-­fare but sold for full-­fare would appear indistinguishable from adult servants, thus masking this intra-­family debt shifting in the records. The ages of a small sample of German servants can be found by matching their names in the 1785–1804 servant-­auction records with those on surviving ship manifests. The evidence is consistent with the logical deduction above. Children aged 10, 11, 12, 13, and 14 were sold for the same price as adults. Additional direct evidence, however, would be helpful in confirming that this particular pattern of child servant sales—children charged half-­freight being sold for full-­fright—was the dominant family debt-­shifting strategy.24

A unique conjunction of evidence: the ship Britannia Most surviving German passenger manifests from the colonial and early national periods recorded only the signatures (oaths of allegiance) of adult German male passengers and, on occasion, the names and ages of the other passengers as well. Of the few surviving German passenger ship manifests that recorded itemized freight charges for each passenger, only the manifest of the Britannia provides a complete itemization of charges by individual and notations on how these charges were paid. The names on it can be matched with those in the surviving servant-­auction records in Philadelphia. No other ship record permits this kind of linkage to other surviving documents.25 The Britannia, captained by James Peters, left Rotterdam on an unknown date, cleared customs in Cowes, England, on August 13, 1773, and arrived in

Debt shifting within German immigrant families   289 Philadelphia on September 18, 1773. The Pennsylvania Gazette reported that Captain Peters had 246 Germans with him. The surviving oath of allegiance administered to the passengers in Philadelphia upon debarkation reported “150 in the list” and “250 freights.” The oath of allegiance was administered only to non-­British males over age 15. It recorded 118 signatures. Joshua Fisher and Sons, to whom the passengers were consigned, placed an advertisement in the Pennsylvania Gazette stating, “Just arrived . . . A number of healthy German passengers whose freights are to be paid to Joshua Fisher and Sons or to the master on board.”26 The mustering book for the Britannia lists 260 individuals: 210 whole freights, 45 half-­freights, and five quarter-­freights, including six adults who died on the voyage. Interpreting and quantifying this information is difficult because of poor penmanship, cryptic terminology, and variable currency for the charges (florins, Pennsylvania pounds, and pounds sterling). Itemized charges for each passenger included “freight,” “head money,” “proportion in the bed place,” “cash advanced,” “interest,” “sea stores,” “clothes at sea,” and “baggage.” Remarks recorded in the mustering book are suggestive of intra-­family debt shifting. For example, next to some passengers’ debts were recorded the phrases “paid by his brother” or “paid by his children.” However, this information does not indicate how the debts were paid—by servitude, by sale of goods that a family brought, by gifts from friends and relatives in America, or by some other method. Although some entries list the names of persons who paid the passengers’ debts, why these people paid the debts and what they received in exchange is not recorded in the mustering book.27 The 1771–3 servant-­auction record ends seventeen days after the Britannia’s arrival—enough time for ninety-­six individuals to complete their sale into servitude to pay their passage debts. Their auction prices can be matched to their passage debts, and the amounts paid toward them, in the ship’s mustering book. For the most part, but not in all cases, the mustering book’s record of money received matched the amount paid in the servant-­auction record. The most reli­ able information comes from the sample of passengers on the Britannia that could be matched to the servant-­auction record, not only because of the independence of the two records but also because of the servant-­auction records’ ability to confirm what was ultimately paid for each individual passenger who entered servitude. The difference between the auction price and the passage debt, when organized and totaled by family, indicates both the existence and the magnitude of intra-­family debt shifting.28 Twenty-­three families with children sailed on the Britannia. Given that usually just under half of German family units entered servitude to pay for their passage, the expectation would be that at least eleven of the families on the Britannia would have done so. As it turns out, the records provide fairly complete information on how thirteen families on the Britannia used servitude to pay their way. Seven family units that had completed their sale into servitude before the records end could be individually matched to servant-­auction sale prices. This matched evidence is presented in Table 15.1. Six more family units appear to

290   German immigrant servitude in America, 1745–1835 have used servitude in part to pay for their passage, but cannot be matched by individual names to the servant-­auction record. Most likely they entered servant contracts sometime after the servant-­auction record ended. Statements in the mustering book about money received from a third party to pay a passenger’s debt for the most part matches this passenger’s sale price in the auction records when available. Thus, the information in the mustering book on the amount received from a third party, when compared to the total debt, can indicate both the existence and the magnitude of intra-­family debt shifting, though slightly less reliably than for the seven families that also appear in the servant-­auction record. This evidence is presented in Table 15.2.29 In Table 15.1, the passage debt is the freight charge, and the total debt is the freight charge plus all additional charges and pre-­embarkation debts added to it. The sale price is the cost of an individual’s servant contract and, along with the sale date and contract length, comes from the servant-­auction record. With the exception of one child, age was not recorded. Children’s ages, however, can be estimated in several ways. The mustering book lists children in descending order by age for each family, and children under 15 were charged half-­freight. Children who signed the oath of allegiance had to be older than 15. Because of the relationship between age and contract length, contract length can serve as a proxy for age. The estimated ages of children are included in both Tables 15.1 and 15.2.30 The pattern of debt shifting within these families in both Tables 15.1 and 15.2 is clear. Children older than 14 who were charged full-­freight passage fares rarely generated much excess value when sold into servitude. The exceptions would appear to be children aged 14 and 15 who had just barely crossed the full-­ freight threshold—for example, Hans Peter Schott and Salomea Reinhard in Table 15.1 and Johan Diederick Beuning in Table 15.2. The required contract length to take them to the age of maturity, 5½ to 6 years, was long enough to generate excess value above their full-­freight passage debt, but only in the range of a 21 to 46 percent mark-­up. For children older than 15, the shorter contract length needed to take them to the age of maturity, 3.8 years on average, reduced the excess value to near zero. Parents did not force these children to serve longer than was needed to pay for their own individual cost of passage. Only when the contract length did not take them to the age of maturity was it extended to that age, permitting parents to extract the value in excess of a child’s passage debt. Children under 15 who were charged half-­freight were the key to intra-­family debt shifting. They were sold into servitude for much more than their passage debt—an average mark-­up of 155 percent in Table 15.1 and 79 percent in Table 15.2. Children in the narrow age range of 9 through 13 were old enough to have high net labor value until the age of maturity, but they were charged only half-­ freight passage fares. The excess sums that their servitude brought paid the debts, and guaranteed the freedom, of other children and of the parents. But why did parents subject this age group, rather than the others, to debt shifting, and why did shippers fail to capture the excess value generated by these children by charging higher passage fares?

Debt shifting within German immigrant families   291

How colonial poor-­relief law constrained parental choice Because society did not want children to die of exposure and starvation or to commit mischief, adults with means and authority were required to look after them. Children’s poor-­relief was largely privatized in early America. The law governing the children of indigent parents followed that governing orphaned children. Overseers of the poor placed orphaned children and children of indigent parents in the homes of private citizens to be raised. The rough modern-­day analog would be foster care, except that the colonial master was to teach the child a skill, as in an apprenticeship, though the skill could be nothing more than doing farm chores. A child was expected to work for his master to pay for his keep. This practice was ubiquitous in the English world and not likely to surprise anyone entering it. As Murray noted, “The practice of binding poor children into new families extended back to the earliest English poor laws and was carried forward into colonial law and practice. Such public apprenticeship was widespread. . . . As in England, overseers of the poor . . . had the authority to seize children without their parents’ consent.”31 Pennsylvania’s poor relief statute of 1705 states, it shall and may be lawful for the said Over-­seers of the poor . . . to set on work the children of all such, whose parents can not be by the said justice thought able to maintain them, and also to put such children out apprentices, for such terms as they in their discretion see fit. This law did not fundamentally change for the rest of the eighteenth century.32 But did this poor-­relief law apply to foreign immigrants? In 1756 the Pennsylvania legislature formally recognized the de facto application of the poor-­ relief law to non-­British immigrants in “An Act for Binding Out and Settling Such of the Inhabitants of Nova Scotia Imported into This Providence as are Under Age.” This Act states, [T]he overseers of the poor . . . [can] bind out such of the children of the said inhabitants of Nova Scotia, whose parents or friends are incapable to maintain or neglect otherwise to provide for, to kind masters or mistresses . . . that the said children shall be taught to read and write the English language and such reputable and profitable occupations as may enable them to support themselves at the expiration of their respective indentures, the male children until they attain the age of twenty-­one years and the female children until they attain the age of eighteen years respectively and no longer.33 The Pennsylvania legislature also formally recognized the de facto application of the basic tenets of the poor-­relief law to German immigrant children in the supplement to the 1750 “Act for the Prohibiting the Importation of Germans or Other Passengers in Too Great Numbers in Any One Vessel,” which was passed in 1765. It recognized the past practice of not allowing German

292   German immigrant servitude in America, 1745–1835 immigrant children who entered servitude to be freed before the age of maturity and, if necessary, forced them to work past that age: [I]f any child or minor shall at the time of his or her said importation be such an advanced age that his or her service until he or she arrives to the age of twenty-[one] years shall not be equal and sufficient to pay and discharge the money due for his or her own freight or passage, then and in such case he or she shall and may be bound to serve for the same until the age of twenty-­four years and no longer, unless the said money can be raised by his or her service for a shorter time.34 Immigrant parents who entered servitude were indigent in that they had exhausted all their resources and still could not pay their passage debts. Hence, these parents could either sell their own children into servitude—supervising their child’s contractual arrangements—for whatever the market would pay, or risk having their children bound out by the overseers of the poor for no upfront compensation to themselves or their children. Furthermore, if parents sold a child into servitude for a term that expired before the child reached maturity, the overseers of the poor had the authority to extend that child’s servitude to the age of maturity, again without compensation. Given that children faced the same outcome in either case—service at least until the age of maturity—parents were wise to sell their own children into bondage rather than let someone else capture their value. Whether German immigrants knew the content and application of colonial welfare law in all its detail at embarkation is unknown. A tremendous amount of information in personal letters and commercial correspondence flowed across the Atlantic between the German communities in America and in Europe. What has survived is probably only a small fraction of what was sent. No letter or broadside used in this study explicitly informed prospective German emigrants about how Pennsylvania’s poor-­relief law would affect them and their children as indigent servant immigrants. Several accounts, however, dealt with what happened to children in Pennsylvania who arrived orphaned. Since the law regarding the disposal of children of indigent parents typically followed that for orphans, these accounts had some relevance to the fate of indigent children as well.35 For example, in a letter sent from Pennsylvania to Germany in 1749, intended as advice on migration for prospective emigrants in Germany, Leonard Melchior commented, The children whose parents have died [on the voyage] have been distributed here and there. . . . Many times a person . . . has promised to care for the children like a father, and afterwards sold them for the sake of gain, so that our gracious authorities would not have interested themselves in the cause of these poor orphans if the sight of their misery had not been intolerable. In a letter of 1752 sent from Philadelphia to his superiors in Germany, Mühlenberg explained that “poor, ignorant, underage worms [children orphaned at sea]

Debt shifting within German immigrant families   293 are given away every year to all sorts of folks, corrupt sects and unbelieving people and are sold at a low price until they reach adulthood.”36 In a similar vein, the Wertheim County government (in Germany) circulated a decree in 1752 to its parishes, meant for prospective emigrants to America, which read, Those with little or no property . . . must then put up with being sold as wretched slaves, to earn the money for their costly passage . . . so that parents and children, man and wife, have to travel many miles from each other in service [as servants in Pennsylvania], indeed often never see, much less speak with each other again their whole life long. . . . Follow this warning for the welfare of yourself and your poor, innocent children!37 This limited available correspondence—along with the accounts of Naas, Mittelberger, Franklin, and Mühlenberg—suggests that prospective German emigrants who intended to arrive in America as servants may well have had some idea of how colonial poor-­relief laws would influence choices regarding their accompanying children. Wokeck concluded, “Most knew from letters to Germany or from traveling newlanders about the procedures for settling up and leaving the ship.” Finally, at debarkation in Philadelphia, providing such information would have been in the interest of the shippers, the German Soci­ eties of Pennsylvania and Maryland, and previous German immigrants.38 The reaction of German immigrant servants to the constraints that colonial poor-­relief laws imposed on them at debarkation explains the debt shifting within the families in this study. Selling children older than 14 into servitude until the age of maturity typically generated prices equal to their passage fare. If it generated less, then a child might be sold for a term that extended past age 21. The sale of children aged nine to 14 until maturity generated money in excess of their passage cost that families could use to defray the passage cost of other family members. For most children younger than nine, however, the net labor value of serving until maturity did not generate market prices above their passage debts, leaving their parents little extra to gain from selling them into bondage.39 Children younger than five were almost never bound out by their parents, and those younger than nine were seldom bound out by the overseers of the poor, who bound out only youths who could earn their keep, roughly those over age eight. This circumstance may explain why so few German immigrant children under age five appear in the servant records. Since they were not under threat from the overseers of the poor, German immigrant parents who entered servitude may have felt safe keeping these children with them.40

Two arbitrage puzzles in the market for bound child labor German immigrant children aged 9 to 14, when required to serve to maturity, sold for 23 Pennsylvania pounds, which was 14 Pennsylvania pounds above their half-­freight passage cost (see Table 15.1). Why were parents able to capture this

294   German immigrant servitude in America, 1745–1835 excess value rather than shippers? Second, the overseers of the poor typically bound out indigent children at a zero price. If German immigrant children and local indigent children of the same age had similar human-­capital attributes, then employers were earning excess profits equal to 23 Pennsylvania pounds on local indigent children. How did employers compete for, and arbitrage, these excess profits? Shippers did not charge nine- to 14-year-­old German immigrant children full freight because competition prevented them. Passage fares were based on freight space. Children under 15 used only half the allotted space, and ate only half the rations, given to adults. If a shipper tried to charge these children more than half­fare, families could have canvassed for competing shippers who would charge the prorated space and provision cost for a half-­ling. Shippers competing for cargo would readily offer such half-­fare prices for children under 15 to fill their cargo holds with family groups. In the end, parents made the determination that their children under 15 would require only half the resources of adults. Parents could have paid full freight for young children, but competition among shippers allowed them to minimize the total family cost of passage.41 Sorting out the second arbitrage puzzle is more difficult. Three solutions are possible. First, German immigrant children of the same age and the same contract length as local indigent children may have sold for higher prices because they had superior human-­capital attributes. Most likely, German immigrant children did not have superior human capital to local children per se. Adult American-­born and British-­born servants were taller than German-­born servants in colonial Pennsylvania, and German children were not likely to have English-­ language skills upon debarkation. The market for German immigrant servants was also not ethnically segmented. More buyers with non-­German names purchased German servants than did buyers with German names. This trend held true for the Britannia, as it did for Büttner and Whitehead (both purchased by English Quaker farmers), both of whom explained the language issues in their respective memoirs. Mittelberger referred to it when he said that everyday “Englishmen” went on board German passenger ships to buy servants.42 But German immigrant child servants may well have been preferable to local indigent children who fell into the hands of the overseers of the poor. The presence of differential “selectivity” between the two markets may have produced differences in human capital that explain the observed price difference. If local indigent parents had already privately bound out their children who could command a positive price, then the overseers of the poor would be “selectively” left binding out the indigent children who could obtain only a zero-­asset price. As such, the local children with a positive price may not have shown up in the formal servant records, whereas all German immigrant children bound out to service would have. The second possible solution to the arbitrage puzzle involves contract benefits. Assuming similar human-­capital attributes between the two groups, local indigent children of the same age and contract length may have sold for less than German immigrant children because they received more within-­contract benefits.

Debt shifting within German immigrant families   295 Analysis of the contracts of the two groups reveals that the contracted provisions of food, shelter, clothing, freedom dues (end-­of-contract payments), and education were basically the same. The only difference was the provision of occupational training. Few German immigrant children received such a contract clause. For example, only 5 percent of the 794 German immigrant children entering servitude between 1787 and 1804 had occupational-­training clauses explicitly in their servant contract. Immigrant servitude was not apprenticeship. Local indigent youths, however, were bound out to a trade by law—the institution typically described as pauper apprenticeship. If the contractual provision of employer-­provided occupational training was the only difference between these two groups of children, then the price difference of 23 Pennsylvania pounds represents the market-­revealed measure of the net opportunity cost of employer-­ provided occupation training. This finding is important because there are no other measures of the opportunity cost of training alone in the apprenticeship contracts of eighteenth-­century America.43 The above explanation has a serious caveat. The overseers of the poor did not necessarily bind out local indigent youths to traditional craft trades. Of the boys bound out by the Philadelphia overseers of the poor, 25 percent were to be trained in “husbandry,” and approximately 50 percent were bound to masters outside of Philadelphia City—suggesting that serious craft training may have been unlikely. By comparison, the occupational and geographical distribution of employers who purchased German immigrant children looks remarkably similar to that for local pauper apprentices. For example, 18 percent of German immigrant children bound out between 1787 and 1804 went to farmers, and the rest to tradesmen, such as millers, bakers, tanners, and glaziers, or to proprietors in the service trades, such as merchants, innkeepers, and storekeepers. More than half were bound to masters who lived in Philadelphia City.44 If occupational training was mostly learning by observing and learning by doing, then the actual training of pauper apprentices and German immigrant children may not have been too different. The training clauses in pauper apprentice contracts versus the absence of training clauses in the servant contracts of German immigrant children may have simply reflected legal formalities. This scenario is consistent with the general decline in formal apprenticeship training via explicit contract observed in America relative to Europe.45 The above possibility leads to the third possible solution to the arbitrage puzzle: Employers competed for the 23 Pennsylvania pounds in excess profit earned on local indigent youth versus German immigrant children either by providing under-­ the-table kickbacks or by providing other nonpecuniary favors to the overseers of the poor. Given that, on average, the overseers of the poor apprenticed only 14 boys and 10 girls per year in Philadelphia, one of the largest cities in the English-­ speaking world at the time, favoritism practiced by the overseers would not have caused an excessive competitive disadvantage to, nor have incurred excessive protest by, employers that they did not favor with cheap child labor.46 Scholars should always be wary of taking either literary evidence or quantitative evidence at face value; both can mislead. Naas, Mittleberger, Franklin, and

296   German immigrant servitude in America, 1745–1835 Mühlenberg were correct in claiming that intra-­family debt shifting existed among German immigrants to early America, but they were wrong about its extent, nature, and cause. Yet the quantitative evidence on servant sales masks the debt shifting that was occurring, making it easy to overlook. Literary and quantitative evidence need to be combined with analytical tools and deductive analysis—including an appreciation of the interplay between markets and legal institutions—to discover their full meaning. The greed or malevolence of the shipping industry or the market for indentured servants was hardly to blame for the debt shifting that occurred. Nor was parental cruelty or selfishness responsible for it. Parents did not typically force their children under age 16 into contracts that required them to serve beyond the age of maturity, nor force their children over 15 into contracts long enough to generate excess value above that child’s passage cost. That some parents went free by selling children between the ages of 9 and 14 into bondage was an unintended outcome of the benefits gained in a competitive shipping market that constraints imposed on them. The net outcome may also have been the best that their world could achieve. If poor German immigrant parents were not allowed to use their young teenage children’s servant contracts to pay for their migration costs, they would have been forced to leave their children behind in Germany or to remain behind in Germany themselves. Either option was likely to have been worse—particularly in light of their actual choice to migrate. Leaving teenage immigrant children unsupervised in America while their parents labored in bondage may also have been an unacceptable alternative for these families, not to mention for society. All things considered, colonial welfare laws may have yielded a socially optimal outcome by encouraging debt shifting within German immigrant families.

Notes   * Originally published as Farley Grubb, “Babes in bondage? Debt shifting by German immigrants in early America,” Journal of Interdisciplinary History, vol. 37, no. 1 (Summer 2006): 1–34. Copyright 2006. © 2006 by the Massachusetts Institute of Technology and the editors of The Journal of Interdisciplinary History. Reprinted with permission. Farley Grubb is Professor of Economics, University of Delaware, and Research Associate, National Bureau of Economic Research. He is the author of German Immigrant Servant Contracts Registered at the Port of Philadelphia, 1817–1831 (Baltimore, 1994); “The Circulating Medium of Exchange in Colonial Pennsylvania, 1729–1775: New Estimates of Monetary Composition, Performance, and Economic Growth,” Explorations in Economic History, XLI (2004), 329–60; “German Immigration to Pennsylvania, 1709 to 1820,” Journal of Interdisciplinary History, XX (1990), 417–36. Preliminary versions were presented at the XIIIth Congress of the International Economic History Association, Buenos Aires, Argentina, 2002; the Economic History Society annual meeting, Durham, UK, 2003; and the Social Science History Association annual meetings, Baltimore, MD, 2003. The author thanks the participants of these seminars and Howard Bodenhorn, George Boyer, Walter Kamphoefner, David Mitch, John Murray, Jean Elliot Russo, and James Watkinson for helpful comments.

Debt shifting within German immigrant families   297 Georgianna Daugherty, Anne Pfaelzer de Ortiz, and George Ward provided research assistance, and Anne Pfaelzer de Ortiz provided editorial assistance. Financial support provided by a 2003–4 American Philosophical Society Sabbatical Fellowship Grant and resource support provided by Harvard University Department of Economics are gratefully acknowledged. © 2006 by the Massachusetts Institute of Technology and The Journal of Interdisciplinary History, Inc.   1 Grubb, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, XXII (1985), 316; idem, “The End of European Immigrant Servitude in the United States: An Economic Analysis of Market Collapse, 1772–1835,” Journal of Economic History, LIV (1994), 818–19; Günter Moltmann, “The Migration of German Redemptioners to North America, 1720–1820,” in Pieter C. Emmer (ed.), Colonialism and Migration; Indentured Labour Before and After Slavery (Boston, 1986), 108–9; Abbot Emerson Smith, Colonists in Bondage (New York, 1947), 336; Marianna S. Wokeck, Trade in Strangers: The Beginnings of Mass Migration to North America (University Park, 1999), 162–3.   2 The literature on this topic notes the practice, sometimes condemning or regretting it, but it usually fails to follow the implications to their logical conclusion. See Bernard Bailyn, Voyagers to the West (New York, 1986), 185; Frank Reid Diffenderffer, “The German Immigration into Pennsylvania through the Port of Philadelphia, and ‘The Redemptioners,’ ” Pennsylvania German Society, X (1899), 191–293, 266, 302–3; Karl F. Geiser, Redemptioners and Indentured Servants in the Colony and Commonwealth of Pennsylvania (New Haven, 1901), 52–3, 73–4; Cheesman A. Herrick, White Servitude in Pennsylvania (Philadelphia, 1926), 187, 191; Moltmann, “Migration of German Redemptioners,” 111; Robert B. Morris, Government and Labor in Early America (Boston, 1981), 321–2; Smith, Colonists in Bondage, 22; Alan Taylor, American Colonies (New York, 2001), 319; Wokeck, Trade in Strangers, 157, 162–3; Klaus Wust, The Virginia Germans (Charlottesville, 1969), 105–6.   3 Bailyn, Voyagers, 210–11; Aaron S. Fogelman, Hopeful Journeys: German Immigration, Settlement, and Political Culture in Colonial America, 1717–1775 (Philadelphia, 1996), 1–12, 61–5; idem, “From Slaves, Convicts, and Servants to Free Passengers: The Transformation of Immigration in the Era of the American Revolution,” Journal of American History, LXXXV (1998), 43–74; Hans-­Jurgen Grabbe, “European Immigration to the United States in the Early National Period, 1783–1820,” Proceedings of the American Philosophical Society, CXXXIII (1989), 192; Grubb, “Colonial Immigrant Literacy: An Economic Analysis of Pennsylvania-­German Evidence, 1727–1775,” Explorations in Economic History, XXIV (1987), 63–76; idem, “German Immigration to Pennsylvania, 1709 to 1820,” Journal of Interdisciplinary History, XX (1990), 421, 428–31; idem, “End of Immigrant Servitude,” 818–19; Wokeck, Trade in Strangers, 45–6, 50–1, 162–3, 172–3.   4 “Book of Redemptioners, 1785–1804 and 1817–1831,” unpub. ms.—microfilm copy obtained from the Historical Society of Pennsylvania, Harrisburg; Grubb, “Incidence of Servitude,” 320, 322; idem, “Redemptioner Immigration to Pennsylvania: Evidence on Contract Choice and Profitability,” Journal of Economic History, XLVI (1986), 407–18; idem, “The Long-­Run Trend in the Value of European Immigrant Servants, 1654–1831: New Measurements and Interpretations,” Research in Economic History, XIV (1992), 167–240; John J. McCusker and Russell R. Menard, The Economy of British America, 1607–1789 (Chapel Hill, 1985), 243; George W. Neible (ed.), “Servants and Apprentices Bound and Assigned Before James Hamilton Mayor of Philadelphia, 1745,” Pennsylvania Magazine of History and Biography, XXX (1906), 348–52, 427–36; XXXI (1907), 83–102, 195–206, 351–67, 461–73; XXXII (1908), 88–103, 237–49, 351–70; “Record of Indentures of Individuals Bound Out as Apprentices, Servants, etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771, to October 5, 1773,” unpub. ms.—microfilm

298   German immigrant servitude in America, 1745–1835 copy obtained from the City Archives of Philadelphia; Michael Tepper (ed.), Emigrants to Pennsylvania, 1641–1819 (Baltimore, 1978), 6–18. Only a few cases of married couples with dependent children can be found among British and Irish immigrant servants in this period. For example, only 30 of the 1,609 English, Irish, and Scots immigrant servants who landed in Philadelphia from October 1771 to October 1773 were listed as married, and only a couple of them appear to have had dependent children with them.   5 Susan E. Klepp, Grubb, and Anne Pfaelzer de Ortiz (eds.), Souls for Sale: Two German Redemptioners Come to America—The Life Stories of John Frederick Whitehead and Johann Carl Büttner (University Park, 2006); Otto Langguth, “Pennsylvania German Pioneers from the County Wertheim,” Pennsylvania German Folklore Society, XII (1947), 259–61; Ralph B. Strassburger, Pennsylvania German Pioneers (Norristown, PA, 1934), III, 131; Tepper, Emigrants to Pennsylvania, 257; Wokeck, Trade in Strangers, 86–91.   6 Grubb, “Redemptioner Immigration”; idem, “Market Structure of Shipping German Immigrants to Colonial America,” Pennsylvania Magazine of History and Biography, CXI (1987), 27–48; idem, “The Auction of Redemptioner Servants, Philadelphia, 1771–1804: An Economic Analysis,” Journal of Economic History, XLVIII (1988), 583–603; idem, German Immigrant Servant Contracts Registered at the Port of Philadelphia, 1817–1831 (Baltimore, 1994); Smith, Colonists in Bondage, 20–2, 40; Strassburger, Pennsylvania German Pioneers, III, 112–14, 131–2, 137–8; Tepper, Emigrants to Pennsylvania, 257–60.   7 See “Adverse Selection” and “Asymmetric Information,” in John Eatwell et al. (eds.), The New Palgrave: A Dictionary of Economics (New York, 1987), I, 32–4, 133–5; George Akerlof, “The Market for Lemons,” Quarterly Journal of Economics, LXXXIV (1970), 488–500; Grubb, “The Market for Indentured Immigrants: Evidence on the Efficiency of Forward-­Labor Contracting in Philadelphia, 1745–1773,” Journal of Economic History, XLV (1985), 855–68; idem, “Redemptioner Immigration,” 407–11; idem, “Contract Labor and the Indenture System,” in Joel Mokyr (ed.), The Oxford Encyclopedia of Economic History, I–V (New York, 2003), I, 535–8; Grubb, “Indentured Servitude,” in Steven Durlauf et al. (eds.), New Palgrave Dictionary of Economics and the Law (New York, forthcoming; orig. pub. 1998); Wokeck, Trade in Strangers, 126, 162–3. Whether some other truth-­revealing mechanism could have been devised to solve the “lemons” problem under the indenture method is unclear. Given that the market had few repeat participants, developing such a mechanism would have been difficult.   8 Diffenderffer, “German Immigration,” 300–3.   9 Gottlieb Mittelberger, Gottlieb Mittelbergers Reise nach Pennsyvanien im Jahr 1750 und Rukreise nach Teutschland im Jahr 1754 (Stuttgart, 1756); idem (trans. Carl Theo. Eben), Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754 (Philadelphia, 1898), 26–8; idem (ed. and trans. Oscar Handlin and John Clive), Journey to Pennsylvania in the Year 1750 and Return to Germany in the Year 1754 (Cambridge, Mass., 1960), 17–18. The italicized phrase is missing in the Handlin and Clive translation—the most commonly used translation—but appears in the translation by Eben and in the original 1756 German version. 10 Leonard W. Labaree (ed.), The Papers of Benjamin Franklin (New Haven, 1963), VI, 474. 11 Diffenderffer, “German Immigration,” 191–3. 12 On the correctness of stylized fact 2, see Grubb, “Long-­Run Trend,” 174, 178–9. 13 Mittelberger should not be regarded as a neutral observer reporting on everyday events. First, his voyage carried more passengers, took longer, and experienced higher mortality than most others. Compare Grubb, “Morbidity and Mortality on the North Atlantic Passage: Eighteenth-­Century German Immigration to Pennsylvania,” Journal of Interdisciplinary History, XVII (1987), 565–85; idem, “Market Structure of

Debt shifting within German immigrant families   299 Shipping,” 47; Mittelberger, Journey; Strassburger, Pennsylvania German Pioneers, I, 445. Second, his traveler’s account was encouraged and probably edited by Wurttemberg authorities to discourage emigrants. Third, his return to Germany was precipitated less by his self-­proclaimed disillusionment than by the loss of his position as organist due to a sexual offense—casting doubt on the general integrity of his travel account. Andreas Brinck notes that Mühlenberg, Mittelberger’s sponsor, wrote that because of Mittelberger’s misconduct, he considered him “unworthy to touch the organ in the church or instruct the children in the school” (Brinck, Die deutsche Auswanderungswelle in die britischen Kolonien Nordamerikas um die Mitte des 18. Jahrhunderts [Stuttgart, 1993], 96). I thank Walter Kamphoefner for pointing out this information and providing a rough translation of it into English. Regarding Mühlenberg’s attempt to generate funds to start an orphanage, see John W. Kleiner and Helmut T. Lehmann (eds. and trans.), The Correspondence of Heinrich Melchior Mühlenberg, II: 1748–1752 (Camden, ME, 1997), 203–24. 14 Diffenderffer, “German Immigration”; Klepp et al. (eds.), Souls for Sale; Christopher Sauer, “Letters to the Governor of Pennsylvania, 15 March, 12 May, and 12 June 1755,” Historical Magazine, IV (1860), 100–4. Another surviving, but unpublished, eyewitness description of German immigrant servant sales in Philadelphia by a German immigrant sold into servitude is in Fogelman, Hopeful Journeys, 77–9, 196—a letter from Maria Barbara Kober to relatives in Germany in 1767. Fogelman does not mention her saying anything about parents’ selling children into servitude in order to pay their passage debt. Kober’s one child died during the voyage. 15 “Book of Redemptioners”; Grubb, “Incidence of Servitude,” 320; idem, “Long-­Run Trend,” 171; idem, German Immigrant Servant Contracts; Neible, “Servants and Apprentices”; “Record of Indentures.” German immigration peaked between 1749 and 1754. Wokeck, Trade in Strangers, 45–6. Unfortunately, no servant-­auction records have survived for those years. Whether servitude on average functioned differently regarding the sale of children during those years compared with the years for which auction records exist is unknown. 16 “Book of Redemptioners”; Grubb, “Long-­Run Trend,” 174; idem, German Immigrant Servant Contracts; Neible, “Servants and Apprentices”; “Record of Indentures.” 17 Grubb, German Immigrant Servant Contracts, 85. Additional examples can be found on 2/7/1772, 2/11/1772, 12/10/1772, 1/19/1773 in “Record of Indentures,” in “Book of Redemptioners,” and in idem, German Immigrant Servant Contracts. 18 Klepp et al. (eds.), Souls for Sale; Neible, “Servants and Apprentices”; Pfaelzer de Ortiz, “German Redemptioners of the Lower Sort: Apolitical Soldiers in the American Revolution,” Journal of American Studies, XXXIII (1999), 281–3; “Record of Indentures.” 19 “Book of Redemptioners”; Grubb, “Long-­Run Trend,” 171; idem, German Immigrant Servant Contracts; idem, “End of Immigrant Servitude,” 818–19. 20 “Book of Redemptioners”; Grubb, German Immigrant Servant Contracts; Neible, “Servants and Apprentices”; “Record of Indentures.” 21 Book of Redemptioners”; Grubb, “The Incidence of Servitude,” 320; idem, German Immigrant Servant Contracts; Neible, “Servants and Apprentices”; “Record of Indentures.” 22 Grubb, “Educational Choice in the Era Before Free Public Schooling: Evidence from German Immigrant Children in Pennsylvania, 1771–1817,” Journal of Economic History, LII (1992), 363–75. 23 Grubb, “Incidence of Servitude.” One strategy would be for the children to go into service and the parents to stay free and earn enough cash to buy their children back. For example, in October 1819, Anna Margaret Weller was sold into six years of servitude with her father’s consent for $50.00 while he remained free. In February 1821, he bought her back for $25. See Grubb, German Immigrant Servant Contracts, 106, 124. This strategy, however, appears to have been rare.

300   German immigrant servitude in America, 1745–1835 24 Grubb, “Long-­Run Trend,” 171. 25 Grubb, “German Immigration”; Herrick, White Servitude, 186–7; “Munstering [mustering] Book for the Britannia Capt. James Peters from Rotterdam 3rd July 1773,” unpub. ms., AM 209, Historical Society of Pennsylvania, Harrisburg, Penn.; “Redemptioners, Philadelphia, 1750–1830,” unpub. ms., Miscellaneous Collection, Box 7a, Folder 7, Historical Society of Pennsylvania, Harrisburg, Penn.; “Record of Indentures”; Strassburger, Pennsylvania German Pioneers, I, 749–51; III, 112–14, 137–8; Tepper, Emigrants to Pennsylvania, 239; Wokeck, Trade in Strangers, 99–100, 102, 153–63. Other scholars have noted this ship manifest but have not systematically linked the names to the servant-­auction records and quantitatively analyzed its information. 26 Pennsylvania Gazette, 22 Sept. 1773; Strassburger, Pennsylvania German Pioneers, I, 749–51. Wokeck, Trade in Strangers, 248, lists the Britannia as departing Cowes on July 24, 1773. 27 “Munstering Book.” See the reprint of a page from this record in Herrick, White Servitude, 186–7. The voyage death rate and time at sea for the Britannia are slightly less than average. See Grubb, “Morbidity and Mortality”; Wokeck, Trade in Strangers, 115–16. What accounts for the difference between the 246 passengers reported by the Pennsylvania Gazette and the 260 passengers (minus the six dead) listed in the mustering book is unclear. The discrepancy might simply reflect passengers discharged from the ship between when it docked and when the Gazette was published. The five passengers listed as being charged only one-­quarter freight were not children. They were charged adult “head money” and were able to sign their names. Why they were charged this odd amount is unknown—perhaps as compensation for services rendered. For example, Traugott Leberecht Buhze, who was charged one-­quarter-freight, was the only person on the ship listed as a physician. Since the mustering (debt account) book listed no “zero freight” passengers, that is, no children under age 5, it is likely that more than the book’s 260 passengers embarked at Rotterdam. The copy of the “Munstering Book” in Tepper, Emigrants to Pennsylvania, 236–8, is neither complete nor a direct transcription. Wokeck, Trade in Strangers, 248, reports a total of 292 as her “best guess” for the number of German passengers on this voyage that were landed in America. It seems high relative to other sources cited. “Head money” refers to the payment of fees required for inspecting the vessel and registering the immigrants upon arrival. See Wokeck, Trade in Strangers, 97–9. 28 “Record of Indentures.” 29 Grubb, “Incidence of Servitude,” 320. 30 Grubb, “Long-­Run Trend,” 171; Strassburger, Pennsylvania German Pioneers, I. 31 See John E. Murray, “Family, Literacy, and Skill Training in the Antebellum South: Historical-­Longitudinal Evidence from Charleston,” Journal of Economic History, LXIV (2004), 776, for the quotation. See also Murray and Ruth Wallis Herndon, “Markets for Children in Early America: A Political Economy of Pauper Apprenticeship,” Journal of Economic History, LXII (2002), 356–82; Christopher Tomlins, “Early British America, 1585–1830, Freedom Bound,” in Paul Craven and Douglas Hay (eds.), Masters, Servants, and Magistrates in Britain and the Empire, 1562–1955 (Chapel Hill, 2004), 123, 131–4, 143, 149. 32 William Clinton Heffner, History of Poor Relief Legislation in Pennsylvania, 1682–1913 (Cleona, Pa., 1913); Herrick, White Servitude, 106–12; Morris, Government and Labor, 384–5; Murray and Herndon, “Markets for Children,” 356–82; Murray, “Family, Literacy, and Skill Training,” 776; Ian M. G. Quimby, Apprenticeship in Colonial Philadelphia (New York, 1985), 99–117; Statutes at Large of Pennsylvania (Harrisburg, 1896–1976), II, 253; IV, 266–7; V, 278–80; VI, 246–7; VIII, 75–9. 33 Statutes at Large, V, 278–80. See also Wokeck, Trade in Strangers, 139. 34 Statutes at Large, VI, 437–8. See also Wokeck, Trade in Strangers, 139.

Debt shifting within German immigrant families   301 35 Regarding the surviving transatlantic correspondence between Pennsylvania and Germany in this period, see Rosalind J. Beiler, “From the Rhine to the Delaware Valley: The Eighteenth-­Century Transatlantic Trading Channels of Caspar Wistar,” in Hartmut Lehmann et al. (eds.), In Search of Peace and Prosperity: New German Settlements in Eighteenth Century Europe and America (University Park, 2000), 172–88; Donald F. Durnbaugh, “Two Early Letters from Germantown,” Pennsylvania Magazine of History and Biography, LXXXIV (1960), 231–3; Fogelman, Hopeful Journeys, 33, 63–4, 71–80, 143–5; John W. Kleiner and Helmut T. Lehmann (eds. and trans.), The Correspondence of Heinrich Melchior Mühlenberg Volume 1: 1740–1747 (Camden, ME, 1993); Langguth, “Pennsylvania German Pioneers,” 149–289; Hannah Benner Roach, “Advice to German Emigrants, 1749,” Pennsylvania Genealogical Magazine, XXII (1962), 226–37; Wokeck, Trade in Strangers, 18–35. 36 Kleiner and Lehmann (eds. and trans.), Correspondence, II, 215; Roach, “Advice to German Emigrants,” 236–7. 37 Langguth, “Pennsylvania German Pioneers,” 190–1. In a letter to the Governor of Pennsylvania in March 1755, Sauer complained, “when merchants binds some people together, whose families were obliged to die and is famished for want, and as a prisoner at the vessel is retained and forced to bind himself,—one for two or three, who are greatly indebted, and who, perhaps, pays his own debt, while others can’t—he is forced to go out of the country, and will go rather than go to prison” [italics from the original] (“Letters to the Governor,” 102). 38 See Wokeck, Trade in Strangers, 154, for the quotation. Fogelman, Hopeful Journeys, 63–4, 76–80, 132, 143–5; Louis P. Henninghausen, History of the German Society of Maryland (Baltimore, 1909); Harry Pfund, History of the German Society of Pennsylvania, 1764–1964 (Philadelphia, 1964); Erna Risch, “Immigrant Aid Societies Before 1820,” Pennsylvania Magazine of History and Biography, LX (1936), 15–33; Henry T. Virchaux, “Report to the German Society of Pennsylvania on the Ship April,” unpub. ms. held at the German Society of Pennsylvania, Philadelphia; Wokeck, Trade in Strangers, 68, 121–3, 127, 141–2, 148, 152, 154–5, 161. 39 Statutes at Large, VI, 437–8. 40 Howard Bodenhorn, “Just and Reasonable Treatment: Racial Differences in the Terms of Pauper Apprenticeship in Antebellum Maryland,” unpub. Working Paper, Economics Dept., Lafayette College, 2002, Table 1; Quimby, Apprenticeship, 108. 41 Grubb, “Redemptioner Immigration”; idem, “Market Structure of Shipping.” Why shippers used only the three-­tier, age-­related passage-­fare structure is unknown. For example, shippers might have charged 10-year-­olds half-­fare, 12-year-­olds three-­ fifths fare, and 14-year-­olds seven-­eighths fare. One possible explanation for why they did not is that the transaction costs and divisibility issues of going to a finer  age-­related tiered structure out-­weighed any potential gain. Keeping track at the   ship’s mess of who was to receive what fraction of a ration and measuring by eye  the various differences in ages and rations was difficult and costly, as well as a  source for potential friction, compared with a simple division in half. Likewise, dividing berths, part of the ship’s fixed structure, into complex proportions while trying not to place children into berths with unrelated adults (with the obvious problems), or having odd fractions of berth spaces go unfilled, may have been costly to arrange. 42 Grubb, “Lilliputians and Brobdingnagians, Statute in British Colonial America: Evidence from Servants, Convicts, and Apprentices,” Research in Economic History, XIX (1999), 172; Klepp et al. (eds.), Souls for Sale; Mittelberger, Journey; “Munstering Book”; “Record of Indentures.” 43 Grubb, “Auction of Redemptioner Servants,” 597; idem, “Educational Choice”; idem, German Immigrant Servant Contracts; idem, “Statutory Regulation of Colonial Servitude: An Incomplete-­Contract Approach,” Explorations in Economic History, XXXVII (2000), 42–75; Quimby, Apprenticeship.

302   German immigrant servitude in America, 1745–1835 44 Grubb, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­Atlantic Economy,” Social Science History, IX (1985), 262–3, 268–9; Quimby, Apprenticeship, 106, 110. 45 See Gillian Hamilton, “The Decline of Apprenticeship in North America: Evidence from Montreal,” Journal of Economic History, LX (2000), 627–64. 46 Quimby, Apprenticeship, 106–7.

16 Processing German servants at the port of Philadelphia, 1817–31 The documents*

Organized markets for German immigrant servants arriving in the U.S. contin­ ued after the War of 1812. When and why such markets ultimately disappeared is a mystery. While the largest peak in German immigration between 1755 and 1830 occurred during the years 1816 through 1819 (Chapter 2), the United States  Federal government did not start collecting passenger manifests until 1820. This lack of Federal documentation of immigration before 1820 has made investigating the end of immigrant servitude difficult. In addition, few sur­ viving port records and state documents before 1820 recorded names and other information about individual immigrants, and those that have survived are incomplete. The last systematic body of records documenting an organized American market in European immigrant servants is in the German immigrant trade to Pennsylvania. After the revolution, the Pennsylvania German Society convinced the state of Pennsylvania to keep a register of German immigrant servant con­ tracts. These books are the only systematic records of European immigrant ser­ vants arriving in the U.S. after 1784. “Book A of redemptioners” spans 1785 through 1804; “Book B of redemptioners” which must have spanned 1805 through November 4, 1817, apparently is lost, and “Book C of redemptioners” spans from November 5, 1817 through 1831. The contents of “Book C of redemptioners” are transcribed in Grubb (1994). The existence of “Book B of redemptioners” and some idea of its content can be taken from references to that record made in “Book C of redemptioners.” “Book C of redemptioners” is also one of the few existing documents that can be used to identify German immigrants who arrived between November 1817 and 1820. Because immigrant servant contracts were registered within 30 days of arrival, the registration date can serve as a proxy for the date of arrival. The register can also be used to reconstruct immigrant families, since many family members entered servitude together. In addition, the register locates the initial residence of these immigrants since the township, county, and state of the immi­ grant’s initial American employer were recorded. This contract register is an indispensible record for studying German immigration to America in the decade after the end of the War of 1812. Because the last organized markets for Euro­ pean immigrant servants in the U.S. were in the German trade, “Book C of

304   German immigrant servitude in America, 1745–1835 redemptioners” is also an indispensible record for studying the nature of immi­ grant servant contracting during the last years of the institution and for studying why European immigrant servitude disappeared. When and why German immigrant servitude in the U.S. ended cannot be understood without knowing the contents of “Book C of redemptioners” and seeing the documentary history that shows how German immigrant servants were legally and contractually processed from embarkation through debarkation after 1815. To this end, a detailed summary of the contents of “Book C of redemptioners” is provided first. The documentary history of each step of the  migration and servitude process is provided second. Examples of ship contracts, servant contracts, health inspection orders, sales advertisements, and  eyewitness descriptions of the servant auction from this post-­1815 era are given. These examples are taken from documents which are concurrent with, and involve the immigrants whose contracts appear in, “Book C of redemption­ ers.” This information lays the foundation for the analysis in the next two chap­ ters which determine when and why German immigrant servitude came to an end. “Book C of redemptioners” does not contain the complete transcriptions of the actual immigrant servant contracts, but only summarizations of the key elements in the contracts. The registrar typically recorded the immigrant servant’s name; the buyer’s name, occupation, township, county, state of residence; who gave the servant consent to contract if the servant was a minor; the length of servitude and the amount paid the shipper for the labor contract, called the “considera­ tion;” any special payments to the servant during or at the end of the contract; and any other non-­standard contractual stipulations (Grubb 1994). Based on the contract example given below, the registrar simply recorded the unique informa­ tion the contracting parties added (indicated in italics in the contract example given below) to the standard format or pre-­printed contract form (indicated by non-­italics in the contract example given below). While the parties typically used pre-­printed forms which left spaces open for the parties to add their unique handwritten information, complete handwrit­ ten contracts consistent with these pre-­printed forms were also used and accepted. Because few actual contracts have survived, contract registers are the key source of information on immigrant servitude. The following example of an actual contract was appended to the “Book C of redemptioners” contract regis­ ter. The handwritten information added by the contracting parties to the pre-­ printed form is indicated here by italics. The one piece of information atypically missing from this contract is the sale price, called “the consideration,” paid by the master for the contract. Contracts were typically a single page in length. This contract from 1824 is not structurally different from the German immigrant servant contracts reproduced in Figures 12.1 and 12.2 from 1767 and 1785, respectively.

Processing German servants, 1817–31   305 PHILADELPHIA. ss. This Indenture Witnesseth, That Susanna Herbster of her own free will, And consent of her Father Laurence Herbster, for the consideration Hereinafter mentioned, to be done and performed by Christian Schenck of the Northern Liberties, Shoemaker as also for other good causes she the said Susanna Herbster hath bound and put herself and by these Presents doth bind and put herself Servant to the said Christian Schenck to serve him his Executors, Administrators and Assigns from the day of the date hereof, for and during the full term of Six Years from thence next ensuing. During all which term the said Servant her said master his Executors, Administrators and Assigns faithfully shall serve, and that honestly and obediently in all things, as a good and faithful Servant ought to do. And the said Christian Schenck his Executors, Administrators and Assigns, during the said term shall find and provide for the said Servant sufficient Meat, Drink, Apparel, Washing and Lodging And also give her Six Months Schooling. And within the first three years of the Term of the Service of the said Susanna Herbster, to have her confirmed at the German Lutheran Church in Philadelphia, and attend the Lectures of the Minister, agreeable to the form of the said Church. And at the expiration of her term to have two complete suits of clothes, one thereof to be new, also a good Bed, Bolster Pillows & Blankets worth at least twenty five dollars. And for the true performance hereof, both the said parties bind themselves firmly unto each other by these Presents. In Witness whereof they have interchangeably set their Hands and Seals. Dated the nineteenth day of May A.D. one thousand eight hundred and twenty four Bound before her Susanna X Herbster mark (“Book C of redemptioners”) Pennsylvania State law regulated the immigration and sale of German immigrant servants (Herrick 1926: 293, 305–8). During the years covered by “Book C of redemptioners” these laws required the registration of all German immigrant servant contracts. The registrar was to be a special officer who was familiar with both the English and German languages. Unregistered contracts were invalid, and all concerned persons were entitled to copies or abstracts of the register. Masters were required to give German redemptioner servants who were minors six weeks of schooling for each year of service, and it was the duty of the registrar to insert this provision into the labor contract if such a statement was absent. Masters were also required to pay, at the end of the contract, freedom dues consisting of two complete suits of clothes, one of which was to be new, or  the equivalent.1 Husbands and wives who entered servitude were not to be

306   German immigrant servitude in America, 1745–1835 separated except by mutual consent. Unless they agreed to separate servant contracts, separating them would void the contract. Those servants bound to serve in Pennsylvania were not to be sold out of the state without the servant’s consent. Pennsylvania state law also required bills of lading to be given to all passen­ gers for all goods brought and stored by the immigrants in the hold of the vessel. Persons were to be put ashore with their goods at no additional charge beyond the price agreed on in Europe before embarkation. The ship’s captain was to give proper food and drink and to care for the passengers on his ship during a term of up to 30 days after arrival—during which time the passengers in debt for their passage could negotiate with buyers who boarded the ship as to the conditions which would cause their passage debts to be paid. Additional time after the 30 days was to be at the expense of the passengers and added to their passage debts (Herrick 1926: 293, 305–8). The captain was to have sick passengers removed and cared for on shore, unless they could not be moved from the ship, in which case they were to be  cared for on the ship. Passengers who died on the voyage or shortly after landing were to be reported to the authorities in Philadelphia within 15 days of arrival and a just account of their goods was to be given to the Registrar of Wills for the benefit of their heirs and creditors. Penalties levied on shippers for violating these regulations ranged from 50 to 100 dollars (Herrick 1926: 293, 305–8). A quantitative summary of the information in “Book C of redemptioners” is pro­ vided next. The first entry in the registry is on November 5, 1817, and the last entry is on December 1, 1831. The register lists 1,258 entries wherein 1,933 names of separate individual immigrants are recorded. The entries are concen­ trated in the years 1817, 1818, and 1819, when 370, 464, and 351 entries, respec­ tively, are recorded. Only 73 entries are recorded between 1820 and 1831. Within these entries 1,035 new immigrant servants are recorded. Other entries include the resale of 60 immigrant servants from one master to another master, the outright release of 58 immigrant servants by their masters, and the reassign­ ment of 192 immigrant servants from the master’s purchasing agent to the master proper. In most cases, each entry records only one immigrant servant contract. The exceptions are most often for married servants and their children. Their con­ tracts are sometimes recorded as separate entries and sometimes recorded together within one entry. A number of children of immigrant servants, while mentioned by name in their parents’ contract entries, were not subject to the con­ ditions of their parents’ servant contracts. Among the 1,035 new immigrant servants were 192 married persons, 249 single adult males, 187 single adult females, and 407 dependent children (those who required consent from an adult to contract). The average contract length for single adult males was 2.7 years, and for married persons was 2.9 years. Given the movement in price indices during this period, the nominal contract sale prices (the passage debts to be paid) lacks meaning. The average deflated con­

Processing German servants, 1817–31   307 tract sale price (1795 = 100) was $100 dollars for single adult males and $74 for married persons (Chapter 18; Bezanson, Gray, and Hussey 1936: 392). Masters came from all walks of life. Some came from well-­respected stations and occupations. For example, Gottlieb Kintzle was sold to James Madison of the State of Virginia, Philipina Betz was sold to Joseph Barnes, a Philadelphia attorney, Frederick Troscher was sold to Richard Bache, a postmaster, Dorothea Werner was sold to John Geyer, a Philadelphia alderman, Barbara Schaffelin was sold to Louisa McIlhenny, a gentlewomen, Fredericca Kalender was sold to John Ross, a physician, Johan George Vogt was sold to John Biddle, a major in the U.S. army, Reginia Zahnin was sold to Augustus L. Baumfort, a professor of mathematics, Frederika Eitel was sold to Charles W. Bazeley, the keeper of a female seminary, and Dorothea Beck was sold to Robert B. Belville, a minister of the gospel (Grubb 1994: 9, 10, 16, 55, 61, 69, 86, 89, 116). Most masters, however, came from a middling station in society and employed those in the common trades. For example, in order from the most pre­ valent, farmers purchased 31 percent of the servants, merchants 19 percent, bakers 3 percent, innkeepers 3 percent, storekeepers 1.6 percent, victualers 1.4 percent, millers 1.3 percent, ironmasters 1.1 percent, grocers 1 percent, confec­ tioners 0.9 percent, papermakers 0.9 percent, mariners 0.8 percent, brewers 0.7 percent, and tanners purchased 0.7 percent of the servants. The rest of the masters were scattered among the numerous trades from brass founders and coopers to tailors and weavers. Masters also came from all parts of the country, including the western and southern territories. For example, 11 percent went to masters from Alabama, 2.3 percent from Missouri, 2.2 percent from Virginia, 2.1 percent from New York, 1.6 percent from North Carolina, and 0.9 percent went to masters from Illinois. One master even came from Cuba. Most masters, however, resided close to Phil­ adelphia. For example, 37 percent went to masters from Philadelphia county, 9 percent from Gloucester county (New Jersey), 6 percent from Lancaster county, 2.1 percent from Bucks county, 1.9 percent from Cumberland county, 1.7 percent from Berks county, 1.6 percent from Chester county, 1.6 percent from Lebanon county, and 1.6 percent went to masters from Burlington county (New Jersey)—all counties in eastern Pennsylvania or western New Jersey. The rest of the masters were scattered mostly across the other counties of Pennsylvania, New Jersey, and Delaware. Most masters, 74 percent, purchased only one servant, and another 16 percent purchased only two servants (often just a married couple). A few, however, made large purchases. The seven largest buyers were, in descending order, Benjamin B. Cooper of Gloucester county, New Jersey who purchased 63 servants; Nathan Farrows of Alabama who purchased 46 servants; Francis Clopper of Alabama who purchased 30 servants; James Brown of Alabama who purchased 29 servants; Woodbridge Odlin of Gloucester county, New Jersey who purchased 22 servants; James T. Watson of New York City who purchased 20 servants; and Jacob Bollinger of St. Louis, Missouri who purchased 18 servants.

308   German immigrant servitude in America, 1745–1835 German immigrants, especially those who would pay through entering servant contracts in America, experienced a series of contractual and documentary steps from embarkation in Europe through final debarkation and registration of their servant contract in America. The following illustrates this paper trail by giving examples of ship contracts, government documents, eyewitness descriptions of the process, and contract registration examples. They trace in order the sequence of steps the immigrants passed through, from boarding the ship in Europe to having their servant contracts registered in America. In Europe, German immigrants signed a passenger contract which gave them a choice of paying a cash price for passage upfront and thus being “free” immi­ grants, or paying a credit price for passage to be paid in Philadelphia and thus being “redemptioner” immigrants. The passenger contract specified the con­ ditions and restrictions on “redeeming” or repaying the passage loan upon arrival in America. The passenger contract also specified the provisions, destination, and what behavior was expected both of all passengers, free and redemptioner, and of the captain. Few original passenger contracts have survived (Chapter 13). One example is for the ship Favorite which arrived in Philadelphia from Amster­ dam on 15 November 1803 with 153 German immigrants. It is reproduced in Figure 16.1. Note that it is written in German. An example of a surviving passenger contract for German immigrants from the period covered by “Book C of redemptioners” that was subsequently trans­ lated into English is for the ship Elizabeth which arrived in Philadelphia from Amsterdam in 1819.2 The names of some of the passengers on this passenger contract subsequently appear in “Book C of redemptioners.” The passenger con­ tract reads as follows (the passengers’ signatures and individual accounts that are at the bottom of the contract are not reproduced here): We the undersigned: I, M. Adams, Captain of the Ship Elizabeth on one part, and we the passengers on the other part do obligate ourselves— First, We the passengers to take our passage with the above mentioned Capt. Adams to Philadelphia in North America, and to conduct ourselves as good passengers ought to do, quiet and orderly, and to be satisfied with the food mentioned at foot as per agreement with the Captain, and with regards to water and other provisions, to follow the Captain’s directions as he shall find necessary through a long passage or other circumstances. Second. We agree to take our passage on the following Conditions, viz. to pay For those who are able to pay in Amsterdam for each person man or woman 180 fr.3 Children under four years of age are free— From four to twelve years to pay 90 fr. From twelve years and older to 180 fr.

Figure 16.1 Passenger contract for German immigrants on board the ship Favorite, captain John Alftan [Alston], bound for Philadelphia from Amsterdam (source: original contract reproduced with permission from The Historical Society of Pennsylvania, Philadelphia (society miscellaneous collection, box 9c, white servitude folder #1). The ship arrived in Philadelphia 15 Novem­ ber 1803 with 153 passengers, see Strassburger (1934: vol. 3, 135–7 [ship #470])).

310   German immigrant servitude in America, 1745–1835 For those who are not able to pay here or only in part, the passage to be Children under four years of age free From four to twelve years 95 fr. From twelve years and older 190 fr. and 200 fr. as specified. Those who have to pay their passage in America shall be obligated to do so in ten days after their arrival.4 No passenger shall be allowed to leave the vessel in America without leave from the Captain and in particular those as have not paid their passage money. Should any one of the passengers die on the voyage, the family of such person shall be obliged to pay his passage, if such deceased took place on more than half the distance of the voyage, but should the person die this side half the distance, the loss of the passage shall fall to the Captain. In return I, M. Adams obligate myself to carry these passengers to Philadelphia, to accommodate them with necessary comfort and give them daily the here below mentioned proportion of victuals—children not to receive anything. Sunday—one pound Beef and half pound Rice, Monday—one pound Flour, Tuesday—one half pound Pork with pease, Wednesday—one pound Beef and barley, Thursday—like Tuesday, Friday—like Monday, Saturday—like Wednesday, One pound Butter, one pound Cheese, six pounds Bread, per week. One glass Gin and three quarter gallons Water per day. There shall also be on board a sufficiency of Vinegar to cleanse the vessel and for the refreshing the passengers. To all this we bind ourselves with our persons and property. Witness Van Olivier & Co. Amsterdam, 4 May 1819. (Tepper 1975: 257–8) After the ship docked in Philadelphia, a health officer inspected the passengers. The following is an example of the health officer’s handwritten report, one which covers immigrants appearing in “Book C of redemptioners” and was appended to that register: Health Office, Philad a 1824 To Jacob H. Hoeckly Register of German Passengers  I do hereby report That I have received all the above named Passengers (25 in number) on board the Ship Jane Capt. John Smith, arrived (this day) at

Processing German servants, 1817–31   311 the Port of Philadelphia from Amsterdam, and that none of them are superannuated, impotent or otherwise likely to become chargeable to the Public, but all of them sound, without any defect in mind or body. NB. Health Officer (“Book C of redemptioners”) Subsequently, the registrar of German passengers passed the health officer’s information on to the secretary of the state of Pennsylvania, as the following handwritten letter which was appended to “Book C of redemptioners” illustrates: Philadelphia 1824 Molten C. Rojeng Esq. Secretary of the Commonwealth of Pennsylvania—Harrisburg Sir, Agreeably to the Act of Assembly (in such case made & provided) I hereby transmit to you the list of the names of German Passengers arrived at the Port of Philadelphia on the day of inst. From Amsterdam on board the Ship Jane Capt. John Smith—all of whom (agreeably to the Report of the Health Officer) have been permitted & Licensed to land. J. F. H. Register of German Passengers (“Book C of redemptioners”) If the health officer determined that the passengers were suitable for landing, the merchant in charge of the cargo would place advertisements in the local news­ papers announcing that immigrants on board the ship were willing to enter servi­ tude in exchange for paying their passage debts. The advertisements attracted buyers from the counties surrounding Philadelphia, from neighboring states, and occasionally from places as far away as Alabama, Missouri, and even Cuba. The buyers came on board the ship and negotiated with the immigrants regarding the terms of the servant contract that the immigrant would agree to, in exchange for the buyer paying the immigrant’s passage debt. The only eyewitness description of the sale of German immigrant redemp­ tioner servants in Philadelphia from the period covered by “Book C of redemp­ tioners,” including servants found in that register, is by Henry Bradshaw Fearon. His description, however, should not be taken at face value. Fearon was a well-­ to-do English traveler, unfamiliar with working class conditions and institutions, and somewhat biased against America. In his description, his errors regarding the ethnicity of the immigrants and how contract prices were determined speak to his unfamiliarity with the institution and process of redeeming. In addition, Fearon observed the servant bargaining process in Philadelphia in the fall of 1817 when a flood of destitute Germans had been driven from the Rhineland by the famine and crop destruction cause by the 1816 agricultural

312   German immigrant servitude in America, 1745–1835 disaster known as the “year without a summer.”5 This was a random and acute climatic crisis unparalleled for several centuries in terms of its severe effect on the Rhineland. Thus, Fearon observed a year which was atypical regarding the conditions of German immigrants. Still, Fearon’s description is all we have from this era and it has some merit. He wrote on October 12, 1817 (the W. Odlin mentioned in the quotation was one of the larger purchasers and resellers of servants recorded in “Book C of redemptioners”): A practice which has been often referred to in connection with this country, naturally excited my attention. It is that of individuals emigrating from Europe without money, and paying for their passage by binding themselves to the captain, who receives the produce of their labor for a certain number of years.6 Seeing the following advertisement in the newspapers, put in by the captain and owners of the vessel referred to, I visited the ship, in company with a boot-­maker of this city: “THE PASSENGERS “On board the brig Bubona, from Amsterdam, and “who are willing to engage themselves for a limited “time, to defray the expense of their passage, consist “of persons of the following occupations, besides wo“men and children, viz. 13 farmers, 2 bakers, 2 butch“ers, 8 weavers, 3 taylors, 1 gardener, 3 masons, “1 mill-­sawyer, 1 white-­smith, 2 shoe-­makers, 3 ca“binet-­makers, 1 coal-­burner, 1 barber, 1 carpenter, “1 stocking-­weaver, 1 cooper, 1 wheelwright, 1 brewer, “1 locksmith. – Apply on board of the Bubona, op“posite Callowhill-­street, in the river Delaware, or to “W. Odlin and Co. No. 38, South Wharves. Oct. 2. As we ascended the side of this hulk, a most revolting scene of want and misery presented itself. The eye involuntarily turned for some relief from the horrible picture of human suffering, which this living sepulcher afforded. Mr. _____ enquired if there were any shoemakers on board. The captain advanced: his appearance bespoke his office; he is an American, tall, determined, and with an eye that flashes with Algerine cruelty. He called in the Dutch language for shoe-­makers, and never can I forget the scene which followed. The poor fellows came running up with unspeakable delight, no doubt anticipating a relief from their loathsome dungeon. Their clothes, if rags deserve that denomination, actually perfumed the air. Some were without shirts, others had this article of dress, but of a quality as coarse as the worst packing cloth. I enquired of several if they could speak English. They smiled, and gabbled, “No Engly, no Engly, —one Engly talk ship.” The deck was filthy. The cooking, washing, and

Processing German servants, 1817–31   313 necessary departments were close together. Such is the mercenary barbarity of the Americans who are engaged in this trade, that they crammed into one of those vessels 500 passengers, 80 of whom died on the passage.7 The price for women is about 70 dollars, men 80 dollars, boys 60 dollars. When they saw at our departure that we had not purchased, their countenances fell to that standard of stupid gloom which seemed to place them a link below rational beings . . . An interesting occurrence is said to have taken place the other day, in connection with the German Redemptioners (as by a strange misnomer the Dutch as denominated).8 A gentleman of this city wanted an old couple to take care of his house; —a man, his wife, and daughter were offered to him for sale; —he purchased them. – They proved to be his father, his mother, and sister!!!9 (Fearon 1818: 149–52) After the servant contract was negotiated (see the example contract illustrated above) and the shipper was paid, the contract was registered by the Pennsylvania state government in “Book C of redemptioners.” Examples of these registered contracts, the three registered on 28 October 1819, follows: Daniel Herneisen, with consent of his father, to George Schott, of Philadelphia physician, for six years, to have six weeks schooling for each year of the term, & at the end thereof two complete suits of clothes, one of which to be new. Consideration $40.00 Carl Ludwig Schenck, with consent of his father, to Henry Rush of Leacock township, Lancaster county, Pennsylvania farmer, for three years & nine months, to have six weeks schooling for each year of the term, & at the end thereof twenty dollars in lieu of freedoms. Consideration $76.00 Simon Mesner, to Thomas Thompson, of Philadelphia mariner, for two years, to Have at the end of the term two complete suits of clothes, one to be new. Consideration $77.20 (Grubb 1994: 117) After the contracts were registered, the servants went off with their American masters. The servants’ fates after this are largely unknown. With only a name, adult versus child status, gender, and an initial location to go on, tracing indi­ viduals through subsequent censuses is a difficult task, especially given the fre­ quent corruption of German names by English speakers and document recorders.10 Tracing the 1,035 immigrant German servants listed in “Book C of Redemptioners” through other American records, such as to subsequent cen­ suses, is a research project that still needs to be undertaken.

314   German immigrant servitude in America, 1745–1835 Chapter 17 uses this information to evaluate whether a collapse in American demand for servants can explain why immigrant servitude disappeared. The most popular stories offered in the scholarly literature are collapse-­in-demand explanations. Discovering that such stories cannot be made consistent with the evidence presented in Chapter 17, Chapter 18 explores whether a collapse in the supply of German immigrant servants can explain why servitude disap­ peared. In the process, both subsequent chapters use the evidence illustrated in this chapter to more firmly establish exactly when organized markets for immi­ grant servants disappeared, something that has not been well known before and something that cannot be well established without the evidence illustrated in this chapter.

Notes   * Adapted from material originally published in the “Introduction” to Farley Grubb, German Immigrant Servant Contracts Registered at the Port of Philadelphia, 1817–1831 Baltimore: Genealogical Publishing Co., Inc. 1994). Copyright 1994. © 1994 by the Genealogical Publishing Company. Used with permission of the Genea­ logical Publishing Company.   1 For the origins, history, and function of these end-­of-contract payments, called “freedom dues,” in the immigrant servant trade, see Grubb (2000: 42–75).   2 Based on impressions gathered from studying this market over many years, the typical passenger contract in the German immigrant trade had not changed much in basic design going back as far as the mid-­1720s. For similar passenger contracts for German immigrants to America, see Klepp et al. (2006: 12–13); Langguth (1947: 259–62); Strassburger (1934: vol. 3, 112–14, 131–2, 137–8).   3 In the contract, “fr.” equals florins or florijns also known as guldens. This was a Dutch silver coin. It was equal to £0.09 pounds sterling in 1781 or about $0.41 silver dollars, see McCusker (1978: 9–10).   4 Pennsylvania law gave immigrants 30 days to redeem their passage debts after arrival. This law voided any shorter repayment interval contracted. See Herrick (1926: 293, 305–8).   5 See Hansen (1940: 79–106); Skeen (1981: 51–67); chapters 2 and 18 for more detail on this phenomenon.   6 Fearon is wrong here. The immigrant was not bound to the captain for a certain number of years. As newspaper advertisement indicates, they were only bound to the captain to pay the expense of their passage. There was no servant contract between the captain and the immigrants, nor did the captain work or sell the servants. While on the deck on his vessel when it was docked in Philadelphia, the immigrant entered the servant contract for the first time, negotiating with an American buyer who would agree to pay his passage debt to the ship’s captain as indicated by the latter part of Fearon’s account.   7 This appears to be an exaggeration. From 1815 through 1819, 62 vessels carrying German immigrants landing in Philadelphia where the ship’s tonnage could be identi­ fied were listed in the “Philadelphia Customs House records.” None of these 62 vessels reported over 450 passengers, and only two of the 62 reported carrying over 400 passengers. Only four of the 62 vessels reported a passenger-­per-ton ratio in excess of one—the maximum being 1.36. This does not count the disastrous voyage of the ship April which could not make it up-­river to Philadelphia but landed at New Castle, Delaware in December of 1817 (see Swierenga and Lammers 1994: 303–23).   8 Again, Fearon is wrong. These were German not Dutch immigrants.

Processing German servants, 1817–31   315   9 I have not been able to corroborate this story or anything like this outcome in any sur­ viving document on the German redemptioner trade from this era. 10 See the discussion in the preface. Ferrie (1999) provides an excellent discussion of immigrant tracing and record linkage strategies and their problems.

References Bezanson, A., Gray, R. D., and Hussey, M. (1936) Wholesale prices in Philadelphia, 1784–1861, Philadelphia: University of Pennsylvania Press. “Book A of redemptioners, 1785–1804,” Philadelphia: Unpublished manuscript held at the Historical Society of Pennsylvania. “Book C of redemptioners, 1717–1831,” Philadelphia: Unpublished manuscript held at the Historical Society of Pennsylvania. Fearon, H. B. (1818) Sketches of America: a narrative of a journey of five thousand miles through the eastern and western states of America, London: Longman, Hurst, Rees, Orme, and Brown. Ferrie, J. P. (1999) Yankeys now: immigrants in the antebellum United States, 1840–1860, New York: Oxford University Press. Grubb, F. (1994) German immigrant servant contracts registered at the port of Philadelphia, 1817–1831, Baltimore: Genealogical Publishing Company. Grubb, F. (2000) “The statutory regulation of colonial servitude: an incomplete-­contract approach,” Explorations in Economic History, 37: 42–75. Hansen, M. L. (1940) The Atlantic migration 1607–1860, Cambridge, MA: Harvard Uni­ versity Press. Herrick, C. A. (1926) White servitude in Pennsylvania, Philadelphia: John Joseph McVey. Klepp, S. E., Grubb, F., and Pfaelzer de Ortiz, A., eds. (2006) Souls for sale: two German redemptioners come to revolutionary America: the life stories of John Frederick Whitehead and Johann Carl Büttner, University Park, PA: Pennsylvania State University Press. Langguth, O. (1947) “Pennsylvania German pioneers from the county of Wertheim,” D. Yoder, trans. and ed., Pennsylvania German Folklore Society, 12: 259–62. McCusker, J. J. (1978) Money and exchange in Europe and America, 1600–1775: a handbook, Chapel Hill, NC: University of North Carolina Press. “Philadelphia customs house records, 1800–1883,” Washington, DC: Unpublished manu­ script held at the National Archives, microfilm #425, rolls 16–50. Skeen, E. C. (1981) “The year without a summer: a historical view,” Journal of the Early Republic, 1: 51–67. Strassburger, R. B. (1934) Pennsylvania German pioneers, 3 vols., Norristown, PA: Pennsylvania German Society. Swierenga, R. P. and Lammers, H. (1994) “ ‘Odyssey of woe’: the journey of the immi­ grant ship April from Amsterdam to New Castle, 1817–1818,” Pennsylvania Magazine of History and Biography, 118: 303–23. Tepper, M., ed. (1975) Emigrants to Pennsylvania, 1641–1819, Baltimore: Genealogical Publishing Company.

17 The disappearance of organized immigrant servant markets in the U.S. Five popular explanations reexamined*

Organized markets for European immigrant servants in North America began in Jamestown around 1620 and ended in the ports of Philadelphia, Baltimore, and New Orleans around 1820. For two centuries these markets survived, even flourished, in spite of numerous wartime interruptions, political revolutions, depressions in the transatlantic shipping market, cyclical recessions in the American economy, and competition from both slave and native-­born free labor. During the eighteenth century roughly half of the European emigrants to British North America entered servitude to pay for their transatlantic passage (Galenson 1981; Grubb 1992b).1 As late as 1800, 45 percent of the German emigrants arriving in Philadelphia sold themselves into servitude (Grubb 1985a). In the decade following the War of 1812, however, the organized market for European immigrant servants disappeared. While instances of European immigrant servitude appear after 1820, they are isolated, few in number, and not of sufficient volume to sustain an organized market. By the time of the great acceleration in European immigration to the United States, in the second quarter of the nineteenth century, the payment of transatlantic passage fares by servitude had disappeared (Erickson 1984; Galenson 1984). The Europeans who settled the western frontier, built the canals and railroads of the nation, and labored in the factories of the northern states were free artisans, farmers, and wage laborers. If half of these immigrants had been servants, as during the eighteenth century, then the course of U.S. development after 1820 would have been different because servitude altered the geographic distribution and social integration of immigrants into the American work force (Grubb 1985b).2 The end of immigrant servitude, therefore, was a watershed event in the history of European emigration to America, as well as a key turning point in the history of American economic and social development. Given the importance of European immigrant servitude, it is surprising how little is known about why it ended. Many hypotheses have been advanced, but no consensus has emerged. This current state of affairs is due to a lack of evidence. Without evidence, explanations proliferate because it is difficult to test and eliminate hypotheses. This paper will rectify this situation by providing quantitative and qualitative evidence for evaluating the five most frequently advanced hypotheses on why the organized market for European immigrant servants dis-

Disappearance of immigrant servant markets   317 appeared, namely that American employers stopped buying servants because (1)  abolition of imprisonment for debt made contract enforcement difficult, (2)  interference from immigrant aid societies made contract enforcement difficult, (3) servants ran away too much, (4) the recession of 1819 made employers reluctant to buy servants, and (5) employers found wage labor superior. For many scholars, the persuasiveness of these five explanations has made further research seem unnecessary. Therefore, before research can progress, these five explanations must be shown to be inconsistent with the historical record. First, evidence on when immigrant servitude disappeared is presented. Second, direct evidence is used to explore the validity of each explanation individually. In addition, all five explanations assume that the demand for servants fell, albeit for different reasons. The last section tests all demand-­driven explanations with evidence on servant market prices. Finally, the most likely causes of the end of immigrant servitude and the most fruitful directions of future research are discussed.

European immigrant servant markets in America after 1815 The last manifestation of European immigrant servitude of any magnitude was in the German passenger trade.3 German immigration to the United States was restrained by the American Revolution, the 1808 embargo, and the Napoleonic Wars. With the return of peace in 1815, the volume returned to the levels experienced during the Peace of Amiens (1802–5) and the 1770s. While ships carrying German immigrants are known to have arrived in Baltimore, New Orleans, New York, and Philadelphia between 1815 and 1820, only the volume of servants landing at Philadelphia is known. Fortunately, the main port of entry for German immigrants before the 1830s was Philadelphia.4 In 1784 the German Society of Pennsylvania convinced the state to register all German servants as they debarked in Philadelphia. The registers included information on the servant’s name, contract length, and sale price, and on the buyer’s name, residence, and occupation. The registers for 1785 through 1804 and for November 1817 through 1831 have survived.5 Registration was beneficial to all contracting parties. The official in charge was bilingual and provided translation, and the records provided legal protection for both employer and servant against unilateral attempts to change the contract. Registration was relatively costless, and unregistered contracts were invalid. These conditions imply that the registers provide a reasonably comprehensive count of German immigrant servants. Even buyers from distant states registered their servants before departing the city (see Table 17.3). The volume, contract prices, and contract lengths for German immigrant ser­ vants debarking at Philadelphia are reported in Table 17.1. The volume averaged 191 per year between 1785 and 1804. After peace returned in 1815, the volume was higher than during the first two decades of the republic, averaging over 300 per year from 1817 through 1819. This later volume is comparable to the number arriving prior to the Revolution, which averaged 353 per year from 1770 through

3,828 274 396 328 18 5 14 1,035

1785–1804 1817 1818 1819 1820 1821 1822–31 1817–31

1,596 54 106 84 4 1 0 249

79.40 77.53 81.81 100.87 91.62 146.96 99.75

3.03 3.14 3.07 2.79 2.50 2.00 2.70

2.89

3.29 3.16 3.12 2.76 2.50

74.36

72.36 60.59 71.62 87.62 77.62

Average contract debt ($)

Notes Evidence for 1817 starts on 5 November. The yearly distribution of servants from 1822 to 1831 was 0, 1, 4, 2, 1, 2, 1, 1, 0, and 2, respectively. Contract debts are converted into U.S. dollars following Davis and Hughes (1960); Grubb (1990b); McCusker (1978); Officer (1983), and then deflated (1795 = 100) following Bezanson, Gray and Hussey (1936). Children and single adult females are not enumerated separately. The totals for single adult males and married servants only include those for which contract lengths and prices were recorded.

602 76 78 32 6 0 0 192

Average contract length (years)

Total (N)

Average contract debt ($)

Total (N)

Average contract length (years)

Married servants

Single adult male servants

Sources: “Book A of Redemptioners”; “Book C of Redemptioners”.

Total number of servants

Years

Table 17.1 The market for German immigrant servants in Philadelphia, 1785–1831

Disappearance of immigrant servant markets   319 1775 (Grubb 1989: 167). The relative size and continuity of this volume indicates that the organized market for German immigrant servants had survived the War of 1812. In 1820 the market suddenly collapsed, never to recover. Only 18 servants debarked in 1820, and only 19 were registered between 1821 and 1831. These last 19 contracts were mostly children.6 In 1831 the register was discontinued. The sudden disappearance of German immigrant servitude at the ports of Baltimore and New Orleans also occurred in 1820.7 This evidence suggests that the organized market for European immigrant servants ended abruptly in 1820. This timing reaffirms the conclusion that immigrant servitude did not end through direct legal proscription either in the United States or in Europe. It was not directly prohibited in the United States until 1885 (Erickson 1984: 38–56; Cloud and Galenson 1987; Steinfeld 1991: 123–95). Servitude among U.S. residents remained legal long after immigrant servitude disappeared, with the most common form being apprenticeship and the indenture of indigent children (Clark 1977; Jacoby 1991a, 1991b).8 As a consequence, other explanations have been sought for the end of immigrant servitude.

Abolition of imprisonment for debt and the collapse of immigrant servitude One popular explanation of the end of immigrant servitude involves the abolition of imprisonment for debt (Erickson 1984: 39; Herrick 1926: 266; Morris 1981: viii–ix, 354–63, 516). German immigrant servants paid for their migration expenses by borrowing from their shipper prior to embarkation. The immigrant agreed in the passage contract to sell himself as a servant in America if necessary to redeem the debt (hence they were called redemptioners). When their ship docked in America, redemptioners negotiated directly with employers to determine the length of time they needed to serve in order to get the employer to pay off the passage debt (Grubb 1988; Smith 1947: 20–5). Without debtors’ prison, the redemptioner could just refuse to sell himself into servitude when the ship docked in America. The shipper would have no recourse except to allow the immigrant to default on the passage loan and debark in America as a free person. Without a debtors’ prison, the redemptioner who had been sold could default on the servant contract by refusing to work, leaving the employer no way of recovering his money and no recourse except to let the servant go free. The legal abolition of imprisonment for debt, however, occurred some time after the disappearance of immigrant servitude. By 1830 only two states, Ohio and Kentucky, had abolished imprisonment for debt. In 1830 the estimated number of debtors in the prisons of Pennsylvania, Maryland, and New York was 7,000, 3,000, and 10,000, respectively, and according to the Prison Discipline Society the debtor-­to-criminal ratio was nearly five to one (Carlton 1908; Erickson 1984: 39). Most of the imprisoned debtors owed less than $50, and many owed under a dollar, which was far less than the debts incurred by immigrant servants. Only by 1840 had nearly every northern state abolished imprisonment

320   German immigrant servitude in America, 1745–1835 for debt. Immigrant servitude disappeared at least ten years before imprisonment for debt was abolished. The source of the hypothesis that abolition of imprisonment for debt ended immigrant servitude is Frederick Geiser (1901: 42). He referred to a Pennsylvania law passed in February 1819 as the legal “death blow” to servitude. The law stated “that no female shall be arrested or imprisoned for or by reason of any debt contracts after the passage of this act.” While the timing coincides with the collapse of immigrant servitude, the impact of the law on immigrant servitude is questionable. First, only single adult females were affected. However, they comprised only about 10 percent of the German immigrant servant trade between 1785 and 1818 (“Book C of Redemptioners”; Grubb 1985a). Ending adult female servitude would have had little impact on the total volume of immigrant servants. Second, the proportion of single adult females among German immigrant servants actually increased after 1819 to 19 percent. Apparently the law did not deter the purchase of female immigrant servants. Debtors’ prison was used during the last years of the servant trade. In 1819 Christopher Mayer, president of the German Society of Maryland and champion of immigrant and servant rights, advised a group of recalcitrant German redemptioners on the ship Vrouw Elizabeth docked in Baltimore that their captain “has a lawful right to have you committed to prison, to remain there on meager fare, until your debt is paid, if you do not consent to hire for a reasonable time . . . and the German Society will assist him in this” (Henninghausen 1909: 85–6). In August 1819 Ludwig Gall arrived in Philadelphia from Germany with 13 immigrants who had borrowed the cost of passage from him. When two refused to enter servitude to repay their debt, Gall had them thrown in prison. This action induced the other 11 to repay their passage debts by entering servitude (Trautmann 1981: 40–1). Even after the collapse of the organized trade in immigrant servants, debtors’ prison was still used against the few servants who crossed the Atlantic. On 31 October 1829 Niles’ Weekly Register reported that a group of men from England who were indentured to the Chesapeake and Ohio Canal Company for four months to pay for their passage had refused to comply with their agreement. Judge Cranch remanded them to prison until they complied with their contract and discharged their debt. Debtors’ prison was used effectively throughout the history of European immigrant servitude to compel redemptioners to repay passage debts and fulfill servant contracts. While court and prison expenses were incurred to force recalcitrant redemptioners to fulfill contracts, there is little reason to believe that these expenses increased over time or became prohibitive by 1820. Bankruptcy laws and the abolition of imprisonment for debt may explain why European immigrant servitude could not be revived later in the nineteenth century, but it cannot explain the disappearance of European immigrant servitude in 1820.

Disappearance of immigrant servant markets   321

Immigrant aid societies and the collapse of immigrant servitude The resurgence of German redemptioner immigration after 1816 led to renewed activism by German immigrant aid societies to correct the abuses occurring under that system. Shippers who violated passenger contracts and employers who violated servant contracts were prosecuted in court. German immigrant aid societies were instrumental in pressuring state legislatures to regulate areas of behavior not directly specified in the private contracts for ocean passage and servitude. By 1818 redemptioner regulations in Maryland and Pennsylvania included oversight by a bilingual official, registration of servant sales (unregistered sales were invalid), provision of education for servant children, a thirty-­day period after arrival for redemptioners to repay their passage debts and during which they had to be maintained at the shipper’s expense, a limitation on adult contract length (Maryland only) stipulating that adults could not be made to serve over four years, a prohibition on the separation of married servants without their consent, a prohibition on selling servants across state lines without their consent, establishment of customary freedom dues and procedures for handling sick redemptioners, and the restriction of forced debt pooling and confiscation of passenger luggage after the voyage as a method of paying for the passage of redemptioners who died during the voyage.9 Could employers have stopped buying immigrant servants after 1819 because regulatory compliance and oversight by immigrant aid societies, who threatened lawsuits when contracts were violated, were too burdensome? This explanation is difficult to support. First, most of the regulations of the redemption trade in Pennsylvania, the primary state of arrival for German immigrant servants, were in force before 1818, with many laws dating back to 1764. The additional regulations passed in 1818 were minor and, in most cases, repeated previous regulations. The Maryland regulations, passed en mass in 1818, closely followed the established body of Pennsylvania law. In addition, no evidence indicates changes in the level of enforcement in this period. If these laws did not stop the immigrant servant trade to Pennsylvania before 1818, why should they stop servitude after 1818? It seems unlikely, therefore, that state regulations caused the sudden collapse of immigrant servitude in 1820. Second, German immigrant aid societies did not try to end servitude. Their goal was to curb contract fraud. Strict enforcement of contracts also helped shippers and employers. German immigrant aid societies insisted that redemptioners honor their contracts. For example, Christopher Mayer berated recalcitrant German redemptioners on the Vrouw Elizabeth for refusing to honor their side of the passage contract (Henninghausen 1909: 85–6). He called their refusal “unjust, unlawful, and ungrateful,” and told them, “The German Society makes it its duty to assist your countrymen when they are in need, and to protect them as far as it is able; but it will also not suffer any injustice to be done by emigrants, and by advice and act induce them to fulfil their obligations.”

322   German immigrant servitude in America, 1745–1835 The actions of the German Society of Pennsylvania regarding the most notorious event in the German immigrant trade illustrates their position. The ship April arrived in New Castle, Delaware, in December 1817 with German redemptioners. The ship had experienced months of quarantine in Holland and a rough voyage that left over 400 dead and the ship incapable of proceeding upriver to Philadelphia. Reports of misconduct by the shipper prompted the German Society of Pennsylvania to send their secretary, Henry T. Virchaux, to investigate (Virchaux 1818). On 22 January 1818 Virchaux reported, “my instructions were to get sight of any contract made.” He explicitly told the shipper that he had no instructions to seek the shipper’s prosecution. His report indicates that he served as an arbitrator and translator between the passengers and their shipper. He convinced the shipper to fullfil his end of the passage contract by providing better provisions for those still on board waiting to negotiate servant contracts to pay their passage debts, and he convinced the passengers to fulfill their end of the contract by not leaving the ship until they had arranged to pay their passage debts. No passengers complained about or resisted entering servitude to work off their passage debts, and Virchaux even helped negotiate some servant contracts. The only regret Virchaux had regarding the passengers’ entry into servitude was that in the absence of good advice “several of the children have been bound out to very improper places such as tipling houses” (Virchaux 1818). This evidence indicates that the actions of immigrant aid societies and the effects of state regulations did not cause immigrant servitude to collapse in 1820. On the contrary, these actions most likely encouraged servant immigration and employment. They established firmer property rights that helped minimize the cost of transferring servant property, and they emphasized contract enforcement that reduced fraud and increased the level of contractual certainty for all parties—employer, shipper, and servant.

Runaway servants and the collapse of immigrant servitude Another popular explanation for the end of immigrant servitude is that servants increasingly ran away (called here the runaway hypothesis). For example, the economist Yoram Barzel (1989: 83–4) argues that indentured servitude disappeared because “at its inception, the population in America was sparse, and servants were easily distinguished from free people. With time, the population grew and diversified, making escape easier.” In 1788 Daniel Kent, a servant in Chester County, Pennsylvania, advised his parents in Ireland “not to send any more servants. . . . There are but four or five out of twenty and upwards that came to this part of the country with me that have stayed with their masters, the rest ran away. Some have been taken up and condemned to serve a longer time” (Barnard 1904: 26). The runaway hypothesis is based on the following simple argument. As the American population grew and diversified, opportunities to hide successfully after escape increased. The implication is that the frequency of running away

Disappearance of immigrant servant markets   323 increased over time, and thus the number of masters willing to buy servants dwindled. At some point the rate of escape became so high that no one would buy servants. While this simple chain of reasoning seems persuasive, three counter-­ arguments and direct evidence on servant escapes cast doubt on its ability to explain the end of immigrant servitude. First, it cannot explain the abruptness with which immigrant servitude ended. The size and diversity of the population had been increasing for over a century, and yet the number of masters who purchased German immigrant servants did not decline. In fact, between 1817 and 1819 that number reached one of its highest peaks (Table 17.1; Grubb 1989: 167). The purchase of servants stopped abruptly in 1820. The suggestion that servants increased their rate of escape in 1819 so suddenly and substantially that no one was willing to buy servants the following year stretches credulity. Such a large and abrupt change should have produced commentary in the press or by employers and shippers, and yet no such commentary can be found. Second, runaway servants may have found it increasingly difficult to escape successfully. Immigrant servants were almost exclusively German during the last 30 years of the institution. The German-­speaking share of the Pennsylvania population reached its peak in the 1750s and declined to relatively low levels during the early decades of the republic. The retreat in the use of the German language among German Americans, particularly by 1815, meant that runaway German immigrant servants were easily betrayed by their language. Third, the design of the servant contract led to a low and declining rate of escape. Servants had an incentive to run away because their passage fare was paid before the labor contract commenced. However, contracts were required by law to include a performance bond paid to the servant only upon completion of the contract. This bond, called freedom dues, consisted of cash and/or goods and could equal half the size of the original contract price (Grubb 1988: 588–91; 1992b: 185). If the servant made it through the first half of the contract, then the allure of freedom dues was a powerful financial inducement to complete the remainder of the contract. Immigrant servants were reluctant to run away until they had learned enough to escape successfully, such as the lay of the land, the best direction to run, the penalties for escape, the chances of capture, where to hide, how to dress and act to avoid detection, and where to find food, work, shelter, transportation, and so on. Before 1785 the typical adult contract was four years (Grubb 1992b). Thus, the servant would have only about two years to learn enough to escape successfully before the allure of freedom dues would entice him to stay. Therefore, it should not be surprising to find that only 5 to 8 percent of immigrant servants are estimated to have run away in Pennsylvania during the middle of the eighteenth century.10 After the Revolution, the length of adult contracts fell, eventually reaching 2.5 years by 1820, while the value of freedom dues stayed constant or even rose slightly (Table 17.1; Grubb 1992b: 185). Thus, the amount of time the servant had to learn how to escape successfully before the allure of freedom dues

324   German immigrant servitude in America, 1745–1835 induced him to stay fell to around 1.25 years. These changes should have caused the frequency of running away to fall over time, reaching its lowest levels as the institution was disappearing. Quantitative evidence on the incidence of running away among German immigrant servants in the Philadelphia region from 1785 to 1820 is consistent with this analysis (see Table 17.2). The number of German servants advertised in the Philadelphia newspapers Pennsylvania Gazette and Poulsan’s Daily American Advertiser as having run away was compared with the registration of German immigrant servants debarking at Philadelphia for the years when the evidence overlaps, 1785–96 and 1817–20. The incidence of running away was calculated two different ways, both constructed to overstate the true rate.11 First, the number of servants under contract at any given time, and thus at risk for running away, was compared with the number advertised as having run away. Second, the names of servants advertised as runaways were matched with the names of servants registered at the port of Philadelphia. For both calculations, the incidence of running away among German immigrant servants between 1817 and 1820 was a fraction of a percentage point. The only exception was in 1820 after immigrant servitude had already collapsed, when three of seventeen ran away. Table 17.2 Incidence of runaway servants among German immigrant servants in the Philadelphia region, 1785–1820

German runaway servants advertised

Total November– 1818 1785–96 December 1817

1819

1820

Total 1817–20

71

1

18

7

 9

35

German runaway servant   3.4 percentage

0.0004

  0.01

0.003

  0.01

  0.01

Advertised German runaway servants linked by name to the register of arriving German servants

0

 4

2

 3

 9

Percentage of German servants registered who ran away

0.0

  1.01

0.61

17.7

  0.01

Source: Grubb 1992a; Poulsan’s Daily American Advertiser; Table 17.1. Notes German runaway servants were identified in the advertisements by being either explicitly listed as German or by other indications, such as the ad stating the servant “spoke little English,” the spelling of the servant’s name when no ethnic designator was used, and by matching the servant’s and master’s names to the register of newly arrived German immigrant servants. Line 2 equals line 1 divided by the estimated number of German immigrant servants under contract in the region in that year and thus at risk to escape. The percentage is biased high because the denominator assumes an average contract life of three years, which slightly understates the average adult contract length and vastly understates the contract length of children (Table 17.1; Grubb 1992b). Line 4 equals line 3 divided by the appropriate servant total in Table 17.1.

Disappearance of immigrant servant markets   325 In addition, the trend in the incidence of running away was decreasing over time. Using similar methods and sources, the incidence of immigrant servants running away in the Philadelphia region in 1745–6 was estimated to be 6.7 percent (Grubb 1989: 159). Table 17.2 indicates that the incidence of running away was falling, from 6.7 percent in 1745 to 3.4 percent by 1785–96, and then to 0.01 percent by 1817–20. Both the minuscule level and the declining trend of the incidence of running away are inconsistent with the runaway hypothesis of the end of immigrant servitude.

The recession of 1819 and the collapse of immigrant servitude The United States experienced a severe financial panic in the spring of 1819. Trade and production contracted, and the demand for labor fell. Hansen (1940: 103) found that “the cotton mills of Philadelphia and the vicinity, which in 1816 had worked 2,325 hands, employed only 149 on October 2, 1819.” This unexpected and sharp fall in labor demand may have affected the market for immigrant servants. Hansen (1940: 105) concluded, The economic breakdown affected no speculators more directly than the redemptioner broker whose servants were waiting on board the ships lying in the harbors of Philadelphia and Baltimore. Since no buyers appeared and every day meant an added expense in feeding the laborers, the brokers sent them out like groups of slaves to the West and the South in the hope of peddling them off along the way. Others, who could not secure credit to maintain the servants any longer, set them free . . . From this blow the redemptioner trade never recovered. While the 1819 financial panic clearly affected transatlantic trade, the hypothesis that it put an end to European immigrant servitude is difficult to sustain. First, unforeseen financial panics and a resulting decline in servant demand were recurring events. The economic crises of 1764 and 1773 had effects similar to those described by Hansen for 1819. Muhlenberg (1942, 2:121) observed that during the financial crisis of 1764 “four shiploads of German emigrants have arrived again [in Philadelphia] this fall, but since money is scarce, the people are not being sold very fast.” Immigrant servitude had always revived after financial panics just like the trade in other transatlantic commodities. Because these earlier crises had not ended the servant trade, why would the panic of 1819? Second, the behavior of prices and contract lengths in the immigrant servant market does not reveal any adverse effects from the panic of 1819. If the panic had reduced the American demand for servants, then shippers, desperate to lure reluctant buyers to the docks, should have reduced passage debts and sold servants at a discount. However, the evidence in Table 17.1 indicates that the opposite happened. The average passage debt did not fall, but actually rose, from 78 U.S. dollars in 1817 to 101 U.S. dollars in 1819.12 If the panic had reduced the

326   German immigrant servitude in America, 1745–1835 American demand for servants, then redemptioners should have offered longer contracts in order to lure reluctant buyers into paying their passage debts. However, the evidence in Table 17.1 indicates that the opposite happened. The average contract length for adults did not rise, but actually fell from 3.14 years in 1817 to 2.79 years in 1819.13 Third, the characteristics of the buyers and the pattern of sales in the immigrant servant market do not reveal any adverse effects from the panic of 1819. Hansen (1940: 105) suggested that the sale of large blocks of servants to buyers from distant states in the South and West was a sign of acute crisis in the immigrant servant market brought on by the panic. Tables 17.3 and 17.4, however, indicate that 1819 was not an atypical year for the German immigrant servant market in these regards. While 22.8 percent of the servants were sold to traders from distant states in 1819, even more (31.2 percent) were sold to traders from distant states in 1818. While traders from Alabama, Illinois, and North Carolina purchased 16.2 percent, 2.7 percent, and 1.8 percent of the servants in 1819, traders from Alabama, New York, and Virginia purchased 15.2 percent, 5.3 percent, and 4.6 percent of the servants in 1818, and traders from Missouri, Virginia, and Mississippi purchased 8.8 percent, 1.8 percent, and 1.1 percent of the servants in 1817, respectively. This evidence indicates that the financial crisis of 1819 did not produce a market characterized by the desperate sale of servants to distant states. In 1819 the largest buyer from a distant state was N. Farrows from the Alabama Territories, who purchased 46 servants. Farrows paid full price and had to offer the shortest contracts in the market, 2.19 versus an average of 2.78 years for adult males, in order to get servants to go to Alabama. Clearly, Farrows received no bargain, and the shipper was not making a desperation sale. While large blocks of servants were purchased in 1819 (one block of 62 and one of 46), large blocks of servants were also purchased in 1818 (one of 30 and one of 21) and in 1817 (one of 29 and one of 18). The largest block of servants sold in 1819 was not to a trader from a distant state, but to Benjamin B. Cooper of Gloucester County, New Jersey, across the river from Philadelphia. In fact, the only anomaly in the pattern of sales in 1819 compared with 1818 and 1817 was the relatively high percentage of servants purchased by masters from Gloucester County. Finally, the percentage of buyers who bought just 1 servant actually increased from 65 percent to 86 percent between 1817 and 1819, and the percentage of servants purchased by buyers who purchased only one servant was the same in 1819 as it was in 1818 (47 percent). The only significant change in the size of purchases between 1817–18 and 1819 was a relative decline in purchases involving groups of two to five servants. The evidence, therefore, indicates that the financial crisis of 1819 did not produce a market characterized by the desperate sale of large blocks of servants. The immigrant servant market did experience changes after the War of 1812 regarding the characteristics of the buyers and the pattern of sales. Table 17.3 compares the geographic distribution of servant buyers in the 1787–1804 Phil­ adelphia market with that in 1817–31. The percentage of servants sold to masters

21.9 8.0 11.0 3.7 4.7 2.2 3.7 3.7 4.4 1.1 1.1 0.7 9.6

75.8 5.5 1.5 2.6 1.5 0.0

11.1

Subtotal

Other nearby states Gloucester Co., nj Burlington Co., nj Other nj countiesc Delaware Maryland

Subtotal

4.9

0.8 1.0 1.5 0.8 0.8

62.8

26.0 14.1 4.0 1.5 1.8 1.8 1.3 1.3 0.3 1.8 1.8 0.5 6.6

Nov.–Dec. 1817 1818 (N = 274) (N = 396)

Pennsylvania Philadelphia City Philadelphia Co.a Lancaster Co. Bucks Co. Cumberland Co. Berks Co. Chester Co. Lebanon Co. Lehigh Co. Montgomery Co. Northampton Co. Delaware Co. Other pa countiesb

Residence of purchaser

28.6

21.7 2.4 2.1 2.1 0.3

47.1

21.0 15.2 4.6 1.8 0.0 1.2 0.6 0.3 0.3 0.9 0.3 0.9 0.0

1819 (N = 328)

Table 17.3 Geographic distribution of servant buyers, 1787–1831 (in percentages)

8.1

5.4 0.0 2.7 0.0 0.0

83.7

37.8 27.0 2.7 0.0 0.0 2.7 0.0 0.0 0.0 0.0 0.0 13.5 0.0

1820–31 (N = 37)

13.9

8.8 1.6 1.7 1.4 0.4

62.0

23.8 13.3 6.0 2.1 1.9 1.7 1.6 1.6 1.4 1.3 1.1 1.1 5.1

Total 1817–31 (N = 1,035)

10.3

5.0 2.7 1.6 0.4 0.6

84.5

44.9 11.8 7.6 1.5 0.2 3.0 2.8 0.0 0.0 4.8 1.4 2.3 4.2

continued

Total 1787–1804 (N = 3,129)

100.0

Grand total

100.0

15.2 0.0 4.6 5.3 2.5 0.0 1.3 0.8 1.0 0.0 0.0 0.5 0.0 0.0 31.2 1.1 100.0

16.2 0.0 0.0 0.0 1.8 2.7 0.0 0.9 0.0 0.0 0.9 0.0 0.3 0.0 22.8 1.5

1819 (N = 328)

100.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 8.2

1820–31 (N = 37)

100.0

10.9 2.3 2.2 2.1 1.6 0.9 0.7 0.6 0.4 0.3 0.3 0.2 0.1 0.0 22.6 1.5

Total 1817–31 (N = 1,035)

100.0

0.0 0.0 0.4 0.1 0.0 0.0 0.0 0.7 0.2 0.0 0.0 0.1 0.0 0.6 2.1 3.3

Total 1787–1804 (N = 3,129)

Notes a Excludes residents of Philadelphia City. b Includes Dauphin, Bedford, Adams, Centre, Columbia, Fayette, Franklin, Huntingdon, Lycoming, Mifflin, Union, Westmoreland, Northumberland, Washington, Luzerne, and York counties. c Includes Cape May, Sussex, Somerset, Cumberland, Hunterdon, and Salem counties. d Includes Florida, Massachusetts, and New Hampshire.

Source: “Book C of Redemptioners”; Grubb 1985b: 262–3.

0.0 8.8 1.8 0.4 0.0 0.0 0.7 0.0 0.0 1.1 0.0 0.0 0.0 0.0 12.8 0.3

Nov.–Dec. 1817 1818 (N = 274) (N = 396)

Other states and countries Alabama Missouri Virginia New York North Carolina Illinois Washington, dc Georgia Ohio Mississippi Cuba Kentucky South Carolina Other statesd Subtotal Unknown

Residence of purchaser

Table 17.3 continued

100.0 (100.2)

99.8 (100.1)

  0.8 (15.2)

  0.4 (5.3)

100.2 (100.1)

   0.6 (14.2)    0.6 (19.1)

   0.6 (1.9)    0.6 (2.5)    0.6 (2.8)

  86.3 (46.6)    8.6 (9.3)    2.3 (3.7)

1819 Buyers (N = 175) (Servants) (N = 324)

100.0 (100.0)

  90.3 (80.0)    6.5 (11.4)    3.2 (8.6)

1820–31 Buyers (N = 31) (Servants) (N = 35)

100.2 (99.7)

   0.2 (6.1)

  73.9 (39.6)   16.4 (17.6)    4.3 (7.0)    1.6 (3.5)    0.9 (2.9)    0.9 (2.9)    0.4 (1.6)    0.2 (0.9)    0.2 (1.4)    0.2 (1.7)    0.2 (2.0)    0.2 (2.8)    0.4 (5.8)    0.2 (4.4)

Total 1817–31 Buyers (N = 555) (Servants) (N = 1,035)

Notes Empty spaces indicate no observations for those categories. Percentages do not add to 100 because of rounding. Four servants did not record buyer names.

Source: “Book C of Redemptioners.”

Total

   0.7 (10.6)

   0.7 (6.6)

  65.2 (32.9)   15.9 (16.1)    8.7 (13.1)    4.4 (8.8)    2.2 (5.5)    2.2 (6.6)

 1  2  3  4  5  6  8  9 14 18 21 29 30 46 62 63

77.3 (46.5) 17.2 (20.5)   1.3 (2.3)   0.8 (2.0)   0.4 (1.3)   0.8 (3.0)   0.8 (4.0)

November–December 1817 1818 Buyers (N = 138) Buyers (N = 238) (Servants) (N = 274) (Servants) (N = 396)

Number of servants purchased per buyer

Table 17.4 Distribution of buyers and servants by the number of servants purchased per buyer, 1817–31 (in percentage)

330   German immigrant servitude in America, 1745–1835 from Pennsylvania counties outside Philadelphia and to masters from the nearby states of New Jersey, Delaware, and Maryland stayed relatively constant. However, the percentage of servants sold to masters from Philadelphia City fell from 45 percent to 24 percent, and the percentage sold to masters from distant states and territories rose from 2 percent to 23 percent. A similar shift in the occupational distribution of servant buyers is shown in Table 17.5. Comparing 1787–1804 with 1817–31 indicates that the percentage of servants purchased by different occupational groups changed little. The one exception was farmers, who purchased 21 percent in 1787–1804 versus 31 percent in 1817–31. This change was probably caused by the same phenomenon that led to the declining percentage purchased by masters from Philadelphia City. At 31 percent, farmers were still underrepresented among servant buyers relative to their share in the population. However, by 1817 they were over-­represented relative to the share of farmers among German immigrants. In general, merchants; employers in the service sector, such as innkeepers; and select tradesmen, such as ironmasters, confectioners, papermakers, glaziers, cordwainers, and hatters, were overrepresented among buyers of servants compared with their share both in the population and among German immigrants. By contrast, the percentage of German immigrants who were bakers, coopers, and tailors exceeded the percentage of servants purchased by buyers in those occupations. The evidence from the geographic and occupational distribution of buyers and on the size of their purchases does not support the hypothesis that the financial panic of 1819 ended European immigrant servitude by altering American demand. The panic was only coincidental with, rather than a cause of, the collapse of the servant market. The changes in the servant market between 1800 and 1820 suggest that the institution was capable of adapting to the nation’s expanding frontier and to changes in demand across occupations. It could deliver contract labor to any occupation, anywhere in the nation. This ability suggests that immigrant servitude was a flexible and dynamic institution, able to survive the changing demands of the nation, rather than an institution on the verge of collapse.

Wage labor and demand explanations of the collapse of immigrant servitude Another popular explanation of the end of immigrant servitude is that employers increasingly found wage labor superior. As employers substituted wage labor, the demand for immigrant servants fell until eventually none were purchased (Erickson 1984: 39–46; Salinger 1987: 148–52). While American society was moving toward shorter labor contracts over the eighteenth and nineteenth centuries, the claim that this movement was caused by a fall in employer demand for long-­term contracts, rather than a fall in worker supply of long-­term contracts, is doubtful. Wage labor contracts were not universally superior to servant contracts. The ability to fire at will is typically cited as the key benefit to employers of wage

Farmer Merchant Unknown Gentlemen Baker Innkeeper Storekeeper Victualer Miller Widow Ironmaster Grocer Confectioner Papermaker Mariner Minister Brewer Tanner Doctor Glazier

0.4 1.1

4.0 0.4 0.4 1.5 0.7 2.2 0.4

0.4 2.2

38.2 23.3 15.6 1.1 1.5 1.5

1.2

1.3 2.0 0.5 0.5 0.3 0.5 0.5 0.5 0.5 0.9 1.2 0.3 0.9 1.5 0.6

35.7 12.6 20.6 2.8 2.8 1.5 1.9 2.8 0.9 2.2

1819 (N = 325)

23.4 20.9 5.0 16.3 3.0 5.3 2.8 1.0 1.0 1.5

Most frequently Buyers of German servants reported occupations November– 1818 December 1817 (N = 398) (N = 275)

2.7

13.5

13.5 10.8 18.9

1820–31 (N = 37) 30.8 18.6 13.2 7.4 2.9 3.0 1.6 1.4 1.3 1.3 1.1 1.0 0.9 0.9 0.8 0.8 0.7 0.7 0.5 0.5 0.8 1.6 0.5 0.9

20.8 18.1 11.8 6.1 2.3 5.2 1.9 0.7 3.0 0.5 1.3 1.3 0.4 0.9 1.6 1.2 0.9 1.1

1.4

7.3

21.3 4.5 5.7

Total Total Total 1817–31 1787–1804 1815–20 (N = 1,035) (N = 3,129) (N = 647)

0.9 0.4 0.8

2.1

4.5

27.3 0.9 10.5

Total 1793–1807 (N = 579)

German immigrants

continued

1.0 >0.8 0.8

2.5

43.2 3.1 >0.8 1.5 >0.8 1.4

pa rural census 1800 a

Table 17.5 Distribution of occupations among buyers of German immigrant servants, 1787–1831, with some comparative occupational distributions for immigrants and residents (in percentages)

100.0

2.9

1.1

0.4

0.7

100.0

0.5 0.5 0.3 9.0

0.5 0.3 0.8 0.5 0.8

100.0

8.7

0.3 0.3

0.3

1819 (N = 325)

100.0

5.4 27.1

8.1

1820–31 (N = 37)

100.0 92

0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.3 0.3 7.6 100.0 120

0.5 0.3 2.1 14.1

1.5 0.3 1.0 0.5

100.0 77

9.6

1.7 2.2

Total Total Total 1817–31 1787–1804 1815–20 (N = 1,035) (N = 3,129) (N = 647)

100.0 67

6.7

1.7 1.7

Total 1793–1807 (N = 579)

German immigrants

100.0 970

4.3

3.5 1.6

pa rural census 1800 a

Notes a N = 47,073. Soltow and Keller (1982) did not report the number of observations for occupations with less than 0.8 percent of the sample. b The other occupations recorded in the 1817–31 evidence on buyers of German servants included alderman, apothecary, army officer, auctioner, bookseller, bottler, brassfounder, bricklayer, broker, brushmaker, cabinetmaker, carpenter, carter, clerk, coachmaker, coachpainter, combmaker, conveyor, corker, customs official, dealer, distiller, drover, druggist, exchangebroker, ferrykeeper, flourmerchant, gilder, gunmaker, hairdresser, health officer, ironmerchant, ironmonger, justice, laceweaver, lock-smith, lumbermerchant, mayor, millwright, oakcooper, oilman, porter, postmaster, printer, rigger, sailmaker, sheet iron worker, ship broker, ship chandler, shoemaker, silversmith, single women, smith, soapboiler, stablekeeper, sugar boiler, tobacconist, watchdealer, weaver, and whitesmith.

Sources: “Book A of Redemptioners”; “Book C of Redemptioners”; Grubb 1985b: 268–9; Grubb 1990a: 432; Soltow and Keller 1982: 36.

Total Number of reported occupations

Attorney Cordwainer Hatter Blacksmith Cooper Pewterer Sugar refiner Tavernkeeper Tailor Otherb

Most frequently Buyers of German servants reported occupations November– 1818 December 1817 (N = 398) (N = 275)

Table 17.5 continued

Disappearance of immigrant servant markets   333 labor. This ability allowed employers to escape labor costs when demand for their product was low and to eliminate workers who performed poorly. However, wage labor could also be costly because workers could quit at will. Employers could never be sure if they would have enough labor to meet peak demand. Workers could quit suddenly and go to a competitor, or demand higher wages to stay. Employers may have hesitated to discipline or drive their employees to work harder for fear that they might quit. The incentives to invest in worker training were also lost because workers could quit and take the newly acquired training elsewhere. Finally, high labor turnover could disrupt production.14 For many employment situations, immigrant servant contracts had the advantages of wage labor contracts without the costs. If demand for the employer’s product was low or if the servant was poorly suited for the task required, the employer could rent or sell the servant in Philadelphia’s active resale market for   contract labor (Grubb 1989: 157). Because servants could not quit until the end of the contract, employers could be certain of their labor supply, could be certain of capturing some of the benefits of training their servants to be more  productive, and could discipline and drive their servants to work harder. Servant labor could also be shifted across different tasks more readily. Wage workers may have resisted such multitasking as a violation of their initial task-­ specific contract. Therefore, whether wage labor or servant labor was superior depended on the employer’s circumstances.15 Given that employers had a variety of labor needs, a mix of workers with different contract lengths should be observed. An example of employers choosing such a mix is shown by Clemens and Simler (1988: 106–43) for Chester County, Pennsylvania, from 1750 through 1820. The secular trend away from long-­term contracts was caused by workers and not employers (Steinfeld 1991; Jacoby 1991a, 1991b). For similar compensation, American workers increasingly chose wage labor over servant contracts in order to avoid the paternalistic control exercised by masters over their servants. Immigrants in debt for their passage fare, however, did not have this choice. Finally, the desire of employers for longer contracts given worker preferences for shorter contracts can be seen by the increasing use over the course of the nineteenth century of performance bonding, efficiency wages, and internal labor markets that turned short-­term contracts into implicit long-­term worker-­employer attachments (Sundstrom 1988). The hypothesis that immigrant servitude ended because employers abandoned servitude for wage labor can be directly tested by measuring the change in immigrant servant prices. If the market ended because the demand for servants fell as employers shifted to wage labor, then the price of servants would be bid down until Europeans would no longer be willing to finance emigration through servitude (see Figure 17.1). By contrast, if servant prices rose, then the end of immigrant servitude must have been caused by a decline in servant supply rather than a decline in employer demand. The direct test of the wage labor hypothesis is also an indirect test of the previous four hypotheses. All five hypotheses presented here are what economists

Price per year of service

334   German immigrant servitude in America, 1745–1835

European suppty 1817–20 Price (1817) Hypothesized prediction Price (1820)

American demand 1817 American demand 1820

0 {known facts} Quantity (1820) Quantity (1817) Quality of European immigrant servants

Figure 17.1  The demand hypothesis of the end of European immigrant servitude. Notes The demand hypothesis accepts the fall in the quantity of servants as fact and predicts that the price of servants must fall as a consequence. As shown in Table 17.1 and the text, this prediction fails. Therefore, any demand-driven hypothesis of the end of servitude lacks merit.

call demand-­side explanations. Each of these five explanations assumes that European immigrant servitude disappeared because American demand for ser­ vants declined, albeit for different reasons. The evidence below rules out any demand explanation of the end of immigrant servitude and, therefore, rules out all five explanations, including any combination thereof. It also rules out any new explanation based on some yet unknown legal or extralegal impediment that lowered the American demand for immigrant servants. Evidence on the per-­year cost of adult servants can be taken from Table 17.1. Assuming that maintenance costs (principally food and shelter) changed little in the three years from 1817 to 1820, and given that the evidence indicates that freedom dues changed little between 1817 and 1820, changes in the per-­year cost of servants can be calculated by dividing the average contract length by the average deflated contract price. This calculation shows that the per-­year price of adult servants did not fall but actually rose as the market collapsed. The average per-­year price of single adult male servants in deflated U.S. dollars was 24.7 in 1817, 26.6 in 1818, 36.2 in 1819, and 36.6 in 1820. The average per-­year price of married servants was 19.2 in 1817, 23.0 in 1818, 31.7 in 1819, and 31.0 in 1820. The only way for prices to rise as quantity falls is for leftward shifts in supply to dominate leftward shifts in demand. Therefore, any demand explanation of the collapse of European immigrant servitude as illustrated in Figure 17.1 is not supported by the evidence.

Disappearance of immigrant servant markets   335

Conclusions The evidence from the port of Philadelphia indicates that German immigrant servitude collapsed abruptly in 1820. Five popular explanations as to why this happened, namely imprisonment for debt, the work of immigrant aid societies, the frequency of servants running away, the financial panic of 1819, and the efficiency of wage labor, were all shown to be inconsistent with the evidence. Because servant prices rose as servant quantities went to zero, any movement in demand must have been dominated by a fall in supply. The explanation for the disappearance of European immigrant servitude in the United States during the nineteenth century must resemble the explanation for the transition from ser­ vants to slaves in the Chesapeake at the beginning of the eighteenth century, where scholars have shown that planters shifted to slave labor because the supply of servants declined, not because the demand for servants declined (Menard 1977). In both cases, America did not abandon European immigrant servitude; immigrant servants abandoned America. Therefore, future research into why European immigrant servitude disappeared should be redirected toward investigating the forces affecting servant supply rather than American demand. Exactly why the supply of European immigrant servants fell after 1819 is unclear. The answer most often given in the literature, that passage fares fell and European wealth rose such that immigrants could finance their migration with cash rather than through servitude, is also inconsistent with the evidence in Table 17.1.16 First, the abruptness of the end of immigrant servitude indicates that changes in European wealth could not have caused the end of immigrant servitude. Wealth could not have risen enough in one year, from 1819 to 1820, to make all immigrants instantly rich enough to purchase their passage with cash. Second, the decline in passage fares did not occur until after 1829, a full decade after immigrant servitude disappeared (Hansen 1940: 198; Page 1911: 738). The passage debts incurred by German redemptioners equaled their passage fare, a 15 percent markup as interest and risk compensation on the loan, plus additional preembarkation expenses (Grubb 1986). From 1817 through 1821 these passage debts were at their highest level ever (Table 17.1; Grubb 1992b: 184). High passage costs increased servitude because fewer immigrants could pay all their migration expenses in cash. Thus, immigrant servitude disappeared exactly when rising passage costs should have increased servitude. While this simple supply-­side explanation might explain why European immigrant servitude could not be revived later in the century, it cannot explain the immediate end of European immigrant servitude in 1820. The supply-­side answer must lie elsewhere, most likely in the complex configuration of events affecting German migration choices. The unique historical events that shaped the relative magnitudes and directions of German emigration between 1815 and 1830 and changes in the availability of alternative financing must have played a critical role. Given that demand explanations have been discredited, it is to such supply-­side complexities that scholars should now turn their research efforts.

336   German immigrant servitude in America, 1745–1835

Notes   * Originally published as Farley Grubb, “The disappearance of organized markets for European immigrant servants in the United States: five popular explanations reexamined,” Social Science History, vol. 18, no. 1 (Spring 1994): 1–30. Copyright 1994. © 1994 by the Social Science History Association. All rights reserved. Reprinted with permission of the publisher, Duke University Press. CCC 0145–5532/94/$1.50. F. W. Grubb is professor of economics at the University of Delaware, Newark. His field of research is immigration and immigrant labor markets in North America from 1650 to 1835, and he has recent publications in this and many other journals. Earlier versions of this article were presented in the Economic History Workshop at the University of Chicago, at the annual meetings of the Social Science History Association, and in the Economics Seminar at the University of Delaware. The author thanks the participants of these seminars and Rick Agnello, Raymond Cohn, John Craig, David Eltis, Stanley Engerman, Joseph P. Ferrie, Robert W. Fogel, David W. Galenson, Henry Gemery, Dirk Hoerder, J. R. T. Hughes, Ken Koford, Daniel Levy, David Mitch, A. G. Roeber, Theodore W. Schultz, Ralph Shlomowitz, Richard Steckel, and Robert Whaples for helpful information and comments on earlier drafts. In addition, the author thanks the Pennsylvania Historical Society and the National Archives for making microfilm copies of manuscript records available. Financial support from the Economic History Association Cole Grant-­in-Aid and from the University of Delaware General University Research Program is gratefully acknowledged. Paul O’Neill and Anne Pfaelzer de Ortiz provided research assistance, and Anne Pfaelzer de Ortiz provided editorial assitance.   1 For the classic works on immigrant servitude, see Diffenderffer 1899; Geiser 1901; Herrick 1926; McCormac 1904; Morris 1981; Smith 1947.   2 Relative labor scarcity influenced regional patterns of American industrial development (Goldin and Sokoloff 1984; David 1987). If a large portion of the nineteenth-­ century immigrants had been under alienable servant contracts and sold at will by American employers, then the regional patterns of labor scarcity, and thus of industrial development, may have been significantly altered.   3 Direct quantitative evidence on British servant emigration to the United States after 1784 could not be found and may not exist. Most of the servant emigration from Britain in this period was from Ireland. It appears from the secondary literature that Irish servant immigration was in decline by the 1790s and may have been, at best, a trickle by 1818. It clearly ended before the demise of German immigrant servitude. See Adams 1932: 113–14; Jameson 1896, 1897; Jones 1969: 51–62.   4 See Hansen 1940: 85–123; Henninghausen 1909: 17–96; “Philadelphia Customs House Records”; Strassburger 1934; Walker 1964: 2–39.   5 The two surviving registers are “Book A of Redemptioners” and “Book C of Redemptioners.” Apparently “Book B of Redemptioners,” which must have covered from 1804 until November 1817, is lost.   6 Children comprised only 25 percent of German immigrant servants between 1785 and 1804 (Grubb 1985a: 320).   7 See note 4.   8 For example, in the state of Delaware the last officially recorded servant contract for an indigent child in New Castle, Kent, and Sussex counties was in 1900, 1926, and 1930, respectively (Hancock 1974).   9 See Diffenderffer 1899: 260–7; Geiser 1901: 59–70; Grubb 1988; Henninghausen 1909: 17–28, 56–86; Herrick 1926: 263, 286–308; Jameson 1896: 472–3; Jones 1969: 57; McCormac 1904: 110; Pfund 1964: 4–5; Risch 1936: 15–33; Wust 1958: 2–16, 41. Examples of servants suing masters and winning can be found in the county court records of Maryland and Pennsylvania. 10 See Bailyn 1986: 64; Grubb 1989: 159, 1992a; Morris 1981: 391–531; Salinger 1987:

Disappearance of immigrant servant markets   337 133; Smith 1947: 226–84. Servants were also deterred from escape because courts enforced contracts, with the penalty for running away being two days’ extra service for every day absent plus additional time to cover the cost of capture and court fees. In addition, the probability of capture was high because vagrants could be jailed. Jail keepers would advertise these individuals as possible runaways and request their masters “if any” to come and pay the jail fee and any rewards for their capture. 11 See the notes to Table 17.2. While newspaper advertisements are not a comprehensive record of runaways, they are the best source available. A search of county court records yielded little information on runaway servants. While advertised runaways are most likely a biased downward estimate of total runaways, the trend in advertised runaways is less likely to be biased. Because the runaway hypothesis is based on changes in the trend, measuring the number of runaways from newspaper advertisements over time should provide an adequate test of the hypothesis. These newspapers were chosen because they represented the primary outlet for advertising runaways for which a continuous record has survived. Other newspapers in the Philadelphia region, including German-­language papers, published few runaway advertisementss that were not duplicated in these newspapers. Masters advertised their runaways in these two Philadelphia newspapers because servants in the region were thought to be especially inclined to run away to Philadelphia and because these newspapers circulated far beyond the Philadelphia region, would publish advertisements in English or German, and specialized in these kinds of advertisements. The fact that masters from distant counties and states, such as New York, Virginia, and most notably Lancaster County, Pennsylvania, and Baltimore County, Maryland, advertised their runaways in these two newspapers indicates that the runaway advertisements from these two newspapers comprised a substantial proportion of the total runaways advertised (Grubb 1992a). This fact also indicates that the estimated runaway rate is biased high, because some of the servants advertised could have entered through other ports, such as Baltimore. In addition, removing the 22.6 percent of the servants sold to masters from distant states (see Table 17.3) from the denominator used in Table 17.2, line 4, does not appreciably affect the magnitude or trend of the percentages reported. It only raises the percentage reported for 1817–20 by one point. Finally, these estimates are for runaways rather than successful escapes. How many advertised runaways were caught is unknown, but many were (see the quote by Daniel Kent above). Therefore, the estimated incidence of running away is a biased high estimate of the incidence of successful escapes. 12 The t-­statistic for the difference in mean passage debts between 1817 and 1819 is 3.18 (significant above the 0.005 level). 13 The t-­statistic for the difference in mean contract lengths between 1817 and 1819 is 2.46 (significant above the 0.01 level). 14 If work was strictly seasonal, employers would still have to pay wage workers enough to carry them through the slack times or else they would not stay for or survive until the next season. Therefore, wage labor did not generate net savings to society over servant labor in terms of economizing on off-­season maintenance expenses. 15 In fact, preliminary estimates indicate that the demand for immigrant servants was highly elastic, thus indicating that employers in general found servant and wage labor interchangeable (Grubb 1992b: 199). 16 For the most recent discussion of these causes see Galenson (1984).

References Adams, W. F. (1932) Ireland and Irish Emigration to the New World from 1815 to the Famine. New Haven, CT: Yale University Press. Bailyn, B. (1986) The Peopling of British North America. New York: Alfred A. Knopf.

338   German immigrant servitude in America, 1745–1835 Barnard, E. K. (1904) Letters and Other Papers of Daniel Kent Emigrant and Redemptioner. Baltimore: New Era Printing. Barzel, Y. (1989) Economic Analysis of Property Rights. New York: Cambridge University Press. Bezanson, A., R. D. Gray, and M. Hussey (1936) Wholesale Prices in Philadelphia, 1784–1861. Philadelphia: University of Pennsylvania Press. “Book A of redemptioners, 1785–1804.” Manuscript, Historical Society of Pennsylvania, Philadelphia. “Book C of redemptioners, 1817–1831.” Manuscript, Historical Society of Pennsylvania, Philadelphia. Carlton, F. T. (1908) “Abolition of imprisonment for debt in the United States.” Yale Review, O. S. 17: 339–44. Clark, D. (1977) “Babes in bondage: Indentured Irish children in Philadelphia in the nineteenth century.” Pennsylvania Magazine of History and Biography 101: 475–86. Clemens, P. G. E., and L. Simler (1988) “Rural labor and the farm household in Chester County, Pennsylvania, 1750–1820,” in S. Innes (ed.) Work and Labor in Early America. Chapel Hill: University of North Carolina Press: 106–43. Cloud, P., and D. W. Galenson (1987) “Chinese immigration and contract labor in the late nineteenth century.” Explorations in Economic History 24: 22–42. David, P. A. (1987) “Industrial labor market adjustments in a region of recent settlement: Chicago, 1848–1868,” in P. Kilby (ed.) Quantity and Quiddity: Essays in U.S. Economic History. Middletown, CT: Wesleyan University Press: 47–97. Davis, L. E., and J. R. Hughes (1960) “A dollar-­sterling exchange, 1803–1895.” Economic History Review 13: 52–78. Diffenderffer, F. R. (1899) “The German immigration into Pennsylvania through the port of Philadelphia, and ‘The Redemptioners.’ ” Pennsylvania German Society 10: 1–328. Erickson, C. (1984) “Why did contract labour not work in nineteenth-­century United States?” in S. Marks and P. Richardson (ed.) International Labour Migration, Historical Perspectives, vol. 24. Hounslow, England: Maurice Temple Smith: 34–56. Galenson, D. W. (1981) White Servitude in Colonial America. New York: Cambridge University Press. Galenson, D. W. (1984) “The rise and fall of indentured servitude in the Americas: An economic analysis.” Journal of Economic History 44: 1–26. Geiser, K. F. (1901) Redemptioners and Indentured Servants in the Colony and Commonwealth of Pennsylvania. New Haven, CT: Tuttle, Morehouse, and Taylor. Goldin, C., and K. Sokoloff (1984) “The relative productivity hypothesis of industrialization: The American case, 1820–1850.” Quarterly Journal of Economics 99: 461–87. Grubb, F. (1985a) “The incidence of servitude in trans-­atlantic migration, 1771–1804.” Explorations in Economic History 22: 316–39. Grubb, F. (1985b) “Immigrant servant labor: Their occupational and geographic distribution in the late eighteenth-­century mid-­Atlantic economy.” Social Science History 9: 249–75. Grubb, F. (1986) “Redemptioner immigration to Pennsylvania: Evidence on contract choice and profitability.” Journal of Economic History 46: 407–18. Grubb, F. (1988) “The auction of redemptioner servants, Philadelphia, 1771–1804: An economic analysis.” Journal of Economic History 48: 583–603. Grubb, F. (1989) “Servant auction records and immigration into the Delaware valley, 1745–1831: The proportion of females among immigrant servants.” Proceedings of the American Philosophical Society 133: 154–69.

Disappearance of immigrant servant markets   339 Grubb, F. (1990a) “German immigration to Pennsylvania, 1709–1820.” Journal of Interdisciplinary History 20: 417–36. Grubb, F. (1990b) “The transition from colonial currency to the U.S. dollar: A market revealed analysis.” Working Paper No. 90–20. Economics Department, University of Delaware. Grubb, F. (1992a) Runaway Servants, Convicts, and Apprentices Advertised in the Pennsylvania Gazette, 1728–1796. Baltimore: Genealogical Publishing. Grubb, F. (1992b) “The long-­run trend in the value of European immigrant servants, 1654–1831: New measurements and interpretations.” Research in Economic History 14: 161–234. Hancock, H. B. (1974) “The indenture system in Delaware, 1681–1921.” Delaware History 16: 47–59. Hansen, M. L. (1940) The Atlantic Migration, 1607–1860. Cambridge, MA: Harvard University Press. Henninghausen, L. P. (1909) History of the German Society of Maryland. Baltimore: Sun Job Printing. Herrick, C. A. (1926) White Servitude in Pennsylvania. Philadelphia: John Joseph McVey. Jacoby, D. (1991a) “The transformation of industrial apprenticeship in the United States.” Journal of Economic History 51: 887–910. Jacoby, D. (1991b) “The legal foundations of human capital markets.” Industrial Relations 30: 229–50. Jameson, J. F., ed. (1896) “Letters of Phineas Bond, British consul at Philadelphia, to the Foreign Office of Great Britain, 1787–1794.” Annual Report of the American Historical Association 1: 513–659. Jameson, J. F. (1897) “Letters of Phineas Bond, British consul at Philadelphia, to the Foreign Office of Great Britain, 1787–1794.” Annual Report of the American Historical Association 2: 454–568. Jones, M. A. (1969) “Ulster emigration, 1783–1815,” in E. E. R. Green (ed.) Essays in Scotch–Irish History. London: Routledge and Kegan Paul: 46–68. McCormac, E. I. (1904) White Servitude in Maryland, 1634–1820. Baltimore: Johns Hopkins University Press. McCusker, J. J. (1978) Money and Exchange in Europe and America, 1600–1775: A Handbook. Chapel Hill: University of North Carolina Press. Menard, R. R. (1977) “From servants to slaves: The transformation of the Chesapeake labor system.” Southern Studies 16: 355–90. Morris, R. B. (1981) Government and Labor in Early America. Boston: Northeastern University Press. Muhlenberg, H. M. (1942) The Journal of Henry Melchior Muhlenberg. Philadelphia: Muhlenberg Press. Officer, L. H. (1983) “Dollar-­sterling mint parity and exchange rates, 1791–1834.” Journal of Economic History 43: 579–616. Page, T. W. (1911) “The transportation of immigrants and reception arrangements in the nineteenth century.” Journal of Political Economy 19: 732–49. Pfund, H. (1964) History of the German Society of Pennsylvania, 1764–1964. Philadelphia: German Society of Pennsylvania. “Philadelphia customs house records, 1800–1883.” Manuscript, National Archives, Washington, dc, microfilm #425, rolls 16–50. Risch, E. (1936) “Immigrant aid societies before 1820.” Pennsylvania Magazine of History and Biography 60: 15–33.

340   German immigrant servitude in America, 1745–1835 Salinger, S. V. (1987) To Serve Well and Faithfully: Labor and Indentured Servitude in Pennsylvania, 1682–1800. New York: Cambridge University Press. Smith, A. E. (1947) Colonists in Bondage. New York: W. W. Norton. Soltow, L., and K. W. Keller (1982) “Rural Pennsylvania in 1800: A portrait from the septennial census.” Pennsylvania History 49: 24–47. Steinfeld, R. J. (1991) The Invention of Free Labor. Chapel Hill: University of North Carolina Press. Strassburger, R. B. (1934) Pennsylvania German Pioneers. Norristown, PA: Pennsylvania German Society. 3 vols. Sundstrom, W. A. (1988) “Internal labor markets before World War I: On-­the-job training and employee promotion.” Explorations in Economic History 25: 424–45. Trautmann, F. (1981) “Pennsylvania through a German’s eyes: The travels of Ludwig Gall, 1819–1820.” Pennsylvania Magazine of History and Biography 105: 35–65. Virchaux, H. T. (1818) “Report to the German Society of Pennsylvania on the ship April.” Manuscript, German Society of Pennsylvania, Philadelphia. Walker, M. (1964) Germany and the Emigration, 1816–1885. Cambridge, MA: Harvard University Press. Wust, K. G. (1958) Pioneers in Service: The German Society of Maryland, 1783–1958. Baltimore: German Society of Maryland.

18 The collapse of the German immigrant servant market Timing and causes*

In North America, organized markets for European immigrant servants flourished for two centuries. The resilience of these markets is quite remarkable. They survived the turbulent events of the seventeenth and eighteenth centuries, including frequent and severe interruptions to transatlantic trade caused by wars and depressions, competition in America from slave and native-­born free labor, and the change in legal sovereignty caused by the American Revolution. During the eighteenth century, roughly one-­half of the Europeans landing in British North America paid for their passage by selling contracts on their future labor.1 Sometime between 1800 and 1830, however, organized markets for European immigrant servants disappeared. Although there were some instances of European immigrant servitude after 1830, they were few in number and not of sufficient volume to sustain an organized market. By the time of the great acceleration in European emigration to the United States in the second quarter of the nineteenth century, the payment of transatlantic passage through servitude had disappeared.2 The Europeans who settled the western frontier, built the canals and railroads of the nation, and manned the eastern factories were free artisans, farmers, and wage laborers. If half had been servants, as during the eighteenth century, the course of U.S. development after 1830 would have been substantially different, because servitude had decisive effects on the geographic distribution and social integration of immigrants into the American work force.3 The end of immigrant servitude was a watershed event in the history of European emigration to America and in the history of American social and economic development. Given the importance of European immigrant servitude, it is surprising how little we know about when and why it ended. Scholars have advanced many hypotheses, but no consensus has emerged. When evidence is lacking, explanations can proliferate. Scholars have hypothesized that European immigrant servitude disappeared because wage labor was superior, because servants escaped too easily, because passage fares fell, because of economic recessions, because of the abolition of debtors prisons, because of legal interference from immigrant-­ aid societies, because immigrants became wealthier, and so on.4 This article will provide quantitative evidence on when the organized market for immigrant ser­ vants disappeared and on how that market changed over its final 60 years. This

342   German immigrant servitude in America, 1745–1835 evidence will eliminate many of the popular hypotheses in the literature, and I will use it to craft an alternative explanation for why servitude disappeared. Studies measure the importance of immigrant servitude in four ways: as a percentage of the American population, as a percentage of the bound- or slave-­ labor force, as a percentage of arriving immigrants, and as a volume sufficient to sustain an organized market in alienable immigrant-­servant contracts. The first two measures are closely related and have been extensively studied. By 1710 the yearly immigration of European servants was of minor importance compared with free and slave populations.5 Organized markets for European immigrant servants, however, continued for at least another century, with servants comprising a significant portion of arriving immigrants and in large enough numbers to sustain a market in servant contracts. It is the disappearance of immigrant servitude as measured by these two criteria that this article will address.6

The data Tracing the demise of immigrant servitude requires several decades of data on the volume of free versus servant passengers. Unfortunately, little quantitative evidence for U.S. immigration exists for the first four decades of the Republic. Most ports did not systematically collect such information, and the federal government did not require its collection until 1819. The data that were recorded, including the federal data, typically excluded information on servitude. The exception was at the port of Philadelphia, where officials collected information on both immigration and servitude for German passengers. Before 1800 the Pennsylvania government required passenger manifests from ships landing German immigrants, and after 1800 from all ships landing passengers, a substantial portion of whom were Germans. Between 1800 and 1811, Germans comprised 42 percent of all immigrants arriving by ship from northern Europe and Canada, and between 1812 and 1823 they comprised 28 percent.7 Although these manifests contained no information on servitude, other port registers recorded the sale of German immigrant-­servant contracts including information about the buyer, the servant, and the contract terms of each sale. In 1772 officials recorded all servant contracts, including those of German immigrants. With the onset of hostilities caused by the American Revolution, immigration ceased and the port discontinued servant registration.8 Because German servant immigration resumed after the war, the German Society of Pennsylvania convinced the state, beginning in 1784, to again register the sale of German immigrant servants. Two volumes of this latter registration have survived, covering 1785 through 1804, and November 1817 through 1831. Although the intervening volume is lost, researchers can estimate the number of servants in this volume using information from the two surviving volumes.9 All German immigrant ser­vants were “redemptioners.” They signed a debt contract with a shipper in Europe that required them to pay a stipulated passage fare before debarkation in America, even if it meant negotiating and selling a servant contract on their own labor to raise the required sum.10

Collapse of German immigrant servant market   343 All contracting parties benefitted from servant registration. The official in charge was bilingual in English and German and provided translation for the contracting parties, and the recorded contracts provided legal protection both for the employer and the servant against unilateral attempts to change the contract. Unregistered contracts were illegal. The government did not use the register for purposes that would be costly to the contracting parties, such as taxation. Contract negotiations took place on immigrant ships immediately after they docked. Thus, the market was centralized and predictable, which made registration convenient. Even buyers from distant states registered their servants before departing the city. The net benefits of registration, along with its convenience, suggest that this record provides a reasonably comprehensive count of German immigrant servants.11 By combining the servant register with the passenger manifests, researchers can estimate the percentage of German immigrants entering servitude to pay for their transatlantic passage to Philadelphia from 1772 through 1835. Although these data are the only post-­Revolution immigrant stream for which we can make such a calculation, they should be representative of the major trends in immigrant servitude between 1772 and 1835.12 Indirect evidence indicates that the last manifestation of European immigrant servitude of any magnitude was in the German passenger trade, and Philadelphia was the principal port of entry for German immigrants before the 1830s.13

When did European immigrant servitude to the U.S. disappear? Table 18.1 records the yearly flow of German immigrants and of German immigrant servants through the port of Philadelphia from 1772 to 1835. Given the paucity of quantitative evidence on immigration during the early Republic, the data this chapter presents on German immigration are in themselves an important contribution.14 Between 1785 and 1815 the periodic wars of the French Republic and the American embargo of 1808 constrained German immigration. After the end of hostilities in 1815, German immigration surged to over 4,000 in 1817. This exceeded the annual averages of 1,431 during the Peace of Amiens (1802 to 1805), and 910 during the early 1770s. The year 1817 represented the largest yearly German immigration through Philadelphia since the great migration between the Wars of the Austrian Succession and the Seven Years’ War (1749 to 1755), when it averaged 5,755 per year.15 The release of migration demand pent-­up during the Napoleonic Wars and one of the worst agricultural catastrophes ever in the southern Rhineland, known as “the year without a summer,” caused the peak in 1817.16 German migration to the United States dropped off during the 1820s as prosperity returned to the Rhineland and as other destinations, such as eastern Europe and Brazil, temporarily became more attractive. By 1830, German migration through Philadelphia had returned to normal levels. The number of servants among these German immigrants averaged 191 per year between 1785 and 1804. After peace returned in 1815, the volume was

1772 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807

Year

863 654 170 363 159 116 61 370 578 413 795 744 1,050 435 356 435 466 630 1,338 1,200 1,888 1,297 1,268 956

479 323 99 238 15 46 25 167 236 118 160 346 305 22 54 25 115 36 435 418 645 360 352 265

55.5 49.4 58.2 65.6 9.4 39.7 41.0 45.1 40.8 28.6 20.1 46.5 29.1 5.1 15.2 5.8 24.7 5.7 32.5 34.8 34.2 27.8 27.8 27.8

   –    –    –    –    –    –    –    –    –    –    –    –    –    –    –    –    –    –   60.77   67.04   55.62    –   62.42    –

Total (#) Servant (#) Servant (%) Free adult passage fares ($)

Table 18.1 German immigration to Pennsylvania, 1772–1835

160 103   52   94   13   14   16   77 116   55   86 125   99    3    5    9   63    8 260 206 192    –    –    –

3.94 – – 3.52 3.45 3.54 3.04 2.93 3.15 3.19 2.42 2.57 2.92 3.00 2.50 2.70 2.89 3.17 3.14 3.32 3.23 – – –

  58.62    –    –   78.14   80.14   76.64   78.43   80.25   81.84   79.53   73.37   62.33   71.63   56.39   79.76   92.92   85.03   90.34   91.29   88.75   82.41    –    –    –

Total (#) Average contract Average length (years) contract price ($)

Single adult male servants

  22   52   10   34    0    4    0   26   34    6   18   64   66   10    2    6   18   10   62   58 122    –    –    –

4.30 – – 3.69 – 3.50 – 3.35 3.57 4.00 2.81 2.99 3.14 2.45 3.50 3.50 3.22 3.05 3.44 3.27 3.25 – – –

50.33  –  – 66.45  – 63.18  – 83.79 77.88 75.07 72.22 56.69 68.42 35.04 72.52 86.86 76.92 92.21 84.23 77.94 68.31  –  –  –

Total (#) Average contract Average length (years) contract price ($)

Married servants

1808 188 1809 22 1810 73 1811 5 1812 8 1813 0 1814 0 1815 99 1816 1,304 1817 4,481 1818 947 1819 1,573 1820 118 1821 105 1822 52 1823 121 1824 86 1825 107 1826 200 1827 292 1828 467 1829 170 1830 684 1831 1,592 1832 1,003 1833 160 1834 252 1835 33 1785–1835 Totals 29,884

8,105

52 6 20 1 2 0 0 40 528 1,890 396 326 18 5 0 1 4 2 1 2 1 1 0 2 0 0 0 0

27.8 27.8 27.8 27.8 27.8 – – 40.4 40.5 42.2 41.8 20.7 15.3 4.8 0.0 0.8 4.7 1.9 0.5 0.7 0.2 0.6 0.0 0.1 0.0 0.0 0.0 0.0

   –    –    –    –    –    –    –   60.00    –    –   68.07   76.57   79.42    –    –   80.32 109.10    –    –    –    –   68.11    –   46.50    –    –    –   22.40

   –    –    –    –    –    –    –    –    –   54 106   84    4    1    –    –    –    –    –    –    –    –    –    –    –    –    –    –

– – – – – – – – – 3.14 3.07 2.79 2.50 2.00 – – – – – – – – – – – – – –

   –    –    –    –    –    –    –    –    –   77.53   81.81 100.87   91.62 146.96    –    –    –    –    –    –    –    –    –    –    –    –    –    –

   –    –    –    –    –    –    –    –    –   76   78   32    6    –    –    –    –    –    –    –    –    –    –    –    –    –    –    –

– – – – – – – – – 3.16 3.12 2.76 2.50 – – – – – – – – – – – – – – –

continued

 –  –  –  –  –  –  –  –  – 60.59 71.62 87.62 77.62  –  –  –  –  –  –  –  –  –  –  –  –  –  –  –

Notes The passage fares are for adults in steerage who paid in advance of embarkation and included meals (fares without meals averaged 62.5 percent of fares with meals). The average contract price included any contracted cash payments made to the servant during or upon completion of the contract; see note 24. The number of immigrants on missing passenger manifests for September 1803; June 15 to July 1, 1805; July 1807; September 1 through 14, 1810; September 1811; and October 1818 was calculated by using the average relative percentage of immigrants arriving during that period in the preceding and following years. For passenger manifests that only listed household heads, the total number of immigrants was estimated by using average family size calculated from manifests that enumerated family size. German immigration was seasonal, with the major break occurring in late winter. Servant sales lagged immigrant arrivals by up to a month. Therefore, both the immigration and servant data were broken into years that ended sometime between February and April. The total numbers of servants reported under the Single Adult Male Servant and the Married Servant columns only count those contracts found directly in the servant registers and only those with complete information on contract length and price (except for 1785 and 1786, when no prices were recorded). Therefore, the number in these categories do not always represent the true percentage of these servants among all German servants arriving in Philadelphia in that year. “Book B of Redemptioners,” which must have covered from 1805 to November of 1817, is lost. The number and distribution of servants in Book B was estimated as follows. The number of pages in Book B was established using page references to Book B found in Book C. This comparison indicated that pages 1 through 100 in Book B registered servants arriving from 1805 to the War of 1812, and pages 101 through 327 in Book B registered servants arriving after the War of 1812 through October 1817. The same official, as judged by the constancy of the handwriting, registered servants for numerous years in succession. Therefore, the number of servants per page in the first 100 pages of Book B was assumed to be the same as the average for the last four years of Book A, and the number of servants per page in the last 226 pages of Book B was assumed to be the same as the average for the first four years of Book C. Finally, this estimated total number of servants in Book B was distributed by the yearly magnitude of German immigration found in the passenger manifests. Exchange rates were taken from The Citizen and Farmer’s Almanac for the years 1802–6 (1 florin = $0.39, and 1 guinea = $4.66); The Directory and Stranger’s Guide for the years 1815–23 (1 florin = $0.40); Grubb, “Transition”; McCusker, Money and Exchange: 186; Nile’s Weekly Register [September 27, 1817] (5 francs = $0.933); Roebling, Diary: 7–9 (1 reichsthaler = $0.813); and Officer, “Dollar–Sterling Mint Parity,” 579–616. All nominal values were deflated (1795 = 100) following Bezanson et al., Wholesale Prices in Philadelphia: 392; and Schumpeter, “English Prices,” 35.

Sources: Total German immigration was derived from “Account of Vessels”; American Watchman, January 7, 1818; Nile’s Weekly Register, Jan. 31, 1818; “Passenger List of the Ship Elizabeth”; “Philadelphia Custom House Records”; “Pilot’s Report for the Port of Philadelphia”; Strassburger, Pennsylvania German Pioneers; Virchaux, “Report to the German Society”; and Wokeck, “Flow and Composition,” 261. German servant immigration, contract length, and contract price data came from American Watchman, January 7, 1818; “Book A of Redemptioners”; “Book C of Redemptioners”; Grubb, “Servant Auction Records,” 155–8, and German Immigrant Servant Contracts; and “Record of Indentures.” The passage-fare data came from Anonymous, Nachrichten und Erfahrungen, as communicated to me by Dirk Hoerder and Agnes Bretting of the University of Bremen; Billigmeier and Picard, Old Land and the New: 70, 190; Duden, Report: 25; Hansen, Atlantic Migration: 198; Page, “Transportation,” 738; “Passenger List of the Ship ‘Elizabeth’,” 257–60; “Philadelphia Custom House Records,” the Ship Elizabeth, Aug. 5, 1815; Roebling, Diary: 3, 67; Strassburger, Pennsylvania German Pioneers, vol. 3: 112–14, 131–8; Von Furstenwarther, Der Deutsche in America, as communicated to me by Hoerder and Bretting; and Wittke, We Who Built America: 118.

Collapse of German immigrant servant market   347 higher than during the first two decades of the Republic, averaging well over 300 per year from 1816 through 1819. This is comparable to the volume of the pre-­ Revolutionary period when German servant immigration averaged 353 per year from 1770 through 1775.17 Therefore, the volume of servants arriving after 1815 must have been sufficient to sustain an organized market in servant contracts comparable with prior decades. In 1820, however, the volume of servants suddenly collapsed, never to recover again. Only 18 debarked at Philadelphia in 1820, and officials registered only another 19 between 1821 and 1831. The latter were almost all children. In 1831 the register was discontinued.18 The sudden disappearance of German immigrant servitude at the ports of Baltimore and New Orleans also occurred in 1820. Thus the evidence indicates that a sufficient volume to sustain an organized market in European immigrant servants disappeared abruptly in 1820. Immigrants’ reliance on servitude stopped at about the same time and almost as abruptly as the collapse in the volume of servants. Figure 18.1 presents the percentage of immigrants entering servitude from 1772 through 1835. The percentage fell from 56 percent in 1772, to 38 percent from 1785 through 1796, and then to 27 percent from 1797 through 1812. With the return of peace in 1815, the servant percentage rose, averaging 42 percent from 1815 through 1818. 100 90 80

Three-year moving average

Percentage

70 60

Yearly percentage

50 40 30 20 10 0

1775

1785

1795

1805

1815

1825

1835

Year of arrival

Figure 18.1 Percentage of servants among German immigrants to Pennsylvania, 1772–1835 (source: Table 18.1). Note The three-year moving average is weighted by the number of observations per year. Breaks in the data are connected by straight lines.

348   German immigrant servitude in America, 1745–1835 Thereafter, the servant percentage plummeted from 42 percent in 1818 to 21 percent in 1819, 13 percent in 1820, 7 percent in 1821, and thereafter only exceeded 1 percent in 1824 (5 percent) and 1825 (2 percent). After 1820, servants never again represented a significant portion of European migration to America. Although the volume of servants prior to 1820 fluctuated widely from year to year and even frequently fell to zero when circumstances temporarily halted migration, the permanent collapse in the servant percentage after 1818 was unprecedented. The only similar movement was from 1785 to 1801, when the percentage declined from 55 to under 15 percent. That downturn was only temporary, however. By 1818 the servant percentage had reached levels comparable with those in the 1770s and 1780s. The meager volume of servants after 1819 represented more than just a fall in immigration. It represented a permanent shift by immigrants away from servitude as a means of financing transatlantic migration. The evidence in Figure 18.1 reaffirms the conclusion reached by other scholars that regulatory or legal proscription did not cause the end of immigrant servitude. Pennsylvania and Maryland had regulated immigrant servant markets since the mid-­eighteenth century, but these regulations never caused a significant constriction of the volume of servant immigration.19 The only new regulation enacted after 1815 required employers to provide children with six weeks schooling for every year of their servitude. It is unlikely that this regulation caused any constraint on servant immigration. Employers and servants had already been voluntarily negotiating the provision of education in 70 percent of the contracts involving minors decades before it became mandatory.20 Parents and children had separate contracts, with the children’s contracts paying for their own passage and education expenses. To compensate employers for the additional schooling expense, the contracts of minors were lengthened. The education regulation, therefore, did not directly affect the supply or the demand for adult servants. In addition, because children were tied movers, it seems unlikely that any costs imposed by this education regulation would reduce their level of immigration. In fact, the relative share of children among German immigrant servants actually increased after 1815, from 25 percent during 1785 through 1804, to 42 percent from 1817 through 1820. By reducing uncertainty over the provision of education, the regulation may have increased the willingness of parents to indenture their children, especially because the child and not the parent paid for the education. Thus, the effect of requiring employers to provide servant children with minimal education may have been to increase rather than decrease the volume of families willing to immigrate as servants. Finally, governments in the United States did not directly prohibit immigrant servitude until 1885. Servitude among U.S. residents remained legal long after immigrant servitude had disappeared, with the most common form being apprenticeship and the indenture of indigent children.21 As a consequence, scholars must seek other explanations for the end of immigrant servitude.

Collapse of German immigrant servant market   349

Why did immigrant servitude disappear? Supply versus demand We can separate the plethora of hypotheses about why immigrant servitude disappeared into supply-­driven versus demand-­driven explanations. The most numerous and popular are demand driven. They argue that American demand for immigrant servants fell (shifted left), driving down the price of servant labor until Europeans were no longer willing to finance emigration through servitude. According to this theory, falling demand drove the quantity of servants supplied to zero. Numerous hypotheses exist for why American demand might have collapsed: recalcitrant servants could not be put in debtors’ prison, interference from immigrant aid societies and state regulations became too burdensome, the recession of 1819 reduced the ability to hire labor, free-­wage labor was superior, and servants escaped too easily.22 The supply-­driven explanations, alternatively, argue that the supply of European immigrant servants fell (shifted left), forcing up the price of servant labor, until American employers were no longer willing to buy servant contracts. According to this theory, falling supply drove the quantity of servants demanded to zero, and servant supplies collapsed either because passage fares fell or because emigrant wealth rose until all emigrants could finance their migration with their own cash rather than through servitude.23 The behavior of servant prices is inconsistent with demand-­driven explanations of the end of immigrant servitude. If falling demand dominated any movement in supply and thus caused servant quantities to go to zero, it should have driven servant prices down as well. However, if servant prices rose as servant quantities went to zero, then falling supply must have dominated any movement in demand. The servant’s contract price equalled the debt owed to the shipper for migration (passage fare, additional preembarkation expenses, and a 15 percent markup to compensate the shipper for interest and risk costs) plus any contracted cash payments given to the servant during or upon completion of the contract.24 The per-­year price of servant labor equalled the contract’s price divided by the contract’s length. Figure 18.2 shows that the real per-­year price of servant labor rose over time, from under 15 dollars in 1772 to well over double that amount by 1820. The steepest rise occurred during the last five years of the market, from 25 to 73 dollars for single adult males and from 19 to 31 dollars for married ser­ vants, between 1817 and 1821. The only way prices could have risen as quantities fell to zero is for leftward shifts in supply to have dominated any movement in demand.25 Therefore, this evidence eliminates all demand-­dominated explanations. There is no escaping the conclusion that falling supply was the dominant cause of the immediate end of European immigrant servitude in America.26 Both rising passage debts and falling contract lengths brought about the rise in the per-­year price of servant labor. Figure 18.2 shows that passage debts rose from 1772 through 1821, with one of the steepest rises occurring between 1817 and 1821. Thus, employers found that paying an immigrant’s passage cost, and thus acquiring the immigrant as a servant, was becoming increasingly expensive.

350   German immigrant servitude in America, 1745–1835 110

Deflated U.S. dollars (1795 = 100)

100 90

Average contract price for a married servant

Average contract price for single adult male servants

Passage fares for free adults

80 70 60 50 40 30

Average per year price for single adult male servants

Immigrant servitude disappears

20 Average per year price for a married servant

10 0

1775

1785

1795

1805

1815

1825

1835

Year of arrival

Figure 18.2 Adult passage fares, contract prices, and per-year contract prices, 1772–1835 (source: Table 18.1). Notes All series except for passage fares are three-year moving averages weighted by the number of observations per year. Breaks in the data are connected by straight lines. Passage-fare data points are indicated with markers and include the cost of meals. Contract prices include any contracted cash payments made to the servant during or upon completion of the contract. Contract prices typically exceeded passage fares of comparable free adult immigrants because they included a risk premium and sometimes additional migration expenses other than the passage fare, see note 24 and note 36. The average per-year price of service divides average contract price by average contract length for each year for each group, respectively.

If nothing else in the market had changed, employers would have required immigrant servants to work more years in order to make up for the higher passage cost. As the monetary cost of passage rose, however, the amount of labor required to pay this cost, the servant’s contract length, fell. Whereas adult contract lengths had remained roughly constant at four years from the mid-­ seventeenth century through 1772, Table 18.1 shows that by 1818 they had fallen to three years and by 1821 to two years. Employers bid down contract lengths as they competed for increasingly scarce servant supplies. In the last years of immigrant servitude, as servants fell from over 40 to under 5 percent of immigrants, both contract prices and contract lengths moved in opposite directions by large magnitudes. The combined effect, dividing contract price by contract length, caused one of the steepest peacetime rises in the per-­year price of servant labor the institution had ever known. This movement indicates that in the

Collapse of German immigrant servant market   351 final years of the market, shrinking servant supplies were reducing the quantity of servants demanded.27 Not only did supply shifts dominate demand shifts during the last years of the market, but supply shifts dominated demand shifts throughout the final 50 years of the market. Demand, far from decreasing, actually increased. Figure 18.3 divides the period 1772 through 1821 into nine subperiods (subperiod dates are in parentheses). The subperiods identify large shifts in servant quantities, as reported in Table 18.1. The figure treats each subperiod as a separate market equilibrium (identified as points 1 through 9), where each subperiod’s average quantity of all immigrant servant arrivals is plotted against its average per-­year adult-­servant contract price. Theoretically, the intersection of a supply curve with a demand curve generated each of the observed equilibria. Therefore, shifts either in the supply or the demand curve, or shifts in both curves, caused movement from one equilibrium to the next. We cannot, however, observe supply and 60 Supply 9

Average adult price per year of service

55 50 45

(1820–1) 9

30

3

20

10

2 (1785–7)

(1788–90)

0

100

200

(1816–17) 7

6 (1802–4)

Demand 4–7

1 (1772)

Supply 1&7

5 0

Supply 2&6

(1818–9) (1797–1801) 8 (1791–6) 5 4

25

15

Supply 3, 4 & 8

Demand 8–9

40 35

Supply 5

300

400

500

Demand 1–3 600

700

800

900 1,000

Quantity of immigrant servants arriving per year

Figure 18.3 The supply and demand for German servants in the Philadelphia market, 1772–1821 (source: Table 18.1). Notes See the text for discussion of the figure’s construction. Market equilibriums are indicated with markers and numbered from 1 through 9 in sequence from 1772 through 1821. The model uses the minimum number of linear supply and demand curves required to generate the given equilibrium points. The average per-year adult contract price was constructed by dividing the average contract price by the average contract length for adult-male servants for each year and then taking the average of this number for the interval of years indicated at each equilibrium point, weighted by the total number of German servants arriving each year. Contract prices are in deflated U.S. dollars (1795 = 100). The 1816–17 quantity excludes the 300 servants who arrived in Wilmington, Delaware, who are included in Table 18.1.

352   German immigrant servitude in America, 1745–1835 demand curves. They are theoretical ideas imposed on the observed data points. The goal of scientific theory is to impose the least amount of modeling onto the data necessary to explain the observed movements in the market. Figure 18.3 does this by drawing the minimum number of linear supply and demand curves needed to generate the equilibria drawn (that is, the fewest number of supply and demand curves that still produce a supply and demand intersection at each equilibrium data point).28 This model shows that the market was subjected to continuous and large fluctuations in supply between 1772 and 1821. The largest shift coincided with the end of the market when, from 1817 to 1821, supply fell by an unprecedented magnitude. By contrast, demand changed little. Instead of decreasing, it actually increased a small amount around 1791, and again after 1818.29 The American demand for immigrant servants was also highly elastic. The model in Figure 18.3 for the demand curve during 1772 through 1790 implies an elasticity of –35; for the demand curve during 1791 through 1817, it is –255; and for the demand curve during 1818 through 1821, it is –18. Apparently, American employers were very price sensitive to the use of servant labor and so must have considered immigrant servants and free-­wage laborers (both locally born and those migrating in from other regions) to be close substitutes. Employers did not abandon servitude because they found wage labor to be superior. The opposite was true. As immigrants abandoned servitude, employers found it easy to shift to wage labor because they regarded the two types of labor to be interchangeable.30 Employers did not react with any loud lament or frantic adjustment to their labor regime when servant supplies dried up, because alternative sources of unbound labor were readily available. Philadelphia experienced significant rates of net in-­migration after 1815. By the nineteenth century, immigrant servitude was a marginal appendage to the region’s labor market.31

Why did immigrant servant supplies collapse after 1817? In order to explain the drop in the supply of servants, one must explore the methods Europeans used to finance their migration. Of course, immigrants with enough savings paid for their passage in cash. Thus, it is possible that servitude declined because passage fares fell or because immigrant wealth (or access to wealth) rose. Significant declines in passage fares in the German passenger trade, however, did not occur until after 1829, a full decade after servitude disappeared. Although falling passage fares may explain why immigrant servitude did not revive later in the century, they cannot be the immediate cause of immigrant servitude’s disappearance. From 1815 through 1825 passage fares and servant passage debts experienced the fastest peacetime rise and reached the highest levels (in real value) ever (see Figure 18.2 and Table 18.1).32 High passage costs should have increased the percentage of immigrants entering servitude, other things equal, because fewer could pay all their migration expenses in cash. Thus, immigrant servitude disappeared exactly when increases in passage costs should have caused the servant percentage to rise.

Collapse of German immigrant servant market   353 Although changes in the price of transportation cannot explain the decline in the supply of servants, the possibility remains that increases in immigrant wealth may have caused the falloff by increasing the number who could pay for passage in cash from their accumulated savings. Indeed, the elimination of state and local emigration taxes and manumission fees by members of the German Confederation after 1791 suggests a mechanism by which average levels of savings could have been increased.33 The evidence on this possibility is, however, thin and somewhat mixed.34 Only one ship manifest has survived for the period from 1816 to 1821 that recorded passenger debts, that for the ship Elizabeth, which arrived at Philadelphia in 1819 from Amsterdam.35 On this ship, 16 percent of the immigrants in debt for their passage nevertheless contributed some amount to defray the cost, with the average contribution being 41 percent of the debt. We can estimate a similar measure indirectly from the evidence in the servant registers. Many immigrants who had partially paid their passage debt, instead of negotiating a shorter servant contract commensurate with their smaller debt, agreed to a full-­term servant contract with the stipulation that the master would refund the immigrant’s contribution to the debt, sometimes with interest, upon completion of the servant contract. The immigrant’s contribution can be calculated by comparing this end-­of-contract cash payment with the initial purchase price paid by the master.36 This contracting behavior was rare prior to the War of 1812. However, during the last three years of the servant market, 1817, 1818, and 1819, 26 percent, 13 percent, and 20 percent of the adult contracts, respectively, possessed these end-­of-contract payments. The average portion of the passage fare these immigrants paid was 46 percent, 35 percent, and 32 percent, respectively. Although this evidence indicates that immigrants had more wealth after 1815, it cannot fully explain the end of servitude. First, many immigrants in these years still could not afford to pay their passage fare in cash. Second, the percentage of servant contracts with these partial passage-­debt payments, as well as the percentage of the debt paid in these contracts, was falling. Thus the ability of immigrants to at least partially defray their passage debts in the late 1810s appears to have been diminishing, rather than increasing, in the years immediately preceding the disappearance of the market. Comparison of the family, age, and occupational structure of German immigrants arriving from 1784 to 1814 with those arriving from 1815 to 1820 also indicates that the average immigrant may have been less wealthy in the latter period. The average number of children accompanying household heads rose from 2.9 to 3.8.37 Additional dependents increased migration expenses for the typical household head, thus reducing the ability of a given wealth level to finance the family’s migration in cash. In addition, the proportion of adult males in the 16–25 age range increased relative to those over age 25, by 3 percentage points. Given that age is the primary determinant of accumulated wealth, immigrants may have been less wealthy when servitude disappeared in 1820 than 30 years earlier when servitude was thriving. The major change in the occupational distribution of adult male immigrants was an increase in tailors, butchers, bakers, and merchants, a combined increase of 11 percentage points, and a decrease in

354   German immigrant servitude in America, 1745–1835 farmers and those listed with no occupation, by 6 and 5 percentage points, respectively. Given the importance of land as an asset and store of wealth, it seems unlikely that tailors, butchers, and bakers had more property than farmers that could be sold prior to emigration to raise the amount needed to pay for transatlantic passage in cash. Finally, the agricultural disaster in the southern Rhineland during 1816 (the “year without a summer”) may have reduced the wealth of the immigrants arriving from 1816 to 1820 compared with those arriving from 1784 to 1814.38 An alternative way to finance passage was to borrow from family and friends. The prohibitive cost of enforcing repayment once the immigrant had left Europe largely ruled out borrowing from European family and friends. Family and friends living in America, however, could more easily monitor and enforce repayment of any loan because of their close proximity and by using such devices as controlling the immigrant’s employment opportunities and access to other forms of assistance within the local kinship network. Moreover, because of the high monitoring and contracting costs associated with servant contracts, access to this source of financing may have been more important in ending immigrant servitude than changes in the immigrants’ own wealth. Exchange within a kinship network is similar to exchange within a firm. Firms are organized to economize on information, monitoring, and contracting costs in such a way that exchange can occur without the necessity of explicit, legally binding contracts specifying each transaction between members.39 When monitoring and contracting costs are high, exchange within firms or kinship networks will be more cost efficient and so preferable to marketplace transactions. The monitoring and contracting costs of servant contracts were high because masters paid passage debts and fixed contract lengths before putting servants to work and because there was little repeat contracting. Thus servants had an incentive to shirk and masters had an incentive to use excessive force to coerce their servants into working harder. The line between normal and excessive coercion, however, was ill defined both legally and contractually. An explicit, legally binding, and recorded servant contract established a legal right to the servant’s labor and defined the key conditions of the exchange. However, it specified nothing about the daily workings of the servant–master relationship, except for the clause stipulating “customary” behavior as recognized by common law, a constraint with considerable plasticity.40 In practice, the master had little legal recourse for losses due to servant shirking, and the servant had little legal recourse for the disutility of excessively harsh masters. It was too costly to negotiate an endless list of stipulations into the written contract regulating work effort, task variety, coercive boundaries, resale conditions, and so on. The servant contract was only viable and cost efficient if the master was given wide discretion over these margins of the institution. The mismatch of expectations between masters and servants over how the contract in practice restricted and directed their behavior did not always involve extreme forms of abuse by either master or servant, but it could nevertheless be costly to both parties. For example, Reverend Muhlenberg, head of the German

Collapse of German immigrant servant market   355 Lutheran Church in Pennsylvania, gave this description of his experience with a German servant he purchased in 1751: “At first he was very happy because he received good food, drink, and clothing in my house and had no hard work to do. He had hardly been with me for nine months, however, being hindered in his movements outside the house and being required to attend morning and evening prayers when I was at home, when he could no longer endure his good fortune and became surly and unruly.”41 Muhlenberg’s only recourse other than coercion was to sell the servant, which he did, at a price 23 percent below what he had initially paid for him. If the exchange were between kin rather than between strangers, employers could largely avoid these monitoring and contracting costs, as well as the costs of drawing up an explicit contract in the first place. Publicly recorded and legally binding contracts are less necessary among kin because shirking or cheating among family members is less likely than between strangers. Related individuals have lower information costs in monitoring each other’s behavior, will not risk the imposition of costly sanctions, both pecuniary and nonpecuniary, by their kinsmen for misbehavior, and may care about the welfare of their kinsmen (that is, in economic terminology, have positive interdependent utility functions). In general, servants would expect better treatment from kin than from strangers.42 Borrowing from kin in America would shift the exchange of labor for passage to the private unrecorded sphere (internalized within the family or immigrant community) and thus eliminate the public market in alienable servant contracts. In other words, the supply of immigrants negotiating alienable servant contracts in Figure 18.3 would shift left until the market disappeared from public view and government records. Although the evidence on alternative methods of repaying passage loans is scarce (because of its unrecorded nature), immigrants appear to have gained increasing access to the resources of family and friends in America. Observing German immigrants in debt for their passage in Philadelphia in 1769, Reverend Muhlenberg commented, “Whoever has well-­to-do friends seeks a loan from them to pay the passage, but there are only a few who succeed.”43 By the nineteenth century, German immigrants had become more successful at tapping these resources because the German community in Pennsylvania as elsewhere had grown, become financially successful, and found a renewed ethnic awareness, as the activities of German immigrant aid societies illustrated. For example, Christopher Mayer, president of the German Society of Maryland, told a group of German redemptioners on the ship Vrouw Elizabeth docked in Baltimore in 1819, “The German Society makes it its duty to assist your countrymen when they are in need, and to protect them as far as it is able.”44 Although the German-­American community directed an increasing amount of resources and interest toward their immigrant kinsmen after 1815, the magnitude of the charity and credit they extended is difficult to gage.45 The limited evidence, however, indicates that German immigrants were increasingly able to have their way bought out of servitude. Three ship manifests reported the names of those who were in debt for their passage and those who had paid cash in

356   German immigrant servitude in America, 1745–1835 advance. We can estimate the percentage of debtors who found alternative repayment methods, and thus avoided servitude, by matching these names with those in the servant register. For the two ships arriving at the turn of the century, 37 percent of the passengers arriving in debt for their passage did not enter servitude to repay the debt. They must have found some alternative way of repayment, such as having American relatives pay off their debts. By 1819, this number had increased to 57 percent.46 The gap between migrating in debt and entering servitude to repay that debt was growing wider. Although servitude appears to have been a minor method of repaying passage debts by the time immigrant servitude disappeared, the collapse of immigrant servitude did not represent the collapse of immigrant indebtedness.

Why was the collapse so abrupt? The abrupt collapse of immigrant servitude between 1818 and 1821 indicates that increases in immigrant wealth or the wealth of the immigrants’ family and friends in America could not have been the immediate cause of the end of servitude. Wealth could not have risen fast enough in this short interval to have eliminated servitude. It should have produced a gradual long-­run decline in the servant percentage rather than the sudden virtual cessation depicted in Figure 18.1. The resolution to this puzzle lies in a sequence of unique historical events that changed a gradual decline into an abrupt collapse. An acute agricultural crisis in the southern Rhineland in 1816 impoverished prospective emigrants and thus temporarily inflated the servant percentage in 1817 and 1818 well above its long-­run trend. As normal conditions returned after 1818, emigrant reliance on servitude fell back to its long-­run trend, thus giving the appearance of a sudden collapse. The quarter-­century of war and turmoil surrounding the birth of the French Republic ruined many lives and livelihoods in the Rhineland, the traditional source of German immigration to America. Emigration from this region during war was too risky. The disruption of commerce along the Rhine and of transatlantic trade from Dutch ports had made the traditional migration routes hazardous and uncertain. The return of peace and normal commerce in 1816 released this pent-­up migration demand.47 The surge of immigrants, whose lives these wars had impoverished, temporarily inflated the servant percentage above normal. Nature compounded this effect with its own disaster in 1816. People on both sides of the Atlantic long remembered “the year without a summer.” The summer was unusually cold and wet, and as a consequence, harvests were scant. In London, John Quincy Adams remarked that there had been “not one evening and scarcely a day in 1816 when a fire would have been superfluous.” Snow fell in London in August. The wine harvest was the latest ever known, the summer the coolest in all the long series of European temperature observations, and the wetness comparable only to the years 1316 and 1675. Whereas in most distressed regions winter famine was avoided by the importation of food from

Collapse of German immigrant servant market   357 less-­distressed regions, such as from America, the southern Rhineland was less fortunate. The summer rains had caused devastating floods and exceptionally poor harvests. The last years of the war had depleted grain reserves, and the inclement weather and an early freeze delayed grain shipments from the Baltic ports from moving down the Rhine until late in the following spring.48 This catastrophe set off a migration of desperate and destitute souls east to Russia and west down the Rhine through Amsterdam to America. The year 1817 produced the largest single-­year spike in German migration to America up to that time.49 The desperation of that year may have led many to migrate even though they did not have established kinship networks in America from which they could secure information and financial assistance. Many emigrants had exhausted their resources escaping starvation by the time they reached Amsterdam, and thus an atypically high percentage had to rely on servitude to pay for the final stage of their escape to America. We can deduce the impoverished state of these immigrants by the strain their care imposed on the private charities and public authorities of Amsterdam, by the increase in disease among the immigrants—both in Amsterdam and on board ship—and by descriptions of their arrival in America.50 For example, Henry Fearon in the fall of 1817 described the German immigrant servants presented to him for sale on the docks of Philadelphia as follows: “Their clothes, if rags deserve that denomination, actually perfumed the air. Some were without shirts, others had articles of dress, but of a quality as coarse as the worst packing cloth. . . . [T]heir countenances fell to that standard of stupid gloom which seemed to place them a link below rational beings.”51 Although these chance historical events temporarily inflated the overall German immigrant-­servant percentage from 1816 through 1818, the shifting composition of servants in this period shows the true underlying long-­run decline in immigrant servitude. From 1785 through 1804 the composition differed little from the traditional colonial pattern. Single adult males were the largest group at 45 percent, with dependent children and married persons coming in a distant second and third, 25 and 19 percent respectively. After servant immigration resumed in 1816, an unprecedented reversal of relative shares occurred, with dependent children and single adult males changing positions. Between 1817 and 1820 dependent children were the largest group, comprising 42 percent of all immigrant servants, whereas the proportion of single adult males shrank to 24 percent.52 The servant percentage among single adult males, the traditional source of immigrant servants, was falling throughout this period, most likely because their increasing wealth, or access to the wealth of family and friends in America, allowed them to purchase passage without recourse to servitude. The rising proportion of dependent children among the immigrants drove the overall servant percentage in the late 1810s above its pre-­War of 1812 level. The “year without a summer” affected families with many mouth to feed more severely than single adults, and thus led them to rely on servitude more heavily to pay for some of their migration expenses, particularly those of their older children.53 These chance historical events inflated and perpetuated the percentage of

358   German immigrant servitude in America, 1745–1835 German immigrants entering servitude. In the absence of these events, the percentage of those entering servitude would have declined more gradually, and the market would have disappeared sooner. Better-­than-normal agricultural conditions returned to the Rhineland after 1819. German migration to America fell off markedly in the 1820s. America fell into disfavor and rival destinations, such as Poland, Russia, and Brazil, siphoned off emigrants, in many cases through subsidized colonization schemes.54 By the time emigration to the United States picked up again in the 1830s, passage fares were low and immigrants’ access to wealth great enough that recourse to alienable servant contracts was no longer necessary.

Conclusion: from redemption to remittance Immigrant servitude disappeared not because employers or shippers abandoned the institution, but because the supply of immigrant servants fell to zero.55 It ended abruptly around 1820 because of chance historical events that temporarily inflated the servant percentage in the years immediately preceding 1820. The supply fell because immigrants were increasingly able to purchase passage with accumulated savings and by borrowing from family and friends upon landing in America. Under the latter method, immigrants still had to sign explicit redemption-­debt contracts with their shippers to secure passage and were thus exposed to being sold as servants or jailed as defaulted debtors if their relatives or friends failed to appear and pay their debts when they debarked. If relatives in America were going to pay for a person’s passage anyway, it would be cheaper to pay before embarkation in Europe (the remittance system) than after debarkation in America (the redemption system). Although the interest foregone was similar under both systems, redemption was riskier for the shipper who consequently charged a 15 percent markup on the passage fare as compensation for the costs of morbidity, mortality, escape, and the marketing of redemptioners in America.56 Redemption was also riskier for immigrants because shippers could sell them to strangers if their relatives failed to appear in time to redeem their passage debts. In addition, it was riskier for the (nonkinsman) employer who consequently required a longer contract to compensate for the servant’s incentive to shirk and escape. Thus, both the monetary and labor cost of passage would be cheaper if it was borrowed within the kinship community under the remittance system. After 1830, the decline in passage fares and rise in immigrant wealth lessened the need for borrowing passage expenses. For example, by the 1840s the average German immigrant carried over 90 dollars in cash, more than enough to cover the cost of passage in 1820, and three to four times the cost of passage in 1845.57 However, some immigrants, especially young adults, were always too poor to pay for their passage; but by the 1830s, they generally used the remittance system. Remittances appeared on the Londonderry-­to-Philadelphia route in the spring of 1816, and were extensive by 1826. By 1834 remittances paid the passage for one-­third of the Irish sailing from Belfast, more than the proportion

Collapse of German immigrant servant market   359 entering servitude in the 1770s. The development of the remittance system in the German passenger trade is less well documented, but researchers think it was similar to that of the Irish trade.58 The remittance system was a superior, cost-­ minimizing institution. The last remnants of European immigrant servitude disappeared from public records as the remittance system absorbed passage debts into an informal market of personal kinship obligations within the immigrant community. The rise of the remittance system required the development of complementary markets. Paying for a relative’s transatlantic passage would be cheaper under the remittance system compared with the redemption system only if there were an inexpensive and reliable way to transfer funds or boat tickets from America to Europe. Before the development of transatlantic shipping lines and regularly scheduled passenger service, Americans may have found the cost of contracting for a westbound passenger berth for a European relative prohibitive. They would have found it difficult to communicate to their European relatives the precise information regarding the particular ship, captain, port, and timing of embarkation for the berth purchased, and they would have found it difficult to ensure that the shipper, once he was paid for the ticket before departing from America, would actually follow through with providing the passenger berth on the return voyage to America. Without repeat passenger business, shippers faced little reputation cost from reneging on the sale of individual passenger berths.59 Similar problems arose when Americans sought to transfer funds to Europe so their European relatives could purchase a westbound passenger berth at their convenience. In the absence of regular transatlantic banking and merchant connections, and the reputation for honesty that repeat business creates, Americans may have found it difficult to locate a trustworthy agent who would take their money to relatives in Europe without embezzling the funds.60 The development of regular passenger shipping routes and merchant connections after the Napoleonic Wars and the development of regular banking relations between transatlantic ports after 1815 lowered the transaction and enforcement costs of the remittance system and allowed it to gain ascendancy over the redemption system as the primary method used by Americans to finance the passage of their European relatives. During this same period, the expansion of regular industrial employment in America allowed American workers to overcome their own capital or savings constraint and borrow against future wages to purchase boat tickets for European relatives.61 In many cases, the expansion of regular industrial employment in America also allowed American workers to guarantee employment to their kin upon arrival, thus allowing easy recovery of the passage loan. Because purchasing a relative’s passage through the remittance system was cheaper than through the redemption system, these developments in industrial employment also tended to promote the ascendancy of the remittance system. All these post-­1815 market developments lowered the cost of tapping transatlantic kinship networks, something that was harder to do during the preceding century when higher costs made immigrant servant markets the primary vehicle for financing transatlantic migration on credit. Chain migration through

360   German immigrant servitude in America, 1745–1835 kinship networks has been shown to be a key determinant of the extent and direction of free European migration and to have been largely absent from servant migration.62 The forces that lowered the cost of chain migration were, therefore, the same forces that eroded the supply of immigrant servants. The replacement of the redemption with the remittance system suggests that the evolution of exchange is not a linear or hierarchical progression from kinship networks to ever wider markets. The kinship–market dichotomy is like the firm– market dichotomy. Just as a firm decides what components are better to produce in-­house and what are better to purchase in the market, a person decides what goods are better to exchange with kin and what are better to purchase in the market. The line dividing the locus of exchange between these two spheres depends on transaction costs whose evolution is not necessarily hierarchical or linear. For example, exchange within firms, as witnessed by the growth of firm size, has grown simultaneously with the expansion of markets. The development of banking, shipping, and industrial markets after the Napoleonic Wars changed the relative cost among different methods of financing transatlantic European migration on credit and so shifted the trade from the public-­market sphere of explicit alienable contracts back into the sphere of informal kinship transactions.

Notes   * Originally published as Farley Grubb, “The end of the European immigrant servitude in the United States: an economic analysis of market collapse, 1772–1835,” Journal of Economic History, vol. 54, no. 4 (December 1994): 794–824. Copyright 1994. © 1994 by Cambridge University Press and the editors of The Journal of Economic History. Reprinted with permission. The Journal of Economic History, Vol. 54, No. 4 (Dec. 1994). © The Economic History Association. All rights reserved. ISSN 0022–0507. The author is Professor of Economics at the University of Delaware, Newark, DE 19716. Earlier versions were presented at the University of Chicago, the University of Delaware, and the annual meetings of the Social Science History Association. The author thanks the participants of these seminars and Rick Agnello, Agnes Bretting, Raymond Cohn, John Craig, David Eltis, Stanley Engerman, Joseph P. Ferrie, Robert W. Fogel, David W. Galenson, Henry Gemery, Dirk Hoerder, J. R. T. Hughes, Ken Koford, Daniel Levy, David Mitch, A. G. Roeber, Theodore W. Schultz, Ralph Shlomowitz, Richard Steckel, and Robert Whaples for their help. Financial support from the Economic History Association Cole Grant-­in-Aid and from the University of Dela­ware General University Research Program is gratefully acknowledged. Anne Pfaelzer de Ortiz provided editorial assistance.   1 See Grubb, “Incidence,” 316–39. For recent discussions of the general magnitude and importance of colonial immigrant servitude, see Bailyn, Voyagers: 85–352; Dunn, “Servants,” 157–94; Engerman, “Servants,” 263–94; Galenson, White ­Servitude; Gemery, “Markets,” 33–54; and Grubb, “Long-­Run Trend,” 167–240. For the classic studies of immigrant servitude, see Geiser, Redemptioners; Herrick, White Servitude; McCormac, White Servitude; Morris, Government; and Smith, Colonists.   2 See Engerman, “Servants,” 263–94; Erickson, “Why Did Contract Labour?” 34–56; and Galenson, “Rise and Fall,” 13–24.   3 Relative labor scarcity influenced regional patterns of American industrial development;

Collapse of German immigrant servant market   361 see David, “Industrial Labor Market Adjustments,” 47–97; and Goldin and Sokoloff, “Relative Productivity Hypothesis,” 461–87. On the effect of servitude on immigrant integration into the American work force, see Grubb, “Immigrant Servant Labor,” 249–75, “Fatherless and Friendless,” 85–108, and “Disappearance.”   4 For recent discussions of these explanations, see Barzel, Economic Analysis: 83–4; Erickson, “Why Did Contract Labour?” 34–56; Galenson, “Rise and Fall,” 12–24; Grubb, “Disappearance”; and Salinger, To Serve Well: 142–52.   5 The transition from servants to slaves south of Pennsylvania has been well documented elsewhere; see Dunn, Sugar: 46–116; Galenson, White Servitude: 117–68; Grubb and Stitt, “Liverpool Emigrant Servant Trade”; Menard, “From Servants to Slaves,” 355–90; and Morgan, American Slavery: 295–315. Dividing the colonial population by a generous estimate of the average number of immigrant servants under contract at any one time indicates that servants represented under 10 percent of the population by 1700. The number of servants was taken as 12,000 (an average of 3,000 arriving yearly times the typical four-­year contract, ignoring mortality); the estimate used the population of New Jersey, Pennsylvania, Delaware, Maryland, and Virginia (the principal non-­Caribbean destinations of servants). See Gemery, “Markets,” 38; Grubb, “Long-­Run Trend,” 167–240; and U.S. Bureau of the Census, Historical Statistics: 1168. Immigrant servants were under 4 percent of the population of Barbados by 1684 (Dunn, Sugar: 87).   6 See the discussion in Grabbe, “Das Ende,” 286.   7 See “Account of Vessels”; Grubb, “Reliability,” 29–54; “Philadelphia Customs House Records”; “Pilot’s Report”; and Strassburger, Pennsylvania German Pioneers.   8 “Record of Indentures.” For the port of Philadelphia, immigrant servant registers have not been found for the years 1773 through 1783 and may not have existed. See Grubb, “Servant Auction Records,” 156–8.   9 See “Book A of Redemptioners”; “Book C of Redemptioners”; and Table 18.1. 10 On the redemptioner system, see Grubb, “Auction,” 583–603, and “Redemptioner Immigration,” 407–18; and Smith, Colonists: 20–5. 11 See Fearon, Sketches: 148–51; Grubb, “Auction,” 583–603, and “Disappearance.” 12 Between 1815 and 1820 German immigrant ships also arrived at Baltimore, New York, and New Orleans. Although the exact volume of immigrants and servants landing at these other ports is unknown, it appears smaller and its variance appears similar to that at Philadelphia. See Grubb, “Reliability,” 29–54; Hansen, Atlantic Migration: 85–123; Henninghausen, History: 17–96; and Walker, Germany: 1–41. In the 1820s Philadelphia was second only to New York in immigrant arrivals and received roughly 12 percent of total arrivals to the United States (Bromwell, History: 21–61). Although pre-­1820 immigration data do not exist for the United States, trade data and anecdotal evidence indicate that Philadelphia was the most important port of entry in the first three decades of the Republic. See Lindstrom, Economic Development: 23–40. 13 Direct quantitative evidence on British servant emigration to the United States after 1784, which was mostly from Ireland, could not be found and may not exist. Indirect evidence indicates that it was in decline by the 1790s and was, at best, a trickle by 1818. It clearly ended before the demise of German immigrant servitude. See Adams, Ireland: 113–14; Hansen, Atlantic Migration: 53–106; Herrick, White Servitude: 263, 307–8; Jameson, “Letters”; Jones, “Ulster Emigration,” 51–62; and Miller, Emigrants: 169–70. 14 For example, see Grabbe, “Das Ende,” 286–7; and Klepp, “Demography,” 111. 15 See Strassburger, Pennsylvania German Pioneers; Wokeck, “Flow and Composition,” 260–1; and Table 18.1. 16 Emigration from the southern Rhineland in 1817, the traditional source of German emigration to America, took on large proportions and went in several directions. The

362   German immigrant servitude in America, 1745–1835 two most important were east towards Russia and down the Rhine to Holland where passage to America was sought. The “year without a summer” refers to 1816. See the discussions in Hansen, Atlantic Migration: 85–123; Skeen, “Year,” 51–67; and Walker, Germany: 1–41. 17 These numbers are from the raw count of contracts registered in “Book A of Redemptioners”; and “Book C of Redemptioners.” The numbers in Table 18.1 incorporate estimates of the volume of servants for the period 1805 to October 1817, and additional numbers were added from servants landing in Wilmington, Delaware in 1818. The data in Table 18.1 indicate that a yearly average of 853 German servants arrived from 1816 to 1819, far more than for any other four-­year period since 1754. See Grubb, “Servant Auction Records,” 167. 18 Exactly why the register was discontinued in 1831 is unknown. A deficient number of servants may have been reason enough. Clearly, it was not due to a lack of German immigration. Table 18.1 shows that the volume of German immigration to Philadelphia from 1831 through 1835 was as large as in any other period after the Revolution, with the exception of 1817. 19 By the end of the eighteenth century, regulations included oversight by a bilingual official, registration of servant sales (unregistered sales were invalid), a 30-day period after arrival for redemptioners to negotiate repayment of their passage debts (during which they had to be maintained at the shipper’s expense), a prohibition on the separation of married servants and on selling servants across state lines without their consent, specification of customary freedom dues to be paid at the end of the contract and of procedures for handling sick redemptioners, and a prohibition on forcing surviving passengers to pay off the debts of those who died during the voyage and on the confiscation of passenger luggage as a method of paying off the debts of those who died during the voyage. These regulations, many dating back to 1764, lowered the transaction and information costs of contracting as well as established firmer property rights in servant labor. By reducing fraud and increasing contractual certainty for all parties—employer, shipper, and servant—these regulations enhanced, rather than impeded, the continuing viability of the market for transporting and selling immigrant servants. The regulations passed in Maryland and Pennsylvania after 1815, with minor exceptions, simply reaffirmed this prior body of law. The exceptions were an education mandate for children, discussed in the text, and a prohibition on adult contracts exceeding four years (Maryland only). Given that adult contract lengths were well below four years (see Table 18.1), this last regulation was nonbinding. See Geiser, Redemptioners: 59–70; Grubb, “Disappearance”; and Herrick, White Servitude: 263, 286–308. Because these regulations only affected the initial contracting and sale of the servant, they did not resolve the shirking and malfeasance problems inherent in the execution of the servant contract discussed later in the text. 20 See Grubb, “Educational Choice,” 367, 372; “Book A of Redemptioners”; and “Book C of Redemptioners.” Children were distinguished in the servant records by the presence in their contracts of stipulations requiring education, parental consent, or other comments indicating that they were underage dependents of other immigrants. 21 See Clark, “Babes,” 475–86; Cloud and Galenson, “Chinese Immigration,” 22–42; Erickson, “Why Did Contract Labour?” 38–56; Galenson, “Rise and Fall,” 1–26; Jacoby, “Transformation,” 887–910, and “Legal Foundations,” 229–50; Morris, Government: 514; and Steinfeld, Invention: 123–95. In Delaware the last officially recorded servant contracts for an indigent child in New Castle, Kent, and Sussex counties were in 1900, 1926, and 1930, respectively. See Hancock, “Indenture System,” 47–59. 22 For discussions of these demand-­driven explanations and evidence of their popularity, see Adams, Ireland: 362; Barzel, Economic Analysis: 83–4; Erickson, “Why Did Contract Labour?” 34–56; Geiser, Redemptioners: 42, 59–70; Hansen, Atlantic Migration: 103–5; Henninghausen, History: 17–28, 56–86; Herrick, White Servitude:

Collapse of German immigrant servant market   363 263–6, 286–308; Jones, “Ulster Emigration,” 57; McCormac, White Servitude: 110; Miller, Emigrants: 169; Morris, Government: viii–ix, 354–63, 391–531; Salinger, To Serve Well: 133, 142–52; and Smith, Colonists: 226–84. 23 For a recent discussion of this explanation, see Galenson, “Rise and Fall,” 12–24. 24 Servants also received maintenance (principally food and shelter) and customary freedom dues, a payment upon completion of the contract of “two complete suits of clothes, one thereof to be new.” The cash value of these payments was seldom recorded. The relative cost of maintenance and freedom dues changed little over this period. Therefore, changes in the total yearly cost of servant labor can be measured by changes in the contract price per year of servant labor. The cash payments during and upon completion of the contract served several purposes. Some immigrants who had partially paid their passage debt agreed to full-­term servant contracts with the stipulation that the master would refund their contribution to the passage debt, sometimes with interest, upon completion of the contract. For others, cash payments were used as an incentive to elicit proper behavior, as a discouragement to running away, and as a calibrating device to adjust the pay of individual servants to more closely approximate their expected productivity. See “Book A of Redemptioners”; “Book C of Redemptioners”; Grubb, “Redemptioner Immigration,” 407–18, “Auction,” 583–603, and “Long-­Run Trend,” 183–6; and Kauffman and Cribari-­Neto, “To Pay.” 25 Increases in the per-­year price of immigrant servants were caused by shifts in supply and demand and not by increases in servant human capital. Although servant occupations are not known, the occupations of their masters are known. Between 1787–1804 and 1817–20 the occupational distribution of employers who purchased servants shifted away from tradesmen and towards farmers (tradesmen declined from 67 percent to 56 percent of the buyers). Servants purchased by employers from distant (frontier and agricultural) states also rose (from 2 to 23 percent). See Grubb, “Disappearance.” Skilled servants were more valuable to tradesmen than to farmers. Thus, tradesmen would outbid farmers for skilled servants. Given that the percentage of tradesmen among servant employers was falling, the percentage of servants with trade skills must have been falling as well. In addition, descriptions of servants arriving between 1817 and 1820 indicate that they were relatively depleted in terms of physical prowess, possibly due to the acute famine in the Rhineland during 1816 and 1817. See Fearon, Sketches: 148–51. These trends indicate that servant human capital was falling, not rising. Thus, the rising trend in servant value may understate the magnitude of the fall (leftward shift) in servant supplies depicted in Figure 18.3. Finally, recent research has shown that differences in human capital among adults were largely unrelated to adult immigrant servant value during the colonial era (see Grubb, “Long-­ Run Trend,” 186–92). Thus, any potential changes in servant human capital of reasonable magnitudes would have had only a minor impact and therefore should be of little concern here. 26 Direct evidence against the most popular demand-­driven explanations is presented in Chapter 17, where it is shown that even as servitude ended, debtors prison was still used effectively against servants, immigrant aid societies and state regulations actually encouraged servitude, few servants ran away, and the recession of 1819 did not alter the American demand for servants or bankrupt the shippers of immigrant servants. 27 Between 1772 and 1820, 38 percent of the rise in the per-­year price of servant labor was caused by rising passage costs and 62 percent was caused by falling contract lengths. For comparison to servant values in earlier years, see Grubb, “Long-­Run Trend,” 167–240. 28 The data are not of sufficient quantity and quality to identify econometrically all the individual supply and demand curves, along with their movement and elasticity. Although the model in Figure 18.3 is more heuristic than econometric, it should be

364   German immigrant servitude in America, 1745–1835 noted that every supply and demand curve is identified with at least two equilibrium points, with the exception of Supply 5 and Supply 9. These last two curves were simply drawn parallel to the other three supply curves in the figure. 29 The shifts in demand for immigrant servants depicted in Figure 18.3 may have been caused by increases in the real wages of local workers. For example, the real-­wage index for the five-­year period from 1818 to 1822 was 17 percent higher than for the preceding five-­year period (Adams, “Standard of Living,” 904). The association of increases in servant demand with increases in the real wages of local workers supports the claim that local wage workers and immigrant servants were substitutes. This substitution effect, the rise in the real wages of local labor causing an increases in the demand for servants, only extended the life of the immigrant servant market briefly, as it was being swamped by the fall in servant supplies. 30 A servant’s total compensation was roughly equal to that earned by a local wage laborer, suggesting that the two were close substitutes (Grubb, “Auction,” 586–91). Employers may have considered free wage labor and servants to be close substitutes because wage contracts could be constructed to replicate many of the features of servitude. For example, wage contracts could be made lengthy, such as for a full year, and, because workers had no legal right to prorated unpaid wages if they departed early, workers could be discouraged from early departure by having part of their wages paid at the completion of the contract. See the examples and discussion in Tomlins, Law: 223–92. 31 See Klepp, “Demography,” 108. The volume of German immigrant servants was too small, averaging under 300 a year between 1785 and 1821, to influence the market price of labor in the Delaware Valley. It comprised well under 1 percent of the region’s work force. Most of the agricultural work was done by landless local workers, tenants, inmates, and cottagers whose numbers had expanded during the previous half century. See Clemens and Simler, “Rural Labor,” 106–43; Salinger, To Serve Well: 178–83; and Simler, “Tenancy,” 542–69, and “The Landless Worker,” 163–99. 32 Passage fares for German immigrants to America had been rising since 1710, with the real value being roughly two to three times higher in 1825 than during the first half of the eighteenth century. The per-­year rise in fares from 1710 to 1775 was more modest than from 1815 through 1825. See Grubb, “Long-­Run Trend,” 184. Why transatlantic passage fares from Dutch ports rose substantially from 1815 through 1825 is not well understood, in part because this evidence has not been widely known. Two possible reasons are that the explosion in transatlantic trade after the Napoleonic Wars bid up the opportunity cost of freight space and that the U.S. passenger regulations of 1819 increased the relative cost of shipping human freight. 33 On emigration taxes and manumission fees, see Selig, “Idea and Practice,” 19. 34 On the one hand, redemptioners appear to have been increasingly able to pay off some or all of their passage debts. Some redemptioners found they could sell their personal possessions in America to repay their passage debts, apparently something they could not or would not do before embarkation (perhaps because prices on many of these goods were higher in America). Evidence for such behavior, however, is confined to a few examples, such as the observation made by Ludwig Gall about one redemptioner debarking in Philadelphia in 1819 who was offered 70 dollars for his double-­barreled gun, which was enough to pay his passage debt (Trautman, “Pennsylvania,” 40–1). Unless their goods would fetch a higher price in America, immigrants would have preferred to sell possessions prior to embarkation and then pay for their passage fare in cash because the credit price for passage was 15 percent higher than the cash price. See Grubb, “Redemptioner Immigration,” 407–18. For advice given to German immigrants encouraging them to carry goods to sell upon debarkation in America, see Durnbaugh, “Two Early Letters,” 231–3; and Grubb, “Auction,” 584–5. 35 “Passenger List of the Ship Elizabeth.”

Collapse of German immigrant servant market   365 36 The following contract registered on October 20, 1819 illustrates this behavior: “Johan Geschroindt willing as above [to go to Alabama, bound himself servant] to E. Clark agent for N. Farrows for two years two months and 10 days, to be paid in hand eight dollars, and [to be paid] 45 dollars 60 cents to him his heirs or assigns at the end of the term, being a part of his passage money paid at Amsterdam, with legal interest thereon. Also to be paid fifty dollars if he conducts himself well. Consideration = $28.” (Words in brackets were added.) The 45 dollars plus the initial consideration price of 28 dollars adds to 73 dollars, which equals the total passage fare paid by the typical adult servant. Although most of the contract registrations are not this explicit, a substantial number with end-­of-contract cash payments have the feature that the sum of this cash payment and the initial consideration price equals the typical adult passage fare and also equals the initial consideration price paid for servants with no end-­of-contract payments. The estimates reported in the text use the adult-­servant contracts for which this feature approximately holds. Because the immigrant took this payment as cash at the end of the contract, rather than using it to reduce the length of the servitude, it seems unlikely that this sum was a transfer from family and friends in America to the immigrant upon debarkation. Instead, this sum most likely originated from the immigrant’s own savings, as the example cited above suggests. See “Book A of Redemptioners”; and “Book C of Redemptioners.” 37 See Grubb, “German Immigration,” 423. 38 Grubb, “German Immigration,” 427, 432. Although the ship manifests do not indicate whether immigrants listed as farmers had been landowners or merely tenants, agricultural laborers were listed separately from farmers and accounted for less than 1 percent in both samples. The effect of the “year without a summer” is discussed in more detail in the next section. 39 For the classic discussion of the dichotomy between firm and market exchange based on transactions costs, see Coase, The Firm: 1–55. 40 See Grubb, “Redemptioner Immigration,” 407–18, and “Does Bound Labor?” On master-­servant law, see Morris, Government: 390–512; Steinfeld, Invention: 15–54; and Tomlins, Law: 223–92, and “Law and Power,” 71–98. 41 Muhlenberg, Journal, vol. 1: 264–5. Another excellent example of the margins of abuse within the servant contract is related in the first hand experiences of Buettner, Narrative: 28–40. 42 For example, see Green, “Ulster Emigrants’ Letters,” 98. On interdependent utility functions, see Becker, Economic Approach: 205–94. 43 Strassburger, Pennsylvania German Pioneers, vol. 1: xxxvii. 44 Henninghausen, History: 85–6. See also Pfund, History; Risch, “Immigrant Aid Societies,” 15–33; and Wust, Pioneers: 1–19. By the late 1810s, the existence of these American humanitarian societies and the fact that they had made conditions for German immigrant servants better than what had prevailed over the previous decades was known in Germany. See Buettner, Narrative: 27. 45 For example, the percentage of contracts in the post-­1817 servant register with explicit “buy-­out” clauses that specified the conditions that would void the contract, as well as the percentage that recorded early releases from their contracts, were relatively small—2.8 percent and 5.6 percent, respectively. The early releases were mostly children purchased out of servitude by relatives. Although these percentages are small in absolute magnitude, they are substantially larger than comparable percentages calculated from the pre-­1804 servant registers. Thus, access to the financial resources of American relatives was increasing over time. In addition, the wording of the buy-­out clauses suggests that the use of these clauses was restricted to the following case: immigrants who had expected American relatives to pay their passage debts but who found that their relatives had failed to appear in time to prevent them from being indentured. Such redemptioners were contractually reserving the right to be purchased out of their indenture by their tardy relatives. For example, Anna Dorothea

366   German immigrant servitude in America, 1745–1835 Schenzin’s contract, registered on September 17, 1818, in “Book C of Redemptioners,” explicitly stated that she was “to be free any time within one month if her freight is paid by her brother within that time.” Finally, because many servant owners resided outside of Philadelphia, some buy-­outs and early releases of servants initially purchased in Philadelphia may not have been recorded in the Philadelphia servant registers, thus biasing these estimated percentages downward. Servants could always be bought out of their contracts by their American relatives making a superior offer to their masters. See “Book A of Redemptioners”; “Book C of Redemptioners”; “Record of Indentures”; and Grubb, “Servant Auction Records,” 157. 46 The brig Pennsylvania arrived in 1796, the ship Belvidere arrived in 1803, and the ship Elizabeth arrived in 1819. See “Pilot’s Report”; Strassburger, Pennsylvania German Pioneers, vol. 3: 112–14, 137–38; “Passenger List of the Ship ‘Elizabeth’ ”; “Book A of Redemptioners”; and “Book C of Redemptioners.” Redemptioners might not appear in the servant register because they escaped or defaulted on their debts. This possibility, however, seems unlikely. If the number of escapes or defaults had grown substantially by 1819, then contract lengths would have been driven up to compensate. Table 18.1 shows that the exact opposite occurred—contract lengths fell. In addition, under 1 percent of German servants are estimated to have escaped between 1817 and 1820 (Grubb, “Disappearance”). 47 See Hansen, Atlantic Migration: 79–85; and Walker, Germany: 1–41. 48 See Hansen, Atlantic Migration: 85–89; Walker, Germany: 1–41; and Skeen, “Year,” 58. 49 See Table 18.1; Strassburger, Pennsylvania German Pioneers; and Wokeck, “Flow and Composition,” 260–1. 50 See Hansen, Atlantic Migration: 85–89; and Walker, Germany: 1–41. 51 Fearon, Sketches: 148–51. That the redemption system was “for want of another institution the only way” for Germans to get to America in 1817 is also suggested in Von Furstenwarther’s report Der Deutsche, published in 1818. See Grabbe, “Das Ende,” 284–5. 52 Between the migration of 1785–1804 and that of 1815–20, the proportion of single adult males fell 9 percentage points, whereas the servant proportion fell 21 percentage points. By contrast, the dependent children proportion among these immigrants rose 11 percentage points, whereas the servant proportion rose 17 percentage points. These figures are derived from Grubb, “Incidence,” 320, and “German Immigration,” 421; “Book A of Redemptioners”; and “Book C of Redemptioners.” 53 One of the common misconceptions perpetuated in the literature on German redemptioner immigration is that parents sold their children into servitude to pay for the parents’ passage debts. For example, see Smith, Colonists: 22. There is no evidence in the servant register of children being sold to pay for the debts of their parents. The average price paid for a child was seldom greater than the child’s own passage fare. 54 See Grubb, “Reliability,” 29–54; Hansen, Atlantic Migration: 107–19; and Walker, Germany: 1–41. 55 Shippers were just transportation intermediaries. The claim that redemptioner shippers went bankrupt, abandoned the business, and so caused the trade to end cannot be sustained by the evidence. See Hansen, Atlantic Migration: 105; and Grubb, “Disappearance.” 56 See Grubb, “Redemptioner Immigration,” 407–18; and Figure 18.2. 57 See North, “United States Balance of Payments,” 612; and Table 18.1. 58 Adams, Ireland: 101–2, 145–9, 175, 180–2, 355–6; Green, “Ulster Emigrants’ Letters,” 98; Grubb, “Incidence,” 327; and North, “United States Balance of Payments,” 614–18. 59 It may have been hard for Americans to directly purchase passage for European relatives during the colonial period because few shippers engaged repeatedly in the passenger trade. For example, out of 317 shipments of German passengers arriving at

Collapse of German immigrant servant market   367 Philadelphia between 1727 and 1775, 75 percent of the captains made only one voyage, and 93 percent made fewer than four voyages. Grubb, “Market Structure,” 41. 60 During the colonial era, trustworthy financial intermediaries serving the American– Rhineland route (who were not personal acquaintances) were scarce. Disreputable freelancing agents were plentiful. Some immigrants found they had to return to Germany personally to retrieve relatives, which was an expensive undertaking. See Roeber, Palatines: 131, 149, 187–91, 218. 61 The importance and timing of such developments on the remittance system based on direct quantitative evidence are still to be determined. Some illustrative examples can be gleaned from the pay books of the DuPont gunpowder mills in northern Delaware (Eleutherian Mills Historical Library). 62 See Grubb, “Fatherless and Friendless,” 85–108.

References “Account of Vessels Arrived at the Port of Philadelphia Kept by Captain Nathaniel Falconer, Health Officer of Said Port, viz., September 30, 1789 to May 3, 1794” (Unpublished manuscript, State Archives of Pennsylvania, Harrisburg, PA, microfilm RG-­41, roll 1). Adams, Donald R., Jr., “The Standard of Living During American Industrialization: Evidence from the Brandywine Region, 1800–1860,” Journal of Economic History, 42 (December 1982): 903–17. Adams, William Forbes, Ireland and Irish Emigration to the New World from 1815 to the Famine (New Haven, 1932). American Watchman (Wilmington, DE), January 7, 1818. Anonymous, Nachrichten und Erfahrungen uber die Uereiuigten Staaten von Amerika (Frankfurt, 1817). Bailyn, Bernard, Voyagers to the West (New York, 1986). Barzel, Yoram, Economic Analysis of Property Rights (New York, 1989). Becker, Gary S., The Economic Approach to Human Behavior (Chicago, 1976). Bezanson, Anne, Robert D. Gray, and Miriam Hussey, Wholesale Prices in Philadelphia 1784–1861 (Philadelphia, 1936). Billigmeier, Robert H., and Fred A. Picard, trans. and eds., The Old Land and the New: The Journals of Two Swiss Families in America in the 1820s (Minneapolis, 1965). “Book A of Redemptioners, 1785–1804” (Unpublished manuscript, Historical Society of Pennsylvania, Philadelphia). “Book C of Redemptioners, 1817–1831” (Unpublished manuscript, Historical Society of Pennsylvania, Philadelphia). Bromwell, William J., History of Immigration to the United States (New York, 1969). Buettner, Johann Carl, Narrative of Johann Carl Buettner in the American Revolution (New York, 1971). The Citizen and Farmer’s Almanac for the Year of Our Lord 1802 [and for years 1803–1806] (Wilmington, DE, 1802). Clark, Dennis, “Babes in Bondage: Indentured Irish Children in Philadelphia in the Nineteenth Century,” Pennsylvania Magazine of History and Biography, 101 (October 1977): 475–86. Clemens, Paul G. E., and Lucy Simler, “Rural Labor and the Farm Household in Chester County, Pennsylvania, 1750–1820,” in Stephen Innes, ed., Work and Labor in Early America (Chapel Hill, NC, 1988): 106–43.

368   German immigrant servitude in America, 1745–1835 Cloud, Patricia and David W. Galenson, “Chinese Immigration and Contract Labor in the Late Nineteenth Century,” Explorations in Economic History, 24 (January 1987): 22–42. Coase, R. H., The Firm, the Market, and the Law (Chicago, 1988). David, Paul A., “Industrial Labor Market Adjustments in a Region of Recent Settlement: Chicago, 1848–1868,” in Peter Kilby, ed., Quantity and Quiddity: Essays in U.S. Economic History (Middletown, CT, 1987): 47–97. The Directory and Stranger’s Guide for the City of Charleston for the Year 1815 [and for years 1816–1823] (Charleston, SC, 1819). Duden, Gottfried, Report on a Journey to the Western States of North America (London, 1980). Dunn, Richard S., Sugar and Slaves (New York, 1972). Dunn, Richard S., “Servants and Slaves: The Recruitment and Employment of Labor,” in Jack P. Greene and J. R. Pole, eds., Colonial British America: Essays in the New History of the Early Modern Era (Baltimore, 1984): 157–94. Durnbaugh, Donald F., “Two Early Letters from Germantown,” Pennsylvania Magazine of History and Biography, 84 (April 1960): 219–33. Eleutherian Mills Historical Library, Unpublished Manuscripts, Greenville, DE. Engerman, Stanley L., “Servants to Slaves to Servants: Contract Labour and European Expansion,” in P. C. Emmer, ed., Colonialism and Migration Indentured Labour Before and After Slavery (Boston, 1986): 263–94. Erickson, Charlotte, “Why Did Contract Labour not Work in the Nineteenth-­Century United States?” in Shula Marks and Peter Richardson, eds., International Labour Migration, Historical Perspectives, vol. 24 (Hounslow, England, 1984): 34–56. Fearon, Henry Bradshaw, Sketches of America (3rd edn., London, 1819). Galenson, David W., White Servitude in Colonial America (New York, 1981). Galenson, David W., “The Rise and Fall of Indentured Servitude in the Americas: An Economic Analysis,” Journal of Economic History, 44 (March 1984): 1–26. Geiser, Karl F., Redemptioners and Indentured Servants in the Colony and Common-­ wealth of Pennsylvania (New Haven, 1901). Gemery, Henry A., “Markets for Migrants: English Indentured Servitude and Emigration in the Seventeenth and Eighteenth Centuries,” in P. C. Emmer, ed., Colonialism and Migration: Indentured Labour Before and After Slavery (Boston, 1986): 33–54. Goldin, Claudia and Kenneth Sokoloff, “The Relative Productivity Hypothesis of Industrialization: The American Case, 1820–1850,” Quarterly Journal of Economics, 99 (August 1984): 461–87. Grabbe, Hans-­Jurgen, “Das Ende des Redemptioner-­Systems in den Vereinigten Staaten,” Amerikastudien, 29 (1984:3): 277–96. Green, E. E. R., “Ulster Emigrants’ Letters,” in E. E. R. Green, ed., Essays in Scotch-­Irish History (London, 1969): 87–103. Grubb, Farley, “The Incidence of Servitude in Trans-­Atlantic Migration, 1771–1804,” Explorations in Economic History, 22 (July 1985): 316–39. Grubb, Farley, “Immigrant Servant Labor: Their Occupational and Geographic Distribution in the Late Eighteenth-­Century Mid-­Atlantic Economy,” Social Science History, 9 (Summer 1985): 249–75. Grubb, Farley, “Redemptioner Immigration to Pennsylvania: Evidence on Contract Choice and Profitability,” Journal of Economic History, 46 (June 1986): 407–18. Grubb, Farley, “The Market Structure of Shipping German Immigrants to Colonial America,” Pennsylvania Magazine of History and Biography, 111 (January 1987): 27–48.

Collapse of German immigrant servant market   369 Grubb, Farley, “The Auction of Redemptioner Servants, Philadelphia, 1771–1804: An Economic Analysis,” Journal of Economic History, 48 (September 1988): 583–603. Grubb, Farley, “Servant Auction Records and Immigration into the Delaware Valley, 1745–1831: The Proportion of Females Among Immigrant Servants,” Proceedings of the American Philosophical Society, 133 (June 1989): 154–69. Grubb, Farley, “The Reliability of U.S. Immigration Statistics: The Case of Philadelphia, 1815–1830,” International Journal of Maritime History, 2 (June 1990): 29–54. Grubb, Farley, “German Immigration to Pennsylvania, 1709–1820,” Journal of Interdisciplinary History, 20 (Winter 1990): 417–36. Grubb, Farley, “Does Bound Labor have to be Coerced Labor? The Case of Immigrant Servitude in English Colonial America” (Department of Economics Working Paper No. 92–08, University of Delaware, 1990). Grubb, Farley, “The Transition from Colonial Currency to the U.S. Dollar: A Market Revealed Analysis” (Department of Economics Working Paper No. 90–20, University of Delaware, 1990). Grubb, Farley, “The Long-­Run Trend in the Value of European Immigrant Servants, 1654–1831: New Measurements and Interpretations,” Research in Economic History, 14 (1992): 167–240. Grubb, Farley, “Fatherless and Friendless: Factors Influencing the Flow of English Emigrant Servants,” Journal of Economic History, 52 (March 1992): 85–108. Grubb, Farley, “Educational Choice in the Era Before Free Public Schooling: Evidence from German Immigrant Children in Pennsylvania, 1771–1817,” Journal of Economic History, 52 (June 1992): 363–75. Grubb, Farley, “The Disappearance of Organized Markets for European Immigrant Servants in the United States: Five Popular Explanations Re-­examined,” Social Science History, 18 (Spring 1994): 1–30. Grubb, Farley, German Immigrant Servant Contracts Registered at the Port of Philadelphia, 1817–1831 (Baltimore, 1994). Grubb, Farley and Tony Stitt, “The Liverpool Emigrant Servant Trade and the Transition to Slave Labor in the Chesapeake, 1697–1707: Market Adjustments to War,” Explorations in Economic History, 31 (July 1994): 376–405. Hancock, Harold B., “The Indenture System in Delaware, 1681–1921,” Delaware History, 16 (April 1974): 7–59. Hansen, Marcus Lee, The Atlantic Migration 1607–1860 (Cambridge, MA, 1940). Henninghausen, Louis P., History of the German Society of Maryland (Baltimore, 1909). Herrick, Cheesman A., White Servitude in Pennsylvania (Philadelphia, 1926). Jacoby, Dan, “The Legal Foundations of Human Capital Markets,” Industrial Relations, 30 (Spring 1991): 229–50. Jacoby, Daniel, “The Transformation of Industrial Apprenticeship in the United States,” Journal of Economic History, 51 (December 1991): 887–910. Jameson, J. Franklin, ed., “Letters of Phineas Bond, British Consul at Philadelphia, to the Foreign Office of Great Britain, 1787–1794,” Annual Report of the American Historical Association, 1 (1896): 513–659; 2 (1897): 454–568. Jones, Maldwyn A., “Ulster Emigration, 1783–1815,” in E. E. R. Green, ed., Essays in Scotch-­Irish History (London, 1969): 46–68. Kauffman, Kyle D. and Francisco Cribari-­Neto, “To Pay or Not to Pay: Positive Incentives as a Calibrating Device in the Seventeenth-­Century White Servitude System” (Wellesley College Working Paper No. 93–03, 1993).

370   German immigrant servitude in America, 1745–1835 Klepp, Susan E., “Demography in Early Philadelphia, 1690–1860,” Proceedings of the American Philosophical Society, 133 (June 1989): 85–111. Lindstrom, Diane, Economic Development in the Philadelphia Region, 1810–1850 (New York, 1978). McCormac, Eugene I., White Servitude in Maryland, 1634–1820 (Baltimore, 1904). McCusker, John J., Money and Exchange in Europe and America, 1600–1775: A Handbook (Chapel Hill, NC, 1978). Menard, Russell R., “From Servants to Slaves: The Transformation of the Chesapeake Labor System,” Southern Studies, 16 (Winter 1977): 355–90. Miller, Kerby A., Emigrants and Exiles: Ireland and Irish Exiles to North America (New York, 1985). Morgan, Edmund S., American Slavery, American Freedom (New York, 1975). Morris, Richard B., Government and Labor in Early America (Boston, 1981). Muhlenberg, Henry Melchior, The Journals of Henry Melchior Muhlenberg (Philadelphia, 1942). Nile’s Weekly Register (Baltimore, MD), Sept. 27, 1817 and Jan. 31, 1818. North, Douglass C., “The United States Balance of Payments, 1790–1860,” in Trends in the American Economy in the Nineteenth Century, NBER Studies in Income and Wealth, vol. 24 (Princeton, NJ, 1960): 573–627. Officer, Lawrence H., “Dollar-­Sterling Mint Parity and Exchange Rates, 1791–1834,” this Journal of Economic History, 43 (September 1983): 579–616. Page, Thomas W., “The Transportation of Immigrants and Reception Arrangements in the Nineteenth Century,” Journal of Political Economy, 19 (November 1911): 732–49. “Passenger List of the Ship ‘Elizabeth’, Which Arrived at Philadelphia in 1819,” Pennsylvania Magazine of History and Biography, 25 (1901): 257–60. Pfund, Harry, History of the German Society of Pennsylvania, 1764–1964 (Philadelphia, 1964). “Philadelphia Customs House Records, 1800–1883” (Unpublished manuscript, National Archives, Washington, D.C., microfilm #425, rolls 16–50). “Pilot’s Report for the Port of Philadelphia, April 26, 1793 to September 5, 1808,” (Unpublished manuscript, State Archives of Pennsylvania, Harrisburg, PA, microfilm RG-­41, roll 1). “Record of Indentures of Individuals Bound Out as Apprentices, Servants, Etc. and of German and Other Redemptioners in the Office of the Mayor of the City of Philadelphia October 3, 1771 to October 5, 1773” (Unpublished manuscript, City Archives of Philadelphia). Risch, Erna, “Immigrant Aid Societies Before 1820,” Pennsylvania Magazine of History and Biography, 60 (January 1936): 15–33. Roeber, A. G., Palatines, Liberty, and Property (Baltimore, 1993). Roebling, Johann A., Diary of My Journey from Muehlhausen in Thuringia via Bremen to the United States of North America in the Year 1831 (Trenton, NJ, 1931). Salinger, Sharon V., To Serve Well and Faithfully: Labor and Indentured Servants in Pennsylvania, 1682–1800 (New York, 1987). Schumpeter, Elizabeth Boody, “English Prices and Public Finance, 1660–1822,” Review of Economic Statistics, 20 (February 1938): 21–37. Selig, Robert A., “The Idea and Practice of the Ius Emigrandi in the Holy Roman Empire from the Reformation to the French Revolution,” Yearbook of German-­American Studies, 27 (1992): 15–22.

Collapse of German immigrant servant market   371 Simler, Lucy, “Tenancy in Colonial Pennsylvania: The Case of Chester County,” William and Mary Quarterly, 43 (October 1986): 542–69. Simler, Lucy, “The Landless Worker: An Index of Economic and Social Change in Chester County, Pennsylvania, 1750–1820,” Pennsylvania Magazine of History and Biography, 114 (April 1990): 163–99. Skeen, Edward C., “The Year Without a Summer: A Historical View,” Journal of the Early Republic, 1 (Spring 1981): 51–67. Smith, Abbot Emerson, Colonists in Bondage (New York, 1947). Steinfeld, Robert J., The Invention of Free Labor (Chapel Hill, NC, 1991). Strassburger, Ralph B., Pennsylvania German Pioneers (Norristown, PA, 1934). Tomlins, Christopher L., “Law and Power in the Employment Relationship,” in Christopher L. Tomlins and Andrew J. King, eds., Labor Law in America (Baltimore, 1992): 71–98. Tomlins, Christopher L., Law, Labor, and Ideology in the Early American Republic (New York, 1993). Trautmann, Frederic, “Pennsylvania Through a German’s Eyes: The Travels of Ludwig Gall, 1819–1820,” Pennsylvania Magazine of History and Biography, 105 (January 1981): 35–65. U.S. Bureau of the Census. Historical Statistics of the United States: Colonial Times to the Present (Washington, DC, 1975). Virchaux, Henry T., “Report to the German Society of Pennsylvania on the Ship April” (Unpublished manuscript, German Society of Pennsylvania, Philadelphia, January 22, 1818). Von Furstenwarther, Moritz, Der Deutsche in America (Stuttgart, 1818). Walker, Mark, Germany and the Emigration, 1816–1885 (Cambridge, MA, 1964). Wittke, Carl, We Who Built America (New York, 1939). Wokeck, Marianne S., “The Flow and Composition of German Immigration to Philadelphia, 1727–1775,” Pennsylvania Magazine of History and Biography, 105 (July 1981): 249–78. Wust, Klaus G., Pioneers in Service: The German Society of Maryland, 1783–1958 (Baltimore, 1958).

Part III

Epilogue: German immigration to the U.S., 1820–1920 From founding migration to mass migration

19 German immigration to the U.S., 1820–1920 Magnitudes, patterns, and relative shares

German immigration after 1835 differed from that which occurred before 1835. The period 1709–1835 was the founding era of German immigration to America. It was characterized by an arduous, dangerous, and expensive journey. The yearly flow of immigrants was relatively low. It was frequently suspended completely during warfare along the migration route. For German emigrants, America was a sideshow. Most went elsewhere. For those who chose America, entering a written contract of servitude upon debarkation to pay for the journey was the dominant experience. These constraints influenced who chose to migrate, which in turn affected the familial, occupational, and educational composition of the immigrants. They also influenced their geographic dispersion within, and social integration into, American society. After 1835, German immigration to America changed. The period from 1835 to 1914 was the era of mass migration. The journey became less arduous and dangerous. It also became considerably cheaper, both in terms of time and money spent. The flow of immigrants was seldom completely interrupted by warfare along the migration route, and that flow burgeoned. Even at its lowest troughs, the yearly volume of German immigration to America in the mass migration era was several times higher than the peak volumes during the founding era. After 1835, America was no longer a sideshow, but instead became the primary destination for German emigrants. Lastly, organized markets in America for German immigrant servants ended in the early 1820s. As such, the era of mass migration was also the era of free passengers and free immigrant labor. These changes altered who chose to migrate, which in turn altered the familial, occupational, and educational composition of the immigrants. They also changed their geographic dispersion within, and social integration into, American society. This book is about analyzing the quantitative dimensions of German immigration to America between 1709 and 1835. Contrasting this period with the subsequent era of mass migration provides a comparative perspective that enriches our understanding of the founding migration era. Knowing what happened to German immigration after 1835, namely after the cost of migration had plummeted and after servitude had disappeared, provides a way of assessing the extent that German immigration during the founding era was constrained and shaped by that era’s institutions and history. This chapter, and the following

376   Epilogue chapter, provides such a comparison and, as such, serves as the conclusion or epilogue to the study of German immigration to America during the founding era. Because the goal is to compare post-­1835 German immigration with that found in Chapters 1 through 8 for 1709–1835, the focus will be on assessing the quantitative dimensions of the migration stream and the immigrants’ characteristics. The information presented is drawn from published primary sources and the voluminous secondary literature on the subject, a body of literature so large that it cannot be adequately summarized in the brief chapters here. Chapter 19 charts the 1820–1920 migration process comparable with that done for the founding era in Chapters 1 through 4. Chapter 20 charts immigrant characteristics during 1820–1920 comparable with that done for the founding era in Chapters 5 through 8. The evidence required for making an exact comparison, however, is not always available in the published primary and secondary sources. Because immigrant servitude disappeared among German arrivals soon after 1820, nothing comparable to Chapters 9 through 18 is presented for the 1820–1920 period. Several large and continuous bodies of quantitative evidence are available for charting the nature of German immigration to America after 1820, evidence not available during the founding era. On the American side, the U.S. Federal government systematically collected passenger data from arriving ships after 1819, and the U.S. decennial censuses recorded the nationality of foreign-­born residents beginning in 1850. On the European side, emigration records were kept by various German states after 1820, at the ports of Bremen and Hamburg, and then more systematically for all of Germany after the formation of the Second Reich in 1871. The year 1920 is used as an end point because German mass migration to the U.S. effectively ended at World War I; the Emergency Quota Act passed by the U.S. Congress in 1921 (the Johnson Act) effectively ended the era of unrestricted open immigration to the U.S.; and the 100-year span from 1820 to 1920 is a comparable timeframe to that studied for the founding era in Chapters 1 through 8.1 Figure 19.1 charts German arrivals to the U.S. per year from 1820 through 1920—a total of 5,490,628 or about 40 times more than during the prior 100 years (Chapters 2 and 18). Compared with German immigration in the founding era during peacetime, German immigration in the 1820s and early 1830s was relatively low (Figure 2.1; Hansen 1940: 105–19; Walker 1964: 37–41, 250). It burgeoned thereafter. Between 1836 and 1915 yearly German immigration, even at its lowest troughs, never fell below the peak years of German immigration during the founding era. Often it was well over 10 times the volume of those founding era peaks. Three major waves of German immigrants appear in Figure 19.1, each with sustained volumes over 70,000 per year—1850 through 1857, 1865 through 1874 (with the exception of 1868), and 1880 through 1893. The peaks of these waves were 215,009 arrivals in 1854, 149,671 in 1873, and 250,630 in 1882.

Magnitudes, patterns, relative shares, 1820–1920   377 260 240

Number of immigrants (thousands)

220

FrancoPrussian War

200 180 160 140

Revolutions of 1848

U.S. Civil War

100

60

Fall of Bismark

Foundation of the second German Reich (Empire)

120

80

Some major U.S. financial panics: 1837, 1847, 1873, 1884, 1893, and 1907

U.S. panic of 1837

World War I

40 20 0

1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 Year

Figure 19.1 Yearly German immigration to the U.S., 1820–1920 (source: Carter, et al. (2006: vol. 1, 560–2, series Ad111)).

The magnitude of immigration between 1836 and 1915 clearly distinguishes this era from the preceding era, and rightly earned it the designation of the era of mass migration. However, Figure 19.1 also shows that considerable yearly variation in German immigration, similar to that observed in the founding era, still existed in the mass migration era. As in the founding era, wars affecting the migration route are associated with major troughs in German immigration between 1835 and 1920 (compare Figures 19.1 and 2.1). The first major trough, 1855 through 1865, is associated with the Crimean War (1853–6), American Civil War (1861–5), and military conflicts within Germany (1864–6). The second major trough, 1874 through 1880, is associated with the aftermath of the Franco-­Prussian War (1870–1) and the formation of the Second German Reich in 1871. Finally, World War I effectively quashed German immigration between 1915 and 1920. With the exception of World War I, these other conflicts did not affect the migration route as directly as did wars during the founding era. As such, they did not force immigration to near zero as wars did during the founding era. Many factors in addition to wars influenced the variation in German immigration between 1835 and 1915. Numerous studies have associated this variation with such factors as the U.S. business cycle (or its component measures, such as

378   Epilogue financial panics, railroad construction, residential building, etc.); changes in ocean shipping technology; variation in German agricultural wealth; cycles in German harvest failures; changes in rye prices, income from linen sales, network connections, unemployment rates, and land and looms per capita in Germany; regional variation in German inheritance laws; German wage rates and wage differentials between Germany and America; German demographic changes; religious issues; and so on.2 German immigration also fell after major U.S. financial panics, surged shortly after the political unrest of 1830 and 1848, and collapsed after the fall of Bismarck (see Figure 19.1). Most of these associations, however, come from univariate or limited multivariate models and casual observations. Disentangling these forces and assigning each their proper relative weight in determining the yearly variance in German immigration in a full multivariate model has not been done. Given the co-­ linearity of many of the variables of interest, and the low degrees of freedom statistically available in a times-­series setting, successfully doing such may not be possible. Factors that triggered and then sustained the high levels of German immigration after 1835 are easier to identify. A substantial gap in economic and social well-­being, often measured as income or wage gaps, had existed between Germany and America well before the onset of mass migration. This gap was sizable enough to justify the expense of moving, yet mass migration did not occur until after 1835. The upfront cost of making the move held Germans back. Most were too poor to front this expense in cash, and they were unable to borrow in Europe or America to pay for migration without collateral or without markets that could credibly enforce borrowing agreements that spanned the Atlantic. Using servitude to front the cost of migration was not an attractive option for most. As such, German migration to America was limited pre-­1836 despite the sizable and persistent economic and social gains that could be realized by migrating. This poverty trap was overcome by two major changes in the 1830s. First, economic development and industrialization had gained enough pace to raise incomes, as well as cause some dislocations, which in turn made the migration decision easier and more affordable. Second, migration costs plummeted. Not only did transatlantic passage fares fall, but the cost of getting to embarkation ports also declined. In addition, transit times, both to embarkation ports and then to cross the ocean, fell which in turn reduced the discomfort, disease exposure, food expense, and lost work time experienced while migrating. Mortality on the ocean passage declined precipitously as well. The journey became shorter, safer, and cheaper. Mass migration during the nineteenth century was so large that it reduced the economic gap between Germany and America. Immigrant additions to American labor supplies held American unskilled wages down, and emigrant subtractions from German labor supplies pushed German unskilled wages up. Given enough time and migration, the wage gap between Germany and the U.S. would close, thus choking off further mass migration. In the meantime, mass migration was

Magnitudes, patterns, relative shares, 1820–1920   379 sustained by rising population pressures, the increasing ease of internal migration within Germany, prior German immigrants inducing and financially supporting continued German immigration, and differential development between the U.S. and Germany.3 While German industrialization lagged behind that in Britain and the U.S., it was on the rise in the first half of the nineteenth century. The formation of the German Customs Union in 1834 was a sign of, and helped to further enhance, the expansion of internal trade. By 1850, Germany had more railroad track than any country except the U.S. and Britain. By 1890, Germany had surpassed Britain for second place. Rail networks made internal trade and migration to embarkation ports easier and cheaper. Industrial development helped Germans overcome the poverty trap so that more could afford to pay the cost of migrating overseas out-­of-pocket.4 With the exception of immigration induced by the 1840s Irish potato famine, mass emigration from Europe to the U.S. over the nineteenth century by national group is correlated with the onset of industrial development—Britain being first, followed by Germany, then Scandinavia, then central and eastern European non-­ German areas, and finally by Italy (see Figure 19.3). This correlation places industrialization as one of the key triggers releasing mass migration to the U.S.5 The onset of mass migrations to the U.S. by national group across the nineteenth century is also correlated with the sequence of dominant immigrant groups from the founding era. Germans were the first non-­English-speaking Europeans to participate in mass migration to the U.S. They were also the dominant non-­ English-speaking migration to English America during the preceding founding era (Chapter 2). The fall in transatlantic migration costs was nothing short of spectacular. From continental Europe, passage fares to America in 1820 were three times higher than they would be by 1835. For Germans migrating to America, transatlantic passage fares average $70 between 1802 and 1830, with a slight upward trend across these years. By 1835, they had fallen to $22 (Chapter 18). A similar 70 percent decline in passage fares is reported for Germans embarking at Le Havre, France for the U.S.—from a high of $77 in 1818 to $23 by the mid-­1830s (Hansen 1940: 198). Between 1849 and 1852, passage fares from Hamburg to New York averaged $26 (Wegge 1998: 975; 2002: 378; 2010). A similar fall in transatlantic passage fares to the U.S. across this period occurred in the British and Irish trade (Cohn 2009: 64–5). After 1835, steerage passage fares from continental Europe to the U.S. stayed in the $22 to $40 range all the way to World War I (Keeling 1999: 42–3, 50, 64–5; 2007: 163–5, 168–9). Such low passage fares after 1835 made the voyage affordable for German emigrants. Friedrich Kapp (1870: 142–3) observed that the New York Commissioners of Emigration found in 1856 that the average emigrant was carrying $68 in cash upon arrival. Kapp also thought this average was biased low. The way the commissioners asked their questions made emigrants suspicious that cash sums would be taxed or confiscated, thus inducing emigrants to conceal and underreport their cash holdings. Kapp also reported that German emigrants in

380   Epilogue the 1840s and 1850s from southwestern Germany left with at least $90 and up to $300 in cash each. Walker (1964: 160) found that between 1840 and 1855 the average German emigrant from Bavaria, Prussia, and Baden left with $101, $169, and $74 in cash, respectively. Among these three groups, the lowest yearly average was $52 and the highest was $291. With transatlantic passage fares in the mid-$20 range after 1835, not only was passage to America affordable, but most German emigrants had capital left over after paying their migration expenses.6 Pre-­1830, when transatlantic passage fares were in the $70 range, only German emigrants with above average wealth could cover that fare plus the other costs of migrating in cash out-­of-pocket. Those possessing below-­average wealth would have to use servitude to finance this cost. The prospect of servitude likely stopped many of the less wealthy from emigrating to America. With passage fares reduced to the mid-$20 range after 1835, paying cash out-­ofpocket to cover migration costs was now within reach of German emigrants of middling and below-­average wealth, though probably not within reach of the truly destitute.7 Keeling (2007: 170) estimated that between 1830 and 1900 only 1.7 to 2.5 years of work in Europe would be needed to accumulate enough savings to pay this migration cost. The forces that led to the rise of mass migration were also those that led to free passenger and free labor migration. As migration costs fell and labor productivity rose, servitude as a way for financing transatlantic migration lost viability. The length of adult servant contracts could not be driven below an approximate two-­year threshold. Servant contracts involved significant fixed-­ cost elements, namely the contracting and employment transaction costs, as well as the mortality, morbidity, acculturation, and acclimatization costs of adapting to the work environment in America for German arrivals. These fixed costs meant that servant contract lengths had to be at least in the two-­year range to be cost-­effective for American employers. Given the sizable fixed-­cost element of servitude, contract lengths could not decline proportionate to the fall in passage fares (Chapters 12 and 18; Baines 1991: 45). At relatively low passage fares, servitude was no longer the best choice for German emigrants for covering the cost of migration. After 1835, setting aside the fixed costs of servant contracting and employment, it would have taken only 0.75 to 1.5 years of labor in America to generate a sum of money to cover the cost of migration of out accumulated saving (Keeling 2007: 170). Under these conditions, it made more sense for emigrants to accumulate the necessary cash for migration before emigrating and to enter America as free laborers, rather than to use servitude to finance migration costs. The end of organized markets for German immigrant servants in the U.S. in the 1820s was driven by falling servant supplies. After 1835, the collapse in passage fares, and the rise in cash remittances sent by prior immigrants to new emigrants, meant that organized servant markets for financing migration could not be effectively revived in the German transatlantic immigrant trade after the onset of mass migration, even for poor migrants.8

Magnitudes, patterns, relative shares, 1820–1920   381 Passage mortality fell as well. This reduced an important non-­pecuniary cost to transatlantic migration, one that worried emigrants and likely stopped many from undertaking the journey. Passage mortality on German immigrants ships to America between 1727 and 1805 was 3.8 percent among all passengers and 3.6 percent among adult men—a monthly mortality rate of about 15 per 1,000 boarded (Chapter 4). Between 1820 and 1860, average passage mortality on voyages from Europe to America was 0.9 percent—a monthly mortality rate of about 6.1 deaths per 1,000 boarded. By the onset of mass migration, passage mortality had fallen 59 to 76 percent over that experience in the late colonial period. As in the colonial period, a few ships with disastrously high mortality were observed. While these disastrous voyages grabbed the media headlines, they were statistically rare and mostly represented unlucky cases where contagious diseases, such as cholera and typhus, were unsuspectingly brought on board.9 Shipboard mortality continued to decline over the nineteenth century. Between 1864 and 1869, passage mortality was 0.38 percent in steerage on ships arriving at New York from Europe—0.9 percent on sailing vessels and 0.18 percent on steamships (Abbott 1924: 48; Kapp 1870: 241). By 1870, over 90 percent of transatlantic passengers were traveling on ships with iron hulls using screw propulsion. This conversion from wooden sailing ships to faster, more spacious iron steam ships reduced passage mortality by 90 percent. In the German transatlantic trade, passenger death rates by the 1870s had fallen to equal that on land for non-­emigrants (Keeling 1999: 42, 44, 48). Post-­voyage mortality also declined—by 65 to 85 percent between the mid-­ eighteenth and the mid-­nineteenth century. These were deaths soon after arrival due primarily to lingering diseases contracted during the voyage, or new diseases contracted soon after landing. In the mid-­eighteenth-century German passenger trade, debarkation morbidity was 3.5 percent, and post-­voyage mortality was about 2.2 percent, of those landed alive (Chapter 4). Between 1820 and 1860, the average post-­voyage mortality for all transatlantic passengers to the U.S. was only around 0.33 to 0.81 percent (Cohn 2009: 149). Even at these lower rates, post-­voyage morbidity and mortality were still concerns to U.S. citizens and U.S. authorities, who worked to provide health assistance to sick and dying immigrants, as well as to segregate or quarantine ill arrivals in marine hospitals. Between 1847 and 1860, some 5 to 6 percent of immigrants landing in New York City were admitted to marine or quarantine hospitals, with the average stay being about six weeks (Cohn 2009: 151–2, 159–66; Kapp 1870: 50–60, 125–41). Similarly, the time it took to cross the Atlantic fell substantially between the founding and mass migration eras. This fall was one of the reasons shipboard mortality declined. It reduced the time that contagious diseases, the primary cause of shipboard deaths, could be spread among the passengers. Average time at sea for Germans migrating to America in the late eighteenth century was about 2.0 to 2.5 months (Chapter 2). By the 1840s, transatlantic transit time on sailing ships averaged about 35 days—a decline of about 50 percent over mid-­ eighteenth times. In the early 1850s, with the introduction of the screw-­driven

382   Epilogue steamship, advertisers in Bremen claimed that a man could leave Bremen on the first day of January and be at work on his new farm in Wisconsin on the first day of February (Walker 1964: 160). By the late 1860s, ocean transit time had fallen to 12 days on steamships—a further decline of 66 percent over the 1840s sailing­ship ocean transit time. Thereafter, it declined slowly to about nine days by 1913.10 Finally, the shift to steamships with iron hulls using screw propulsion led to dramatic increases in ship size. Between 1850 and 1914, ship size increased ten-­ fold. Steerage passengers enjoyed more space and comfort on transatlantic routes. On transatlantic passenger steamers, average human-­cargo capacity doubled between the 1860s and 1910s. After 1870, improving technology and increasing ship size did not yield lower passage fares as much as it increased the safety, amenities, space, and comfort for those making the ocean crossing (Keeling 1999: 47–53, 68). During the founding era, America was seldom chosen by German emigrants. During the era of mass migration, German emigrants stopped going elsewhere and overwhelmingly chose the U.S. as their destination. This shift in emigrant destinations was a major change between the founding and mass migration eras and, in part, explains the substantial increase in the magnitude of German immigration to the U.S. between the two eras. Figure 19.2 shows several estimates of the percentage of German emigrants going to the U.S. between 1820 and 1920. In the 1820s, it was only 20 to 30 percent—more like that in the eighteenth century than what was shortly to follow. In the 1820s, eastern Europe and South America were strong competitors for German emigrants. From 1835 to 1915, 80 to 95 percent of German emigrants went to the U.S. As such, emigration data collected in Germany from 1835 to 1915 is informative of German immigrant arrivals to the U.S. during these years.11 Figure 19.2 also shows that German emigration rates (emigrants per 10,000 resident population) were high during the era of mass migration. Given that most German emigrants selected the U.S. as their destination, the variation in the emigration rate in Figure 19.2 closely mirrors the variation in the volume of German arrivals to the U.S. shown in Figure 19.1. This high emigration rate, between 20 and 50 per 10,000 resident population during the three major waves of German immigration to the U.S., was large enough to affect local German labor markets. It reduced the growth of the German labor force, and, thereby, perceptibly raised real wages, raised real GDP per worker, and reduced income inequality in Germany. These effects contributed to the erosion of the economic opportunity gap between America and Germany, and so helped bring the era of mass migration of Germans to the U.S. to an end after 1895 (Hatton and Williamson 2005: 104–25). Finally, Figure 19.2 shows that the percentage of moves that were internal, crossing from one German state to another, was inversely correlated with overseas German emigration rates. Internal migration was relatively high in the

Magnitudes, patterns, relative shares, 1820–1920   383

Percentage

100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Percentage of German emigrants going to the U.S. 1820– 1919 (5-year average)

Percentage of embarkations at Bremen and Hamburg Percentage of bound for the U.S. 1847–70 all movers (yearly) known to go only to German states

Percentage of German emigrants going to the U.S. 1871–1920 (yearly) German emigration (rate) per 10,000 population

1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 Year

Figure 19.2 Yearly German emigration rates and destination percentages, 1820–1920 (sources: Glazier and Filby (1988: vol. 1, xi); Ferenczi and Willcox (1929: vol. 1, 691, 700–1); Köllman and Marschalck (1973: 518); Lubinski (1995: 59); Willcox (1931: vol. 2, 333–5, 353)).

1830s and early 1840s, and then plummeted when overseas emigration rates took off in the mid-­1840s. It stayed relatively low thereafter, with only a brief upward blip when overseas emigration rates experienced a major trough in the early 1860s. Similarly, as overseas emigration rates declined from 1880 through 1910, internal migration picked up again—not shown in Figure 19.2. Bade (1980: 357) found that migration within Germany from east to west, from Prussia and Posen to Rhineland-­Westphalia, grew significantly between 1880 and 1910. Opportunities within Germany that could be gained by migrating from one German state to another were an important alternative for prospective movers when overseas opportunities appeared less attractive (Baines 1991: 49–51). Germans were the largest non-­English-speaking group among U.S. immigrants in the nineteenth century, just as they had been in the eighteenth century. They were, along with the British and Irish, one of the three largest ethnic groups among U.S. immigrants from 1820 into the late 1890s, again just as they had been in the eighteenth century (Chapter 2). Figure 19.3 charts the percentage distribution of the major national/ethnic groups arriving to the U.S. from 1820

384   Epilogue 60 55

Percentage of total U.S. immigration

50 45

Irish German

40

Central and Eastern European (non-German)

35 30 25 20 15 10 5 0

Italian British Scandinavian 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 Year

Figure 19.3 U.S. immigration: ethnic shares, 1820–1920 (source: Carter, et al. (2006: vol. 1, 541–2, 560–2). The ratios plotted from this source include: Ad107/Ad1, Ad108/Ad1, Ad109/Ad1, Ad111/Ad1, (Ad112 + Ad113 + Ad114 +Ad115)/Ad1, and Ad117/Ad1). Notes Central and Eastern Europeans excludes Germans and includes Poles, Russian and Baltic state citizens, and the two ‘other’ columns in the source. The cumulative percentage for the listed groups falls below 50 percent after 1914 because the denominator includes a sizable number of returning citizens during the years spanned by World War I who were not given ethnic designators.

through 1920. In the 1820s, Germans were under 10 percent of the immigrants— a distant third behind the Irish and the British. The German share rose from 7 percent in 1830 to 50 percent in 1854—its nineteenth-­century peak. During these years Germans surpassed the British, becoming second only to the Irish. From 1854 to 1890, while the German share declined to 20 percent of U.S. immigrants, it remained the largest group over this period (1863 excepted). From 30 percent of U.S. immigrants in the early 1880s, the German share fell to under 5 percent by 1910, and finally to zero during World War I. Germans were surpassed for first place by non-­German immigrants from central and eastern Europe in 1889. They lost second place to the Italians after 1894. In the first two decades of the twentieth century, Germans fell to sixth place, behind non-­German immigrants from central and eastern Europe, Italians, British, Scandinavians, and Irish (Jones 1992: 152–76). Overall, the German share of U.S.

Magnitudes, patterns, relative shares, 1820–1920   385 immigrants was either first or second largest between 1830 and 1895 (1863 excepted). Underlying this dominant position was a significant shift in the origin of German emigrants over the nineteenth century. During the eighteenth century, the dominant region of origin for German immigrants to America was the south Rhine region—south and southwest Germany. Using the easiest corridor to ocean ports, this migration followed the Rhine and its tributaries into Holland with embarkation taking place at the Dutch ports of Amsterdam and Rotterdam. Before the American Revolution, 95 percent, and from 1785 through 1804, 65 percent, of German immigrants to America embarked from these two Dutch ports. By the end of the eighteenth century about 30 percent embarked at German ports, such as Bremen and Hamburg. This 30 percent represented immigrants originating from a different region within Germany (Chapter 2). German ports continued to rise in significance through the early decades of the nineteenth century, until securing a dominant position among German embarkations that lasted throughout the era of mass migration. This change in embarkation ports is consist with a significant shift in the source of emigrants from within Germany. Because German overseas emigration after 1835 is largely synonymous with German immigration to the U.S., this change in the source of German emigrants is also a change in the source of German immigrants to the U.S. Figure 19.4 shows the ports of embarkation for German overseas emigrants from 1848 to 1920. Between 1848 and 1854, German ports, principally Bremen and Hamburg, embarked just over half the emigrants. From 1854 to 1860, their share rose to 80 percent. It hovered between 80 and 100 percent from 1860 to 1885, and remained in the 75 to 80 percent range from 1886 to 1915. By comparison, the Rhine-­corridor ports of Rotterdam, Amsterdam, and Antwerp embarked only 10 to 15 percent from 1848 through 1884, and only 15 to 25 percent from 1885 to 1914. Le Havre was the only other port that embarked significant numbers of Germans for America. It was only important early in the mass migration era, embarking perhaps 25 percent of German emigrants between 1830 and 1850. From 1870 to 1900, Le Havre’s share fell to 5 percent. Moving from the founding migration era to the mass migration era involved a radical transformation in embarkation port usage and, thus, in the exit conduits within Germany used by emigrants. While most of this change reflected a shift in emigrant origins from within Germany, some political and commercial events were important as well. Bremen enhanced its port capabilities in 1830 with the completion of Bremerhaven downriver from Bremen at the mouth of the Weser river. This, along with securing treaties with the U.S., allowed Bremen to dominate the U.S. tobacco trade into central Europe. Because of this, shipping ser­ vices for emigrants on the return voyage to America became readily available. Similarly, after the Napoleonic War, France was able to secure treaties with the U.S. whereby Le Havre became a significant importer of U.S. cotton. Much of this cotton was bound overland for industrial use in the Alsace-­Lorraine region. This provided a transport conduit for German emigrants to America who resided

Percentage

100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Bremen and Hamburg

All German ports

Those going to the U.S. only

All intercontinental German citizen emigrants

Bremen and Hamburg Antwerp, Rotterdam alternative and Amsterdam estimate Le Havre All French ports

London and Liverpool 1850

1860

1870

1880

1890

1900

1910

1920

Year

Figure 19.4 Ports of embarkation for German emigrants, 1847–1920 (percentage distribution) (sources: derived from Carter, et al. (2006: vol. 1, 560–2, column Ad111); Ferenczi and Willcox (1929: vol. 1, 697); Glazier and Filby (1988: vol. 1, xiii); Willcox (1931: vol. 2, 333, column 2)). Notes Two data sets, one for 1847–70 and one for 1871–1920, were aligned by embarkation ports. A vertical line at 1870 is used in the figure to note that data sets with slightly different constructions were used. The 1847–70 data include only those emigrants bound for the U.S. Because of this, U.S. German immigration totals are used as the denominator for pre-1871 calculations. U.S. German immigration totals exceed German emigration to the U.S. from Bremen and Hamburg in all years except 1862, 1863, and 1868. In those three years the percentage embarking from Bremen and Hamburg for the U.S. was counted as 100. The continuous Bremen and Hamburg estimate from 1847 to 1870 includes all embarkations for the U.S. both German and foreign. However, Figure 19.3 shows that almost no non-German immigrants from central and eastern Europe were arriving in the U.S. before 1870. So this estimate is almost exclusively for German emigrants. The alternative estimate for 1850 undercounts German emigration to the U.S. by 31 percent compared with U.S. German immigration totals, thereby biasing the reported percentage estimates for the listed ports down. The 1871–1920 data include only German citizens and counts all intercontinental embarkations, not just those bound for the U.S. Because of this, German emigration totals rather than U.S. German immigration totals were used as the denominator for post-1870 calculations. Post-1870, German ports that were not either Bremen or Hamburg account for only 0.2 percent of German embarkations from German ports (Ferenczi and Willcox 1929: vol. 1, 703). In addition, Figure 19.2 shows that 85 to 92 percent of Germans were bound for the U.S. in this period. As such, the 1871–1920 data is reasonably comparable and aligns well with the 1847–70 data.

Magnitudes, patterns, relative shares, 1820–1920   387 nearby. They could use the conveyances employed in this trade on their return journeys to Le Havre and then to the U.S. Finally, the political, commercial, and military conflicts in the Netherlands over granting independence to Belgium during the 1830s rendered Dutch ports unattractive to German emigrants. These troubles, in part, explain the temporary importance of Le Havre as an embarkation port during the onset of mass migration.12 Figure 19.5 charts the regional origins of German overseas emigrants from 1830 to 1910. Before 1840, as during the eighteenth century, southern and southwestern Germany contributed the dominant share of German immigrants to America, 50 percent or more (Hansen 1940: 289; Walker 1964: 1–69). The share of emigrants from southern and southwestern Germany plummeted from 1840 to 1860. Thereafter, emigrants from that region made up only about 25 percent of the emigrant stream. By the mid-­1850s, central Germany had taken over as the region contributing the largest share of emigrants. It held this dominant position, comprising 40 to 55 percent of the emigrants, until the 1870s. Thereafter its share declined to about 25 percent, similar to that from southern and southwestern Germany. Northeastern Germany contributed under 10 percent to the emigrant stream in the early 1850s. Its share rose to about 40 percent by the early 1870s. From 1870 through 1910, it was the largest regional contributor to the German emigration stream. Evidence from the U.S. decennial population censuses reveals a similar outcome—corroborating to some extent the evidence from the German side of the Atlantic presented in Figure 19.5. In the 1860 and 1870 censuses, German-­born residents often stated from which polity in Germany they had come. Table 19.1 shows the percentage distribution of these statements. It measures the origins of accumulated immigration from earlier years. As such, the magnitudes of, and shifts in, the origins of the immigrants in Table 19.1 will lag behind that reported for current-­year emigrants in Figure 19.5. A substantial percentage stated no polity. This large absent category also adds some margin of error to the measured distribution of origins from the U.S. censuses. In the 1860 and 1870 censuses, at least 32 to 39 percent, respectively, reported that they originated from polities in southern and southwestern Germany. This is consistent with the dominance of this region among German overseas emigrants in the 1830s and 1840s shown in Figure 19.5. The increase in the percentage from northwestern Germany between the 1860 and 1870 censuses is consistent with the substantial share of German overseas emigrants from central Germany (a category which includes Hanover in Figure 19.5) appearing in the 1850s and 1860s. The increase in the percentage from north and east Germany between the 1860 and 1870 censuses is also consistent with the rising share of German overseas emigrants from northeastern Germany between 1850 and 1870 shown in Figure 19.5. In conclusion, during the era of mass migration, major shifts in the origins of German emigrants occurred. At the beginning of the era of mass migration, 1835 to 1845, their dominant origins were the same as they had been in the founding era, namely southern and southwestern Germany. After 1845, the dominant

Percentage

388   Epilogue 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

Central Germany Northeast Germany

Southwest Germany

West Germany 1830

1840

1850

1860

1870

1880

1890

1900

1910

Year

Figure 19.5 Origins of German overseas emigrants by region, 1830–1910 (source: derived from Köllman and Marschalck (1973: 510, 520–1, 535–6). See also Bade (1980: 356); Hansen (1940: 289)). Notes Reported in five-year averages in the original source. Markers are placed in the figure to indicate the middle of each average. The regional construction in the original source differed between the 1830–69 and the 1871–1910 samples. The sub-categories in the original source were re-combined to best align these different regional constructions. Some shift in regional coverage between the 1830–69 and 1871–1910 samples, however, could not be avoided. A vertical line at 1870 is placed as a reminder of this fact. The percentages in the original source for the 1830–69 sample do not add to 100 percent. For 1845–69, the residual is assumed to be central Germany, a region not reported in the original source for these years. Pre-1845, the residual is assumed to be the remainder of Germany outside of the southwest, as no other regions are reported in the original source pre-1845. Southwest Germany for 1830–69 includes Württemberg, Baden, the Palatinate, Hohenzollern, and Bavaria. Southwest Germany for 1871–1910 adds Alsace-Lorraine and the Hesse Grand Duchy to that list. West Germany for 1830–69 includes Westphalia and the Rhineland. West Germany for 1871–1910 adds HesseNassau, Waldeck, and Lippe to that list. Northeast Germany for 1830–69 includes East Prussia, West Prussia, Pomerania, Posen, Silesia, Brandenburg, and Mecklenburg. Northeast Germany for 1871–1910 subtracts Silesia from that list, but includes the Hanseatic cities. Central Germany for 1845–69 includes Hanover, Oldenburg, Schleswig, Holstein, Lauenburg, Anhalt, Brunswick, Saxony, Thuringian States, Hesse-Nassau, Hesse Grand Duchy, Waldeck, and Lippe. Central Germany for 1871–1910 adds Silesia, but excludes Hesse-Nassau, Hesse Grand Duchy, Waldeck, and Lippe, from that list.

Magnitudes, patterns, relative shares, 1820–1920   389 Table 19.1 Percentage distribution of the German-born in the 1860 and 1870 U.S. population censuses by their German region or state of origin Region or State South and Southwest Palatinate Baden Württemberg Bavaria Central Germany Hesse Thuringia Northwest Germany Westphalia Hanover North and East Germany Prussia Saxony Other North German areas Germany “not specified” Total number listed

1860 U.S. Census 32.3% 5.3 8.8 6.4 11.8 9.1 9.1 ? 5.2 5.2 ? 6.5 6.5 ? ? 46.9 1,276,075

1870 U.S. Census 39.2% 10.5 9.1 7.5 12.1 10.4 10.0 0.4 17.5 10.2 7.3 18.5 13.0 2.7 2.8 14.4 1,690,410

Source: Derived from Cohn (2009: 44).

origin shifted to central Germany, which remained the dominant origin through the first and second major waves of German emigration to the U.S. (1845–70). Thereafter, it shifted to northeastern Germany for the third and final major wave of German emigration to the U.S. (1875–94).13 Where German immigrants landed and where they ended up residing in America also changed significantly between the founding era and the mass migration era. Between 1727 and 1775, 75 percent of German immigrants landed at Philadelphia (Chapter 2). As a consequence, by 1790, just over half of German-­ Americans were concentrated in Pennsylvania, with another 25 percent in the neighboring states of New York, New Jersey, and Maryland (Chapter 1). That over half of German immigrants during the founding era had to enter servant contracts with American employers at debarkation to pay for their journey explains, in part, why German immigrants resided predominantly in eastern Pennsylvania and, to a lesser extent, in the states bordering Pennsylvania (Chapters 11 and 17). The predominant use of the port of Philadelphia, the existence of a large and long established German immigrant servant market there, and the existence of a sizable German community in the Delaware valley were mutually reinforcing. These factors combined to maintain Philadelphia and the surrounding region as the dominant port of debarkation and area of residence for German immigrants during the eighteenth century. The disruption to immigration caused by the Napoleonic War and the War of 1812, and disappearance of the German immigrant servitude by the early 1820s,

390   Epilogue

Percentage

was accompanied by a significant change in how German immigrants entered America (Jones 1992: 63–6). By the 1820s, only 9 percent of German immigrants to the U.S. entered through Philadelphia (Chapter 2). Philadelphia’s share would fall to under 2 percent for the rest of the nineteenth century. After the 1820s, the port of New York rose to dominate not just German immigrant debarkations but all European immigrant debarkations. Other ports were insignificant by comparison. Figure 19.6 shows the percentage of German immigrants landing at various ports from 1840 to 1890. In 1840, New York landed 65 percent of German immigrants. Through the rest of the century, New York’s share varied between 70 and 85 percent. In 1840, New Orleans landed about 17 percent, with its share 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

German immigrants landing at New York City

All immigrants landing at New York City

New Orleans

German immigrants landing at:

Philadelphia

Boston

1840

1850

1860

Baltimore

1870

1880

1890

Year

Figure 19.6 Percentage distribution of German immigrants to the U.S. by port of arrival, 1840–90 (source: derived from Bromwell (1969: 105, 109, 113, 117, 121, 125, 129, 133, 137, 141, 145, 149, 153, 157, 161, 165, 169); Glazier (2002: vol. 1, xii), see also Glazier and Filby (1988: vol. 1, xiii). Notes Markers indicate the year for which data was reported (1840, 1850, and so on). Glazier (2002: vol. 1, xii) incorrectly reports the numbers recorded as percentages. They are, in fact, raw immigration totals expressed in 1,000s. These numbers were converted to percentages using the totals reported in Glazier (2002: vol. 1, xii). While these totals are slightly different than the totals reported in Carter, et al. (2006: vol. 1, 560–2, series Ad111) that were used in Figure 19.1, they were nevertheless used in this figure as the base to be consistent with the source reporting the subtotals by port. Bromwell (1969) only reports port totals yearly through 1855.

Magnitudes, patterns, relative shares, 1820–1920   391 declining to 10 percent by 1860 and then to near zero by 1880. Baltimore landed 16 and 20 percent in 1840 and 1890, respectively, but only between 5 and 10 percent between 1850 and 1880. Boston and Philadelphia landed under 5 percent in the nineteenth century. From 1840 through 1855, the share of German immigrants using the port of New York was slightly higher than that for all European immigrants landing in the U.S. From the ports of New York and, to a lesser extent, New Orleans, German immigrants spread across America. The relatively confining settlement patterns of the founding era, namely the areas of eastern Pennsylvania and its borderlands, were replaced with less confining settlement patterns. The use of immigrant servitude during the founding era confined many German immigrants to the environs of their ports of debarkation, principally Philadelphia and to a lesser extent Baltimore. Servant contracts had to be signed with American employers on board ship, while docked in port, before debarking. These employers, given the travel costs, came primarily from that port and its hinterlands (Chapters 11 and 17). With the disappearance of servant markets in the 1820s, German immigrants were freer to chose other places in the U.S. to reside. Table 19.2 lists the top dozen locations of German-­born residents as reported in the 1850 U.S. census. The 10 states with the highest percentage of their population being German-­born, in descending order, were Wisconsin, Missouri, Louisiana, Ohio, Maryland, Texas, Illinois, New York, Iowa, and Pennsylvania. Of all the German-­born listed in the 1850 census, the eight states with the most, in Table 19.2 Top dozen U.S. states with German-born U.S. residents in 1850 U.S. states

All northeastern states New York Pennsylvania All southern states Maryland Louisiana Texas Kentucky Missouri All midwestern states Ohio Indiana Illinois Iowa Wisconsin Total number listed

State population Percentage of state in 1850 population that is German-born in 1850 8,626,629 3,097,394 2,311,786 6,460,501 492,666 272,953 154,431 771,424 594,622 4,721,649 1,980,427 988,416 851,470 192,214 305,391 19,987,571

Source: Derived from Cohn (2009: 168, 170).

2.5% 3.8% 3.4 1.9 5.5 6.4 5.3 1.8 7.5 4.9 5.6 2.9 4.5 3.7 11.3

Percent distribution of all German-born residents in 1850 among the states 37.4% 20.7% 13.7 21.9 4.7 3.1 1.4 2.4 7.7 40.1 19.4 5.0 6.7 1.2 6.0 573,225

392   Epilogue descending order, were New York, Ohio, Pennsylvania, Missouri, Illinois, Wisconsin, Indiana, and Maryland. Table 19.3 shows a similar dispersion of German immigrants across America with respect to urban residences. Among the 15 largest U.S. cities, the German-­ born represented a significant share of those cities’ overall populations, especially in cities in the center of the country, such as Cincinnati, St. Louis, Louisville, Chicago, and Milwaukee (17 to 36 percent). On the east coast, just over 10 percent of the populations of Baltimore and New York City were German-­born. The German-­born were also the majority of the foreign-­born populations in Cincinnati, Louisville, St. Louis, Milwaukee, and Baltimore. While German immigration was largely a Delaware valley affair during the founding era, after 1820 it became a broadly American affair with German settlement concentrations scattered across the length and breadth of the U.S.14 This dispersion pattern continued. The results can be seen in the distribution of the German-­born in the 1920 census. In 1920, three states had 30 percent or more of their foreign-­born population being German-­born, namely Wisconsin, Iowa, and Kentucky. Another six states had between 20 and 30 percent of their foreign-­born population being German-­born, namely Maryland, Indiana, Arkansas, Missouri, Kansas, and Nebraska. Only seven states had less than 5 percent of their foreign-­ born population being German-­born, namely Arizona, New Mexico, Rhode Island, Massachusetts, Vermont, New Hampshire, and Maine. Finally, the German-­born comprised the largest single nationality among the foreign-­born in 11 states, namely Nebraska, Kansas, Oklahoma, Iowa, Missouri, Arkansas, Wisconsin, Illinois, Indiana, Kentucky, and Ohio (Carpenter 1927: 122, 124). The accumulation of Table 19.3 Percentage of German-born residents in the 1850 U.S. population census in the largest U.S. cities U.S. cities

Population in 1850

Percentage of the city’s population that are German-born residents

Percentage of the city’s foreign-born population that are German-born

New York Philadelphia Baltimore Boston Providence, RI Newark, NJ Washington Portland, ME New Haven, CT Cincinnati New Orleans St. Louis Louisville Chicago Milwaukee

513,485 408,045 165,983 135,625 41,434 38,883 37,812 20,777 20,338 115,099 99,071 74,926 37,540 29,375 19,963

10.9% 5.6 11.7 1.3 0.2 9.8 3.3 0.2 1.4 29.1 11.5 30.1 20.0 17.3 36.4

23.8% 18.8 54.7 3.8 0.9 30.9 29.2 1.2 7.7 61.4 23.4 58.8 60.2 32.4 56.9

Source: Derived from Cohn (2009: 172).

Magnitudes, patterns, relative shares, 1820–1920   393

Percentage

German immigration over the prior decades had created a contiguous core of German settlements across the middle states of the U.S., from Wisconsin to Arkansas and Nebraska to Ohio. In this region, the German-­born were the dominant group among the foreign-­born and a sizable influence among the general population. Overall, the German-­born comprised 3 to 4 percent of the U.S. population between 1850 and 1920 (see Figure 19.7). Among the foreign-­born in the U.S. population, the German-­born were the second largest behind the Irish-­born from 1850 to 1870. From 1880 through 1920, the German-­born became the largest. From 1850 to 1900, the German-­born were 26 to 31 percent of all the foreign-­ born in the U.S. While this percentage declined from 26 percent in 1900 to 12 percent in 1920, the German-­born were still the largest national group among the foreign-­born. What German immigrants had been to Pennsylvania and the Delaware valley in the founding era—in terms of their relative numbers to other immigrant 46 44 42 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0

Irish foreign-born/ all foreign-born German foreign-born/ all foreign-born

British foreign-born/ all foreign-born

All foreign-born/U.S. population German foreign-born/U.S. population

1850

1860

1870

1880

1890

1900

1910

1920

Year

Figure 19.7 U.S. decennial censuses: foreign-born percentages, 1850–1920 (source: derived from Gibson and Lennon (1999: Tables 4 and 13)). Notes From 1850 through 1900, the Irish, Germans, and British are the three largest national groups among the foreign-born in the U.S. population. After 1900, while Germans remain the single largest national group of foreign-born in the U.S. population through 1920, the Irish fall from second in 1910 to fifth in 1920. The British fall to fourth in 1910 through 1920. Italians took over third place in 1910 and second place in 1920. The foreign-born from what would become the Soviet Union took over third place in 1920.

394   Epilogue groups and to that resident population—they now were to America in general during the mass migration era. The dispersion of German immigrants across America by the end of the nineteenth century diluted the impact that the stock of prior German migrants had on channeling the intended settlement of new German immigrants. In other words, by the end of the nineteenth century, the chain migration effect that led to relatively narrow concentrations within the U.S. of where non-­English-speaking immigrant groups congregated was weaker among the Germans than among other non-­English-speaking immigrant groups (Baines 1991: 29; Dunlevy and Gemery 1978: 910–1). In many ways, America was now German-­America, especially in the U.S. heartland.

Notes   1 See Baines (1991: 11–14; 67–8); Bromwell (1969); Carpenter (1927); Carter (2006: vol. 1, 523–64); Ferenczi and Willcox (1929: vol. 1, 686–709); Gibson and Lennon (1999); Glazier (2002–4); Glazier and Filby (1988–2002); Grubb (1990); Hatton and Williamson (2005: 181–8, 199–200); Hutchinson (1981: 11–213, 535–6); Jones (1992: 212–38); Köllman and Marschalck (1973); Willcox (1931: vol. 2, 313–89).   2 See Bade (1980); Baines (1991: 7, 15–25, 31–2); Hansen (1940: 105–306); Hansen (1940: 120–71); Hatton and Williamson (2005: 31–76); Hughes and Cain (2011: 323); Jones (1992: 78–99); Kamphoefner (1987: 23, 25, 53, 55, 63–9, 95; 1995: 28); Keeling (1999: 68); Köllman and Marschalck (1973: 522–41); Orr (1980); Reichmann, Rippley, and Nagler (1995); Robertson (1967: 105–85); Thomas (1954); Walker (1964); Wegge (1998; 1999; 2002); Willcox (1931: vol. 2, 341–4).   3 See Baines (1991: 28–30); Hatton and Williamson (2005: 31–76); Hoerder (1985: 14); Kamphoefner (1987: 70–105); Wegge (1998).   4 See Hansen (1940: 191–3); Hatton and Williamson (2005: 37); Walker (1964: 88).   5 See Bade (1980); Baines (1991: 7, 18); Hatton and Williamson (2005: 31–76); Hochstadt (1996); Hoerder (1985); Jackson (1995); Kamphoefner (1987: 12–69); Meyer (1995); Thomas (1954: 120); Walker (1964: 50–3, 157–9, 181–94). France is an exception. France’s unusual demography compared with other European industrializing counties may account for this (Baines 1991: 15–6).   6 See also Wegge (2002: 379–88) on the cash carried by Hessian emigrants in the 1850s.   7 See Hoerder (1985: 9); Kamphoefner (1987: 40–69); Wegge (2002: 379–86; 2010).   8 See Chapters 2, 12, 17, and 18; Baines (1991: 38, 40, 45–6); Galenson (1984: 13–24); Hatton and Williamson (2005: 39–41); Jones (1992: 160); Kapp (1870: 121–2).   9 See Abbott (1924: 16–17, 29–33, 42–8); Baines (1991: 40); Cohn (2009: 137–50); Kapp (1870: 23–40). 10 See Baines (1991: 40–1); Hansen (1940: 172–80); Keeling (1999: 44, 46). The average time spent in steerage from Britain to the U.S. also fell by about 25 percent across the 1840s and 1850s (Hatton and Williamson 2005: 38–9). 11 See Hansen (1940: 105–19); Köllman and Marschalck (1973: 519); Walker (1964: 31–42, 205, 250). 12 See Cohn (2009: 133–7); Hansen (1940: 121, 185–98); Walker (1964: 87–92, 142–3, 154). 13 These patterns are reflected in the location of case studies of emigration across the nineteenth century. See Andelson (1995: 63); Bade (1980); Höndgen (2004: 20); Jackson (1995); Kamphoefner (1987: 12–105, 170–200); Kattner (1995: 240–8); Lubinski (1995); Meyer (1995); Wegge (1998, 1999, 2002, 2010). A general impression of this same shift in origins over the nineteenth century is also provided in Walker (1964).

Magnitudes, patterns, relative shares, 1820–1920   395 14 Such patterns can also be seen in the variety of local-­area studies of German-­ immigrant communities in the U.S. For example, see Brinkmann (2004); Höndgen (2004); Gutmann, et al. (2004: 144–5, 153); Jones (1992: 78–211); Kamphoefner (1987); Mikkelsen (1985); and the numerous studies in Reichmann, Rippley, and Nagler (1995).

References Abbott, E. (1924) Immigration: select documents and case records, Chicago: University of Chicago Press. Andelson, J. G. (1995) “From the Wettweau to Ebenezer and Amana: a demographic profile of the inspirationists in America,” in E. Reichman, L. J. Rippley, and J. Nagler (eds.) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­American Center. Bade, K. J. (1980) “German emigration to the United States and continental immigration to Germany in the late nineteenth and early twentieth centuries,” Central European History, 13: 348–77. Baines, D. (1991) Emigration from Europe, 1815–1930, New York: Cambridge University Press. Brinkmann, T. (2004) “The dialectics of ethnic identity: German Jews in Chicago, 1850–1870,” in W. Helbick and W. D. Kamphoefner (eds.) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German-­American Studies. Bromwell, W. J. (1969) History of immigration to the United States, New York: Arno Press. Carpenter, N. (1927) Immigrants and their children, 1920, Washington, D.C: United States Department of Commerce, Bureau of the Census, Government Printing Office. Carter, S. B., Gartner, S. S., Haines, M. R., Olmstead, A. L., Sutch, R., and Wright, G. (2006) Historical statistics of the United States: earliest times to the present, millennial edition, volumes 1–5. New York: Cambridge University Press. Cohn, R. L. (2009) Mass migration under sail: European immigration to the antebellum United States, New York: Cambridge University Press. Dunlevy, J. A. and Gemery, H. A. (1978) “Economic opportunity and the responses of ‘old’ and ‘new’ migrants to the United States,” Journal of Economic History, 38: 901–17. Ferenczi, I. and Willcox, W. F. (1929) International migrations, volume I, New York: National Bureau of Economic Research. Galenson, D. W. (1984) “The rise and fall of indentured servitude in the Americas: an economic analysis,” Journal of Economic History, 44: 1–26. Gibson, C. J. and Lennon, E. (1999) “Historical census on the foreign-­born population of the United States: 1850–1990, tables 4 and 13,” Washington, DC: U.S. Bureau of the Census, Population Division. Online (internet release date: 9 March 1999). Available http://www.census.gov/population/www/documentation/twps0029/twps0029.html (accessed 11 October 2010). Glazier, I. A. (2002–2004) Germans to America, series II: lists of passengers arriving at U.S. ports in the 1840s, volumes 1–7, Wilmington, DE: Scholarly Resources Inc. Glazier, I. A. and Filby, P. W., eds. (1988–2002) Germans to America: lists of passengers arriving at U.S. ports, volumes 1–67, Wilmington, DE: Scholarly Resources Inc. Grubb, F. (1990) “The reliability of U.S. immigration statistics: the case of Philadelphia, 1815–1830,” International Journal of Maritime History, 2: 29–54.

396   Epilogue Gutmann, M., Pullum-­Pinon, S., Baker, S. G., and Burke, I. (2004) “German-­origin settlement and agricultural land use in the twentieth-­century great plains,” in W. Helbick and W. D. Kamphoefner (eds.) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German-­American Studies. Hansen, M. L. (1940) The Atlantic migration 1607–1860, Cambridge, MA: Harvard University Press. Hatton, T. J. and Williamson, J. G. (2005) Global migration and the world economy, Cambridge, MA: MIT Press. Hochstadt, S. (1996) “The socioeconomic determinants of increasing mobility in nineteenth-century Germany,” in D. Hoerder and L. P. Moch (eds.) European Migrants, Boston: Northeastern University Press. Höndgen, A. (2004) “Community versus separation: a Northwest German emigrant settlement region in nineteenth-­century Ohio,” in W. Helbick and W. D. Kamphoefner (eds.) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German-­American Studies. Hoerder, D. (1985) “An introduction to labor migration in the Atlantic economies, 1815–1914,” in D. Hoerder (ed.) Labor migration in the Atlantic economies, Westport, CT: Greenwood Press. Hughes, J. and Cain, L. P. (2011) American economic history, 8th edition, Boston: Addison-­Wesley. Hutchinson, E. P. (1981) Legislative history of American immigration policy, 1798–1965, Philadelphia: University of Pennsylvania Press. Jackson, J. H. Jr. (1995) “Migration in Duisburg, 1821–1914,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Jones, M. A. (1992) American immigration, Chicago: University of Chicago Press, second edition. Kamphoefner, W. A. (1987) The Westfalians: from Germany to Missouri, Princeton, NJ: Princeton University Press. Kamphoefner, W. A. (1995) “German emigration research, north, south, and east: findings, methods, and open questions,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Kapp, F. (1870) Immigration and the commissioners of emigration, New York: Nation Press. Kattner, L. A. (1995) “Land and marriage: German regional reflections in four Texas towns, 1845–1860,” in E. Reichman, L. J. Rippley, and J. Nagler (eds.) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­American Center. Keeling, D. (1999) “The transportation revolution and transatlantic migration, 1850–1914,” Research in Economic History, 19: 39–74. Keeling, D. (2007) “Costs, risks and migration networks between Europe and the United States, 1900–1914,” in T. Feys, L. R. Fischer, S. Hoste, and S. Vanfraechem (eds.) Maritime transport and migration: the connections between maritime and migration networks, Research in Maritime History, No. 33, St. John’s, Newfoundland, Canada: International Maritime Economic History Association. Köllman, W. and Marschalck, P. (1973) “German emigration to the United States,” Perspectives in American History, 7: 497–554.

Magnitudes, patterns, relative shares, 1820–1920   397 Lubinski, A. (1995) “Overseas emigration from Mecklenburg-­Strelitz: the geographic and social contexts,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Meyer, S. (1995) “In-­migration and out-­migration in an area of heavy industry: the case of Georgsmarienhütte, 1856–1870,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Mikkelsen, R. L. (1985) “Immigrants in politics: Poles, Germans, and the social democratic party of Milwaukee,” in D. Hoerder (ed.) Labor migration in the Atlantic economies, Westport, CT: Greenwood Press. Orr, W. J. Jr. (1980) “East Prussia and the revolution of 1848,” Central European History, 13: 303–31. Reich, U. (1995) “Emigration from Regierungsbezirk Frankfurt/Oder, 1815–1893,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Reichman, E., Rippley, L. J., and Nagler, J., eds. (1995) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­ American Center. Robertson, P. (1967) Revolutions of 1848, Princeton, NJ: Princeton University Press. Thomas, B. (1954) Migration and economic growth, Cambridge, UK: Cambridge University Press. Walker, M. (1964) Germany and the emigration, 1816–1885, Cambridge, MA: Harvard University Press. Wegge, S. A. (1998) “Chain migration and information networks: evidence from nineteenth-­century Hesse-­Cassel,” Journal of Economic History, 58: 957–86. Wegge, S. A. (1999) “To part or not to part: emigration and inheritance institutions in nineteenth-­century Hesse-­Cassel,” Explorations in Economic History, 36: 30–55. Wegge, S. A. (2002) “Occupational self-­selection of European emigrants: evidence from nineteenth century Hesse-­Cassel,” European Review of Economic History, 6: 365–94. Wegge, S. A. (2010) “What drives the self-­selection of emigrants? Evidence from 19th century Germany,” Unpublished working paper, Economics Department, College of Staten Island, City University of New York. Willcox, W. F. (1931) International migrations, volume 2, New York: National Bureau of Economic Research.

20 German immigrants in the mass migration era

Emigration is a choice. Personal characteristics such as age, gender, family status, occupation, and education affect that choice. Across these characteristics, emigrants are seldom a random draw out of the populations from which they came, nor are they representative of the populations into which they enter. Emigration is not only a transfer of raw labor, but a transfer of wealth—human-­ capital investments in occupational skills and education—from one country to another. Knowing the characteristics of the emigrants is important to assessing the impact this transfer had on the economies of the sending and the receiving countries.1 The economic gains from permanently changing locations are realized over a lifetime. When the ratio of skilled to unskilled wages is higher in the U.S., these lifetime gains are greater for those with non-­location-specific trade skills. The gains from emigrating attract adults who can realize the most benefits over their remaining lifetime, namely those who are younger, healthier, male, and already possessing some non-­location-specific occupational and educational skills. By contrast, the cost of migrating is paid up front. The cost of emigration favors adults with enough wealth to pay the cost of migration out of accumulated savings and adults without difficult-­to-transfer familial obligations and occupational connections. The costs and the benefits of migration pull adult emigrant characteristics in opposite directions. Adults who are very young typically do not have enough savings to pay the upfront cost of migration. In addition, their trade skills are not fully developed. Skill quality and wealth increase with age, making migration easier and more attractive. However, as adults age the time over which the accumulated benefits from relocating can be realized is shortened, reducing the incentive to migrate. In addition, as adults age they form families, acquire dependents, and possess location-­specific investments in skills, occupations, and capital goods, which increase the cost of migrating (Chapter 7). In the colonial era, the high cost of migrating and the presence of servant markets affected the emigration choice and so the characteristics of the emigrants. Transatlantic immigrant servants were overwhelmingly male, single, and between the ages of 15 and 25, though less so among Germans than among the British and Irish. Young single adults, especially males, could overcome their

German immigrants in the mass migration era   399 inability to pay for transatlantic passage out of accumulated savings by purchasing it through servitude. The same was true for young families. Accompanying spouses and children raised the cost of migration, placing it beyond the resources of many young families. However, this added cost could be covered by having some dependents enter servitude to pay for that cost. The prospect of being an indentured servant in America, however, held many back from emigrating.2 The transition from the founding to the mass migration era was accompanied by a dramatic fall in the cost of migration and the disappearance of organized markets for German immigrant servants. These changes altered the emigration choice. It clearly increased the numbers moving (Chapter 19). Whether it changed the characteristics of those moving is harder to deduce. Among adults, the fall in the cost of migration widened the age and occupational range within which migration was a net benefit. It also increased the ability of single females and families with dependents to pay for migration out of accumulated savings. The disappearance of organized markets for immigrant servants, however, meant that adults without wealth or some other assistance—typically the young, unskilled, and newly married—could no longer secure passage to America by borrowing on their future labor through selling themselves into servitude. Whether the age, gender, family status, occupational, and educational characteristics of German immigrants in the founding era (Chapters 5–8) differed from that in the mass migration era can be settled only through empirical investigation. Yearly time series of these characteristics for German immigrants to the U.S. during the mass migration era, comparable to the characteristics measured in the founding era, potentially can be constructed. The complete age, family, and occupational distributions per year are distillable from the U.S. customs passenger lists (Carter 2006: vol. 1, 523–64; Cohn 2009: 19–20). These lists for Germans arrivals from 1840 to 1897 have been transcribed and published (Glazier and Filby 1988–2002; Glazier 2002–4). In addition, the lists for all passengers landing between 1820 and 1834 at select U.S. ports have been transcribed and published (Bentley 1999; Tepper 1982). To my knowledge, however, a fully controlled completely disaggregated cross-­tabulation of this evidence has yet to be done. Doing such would require a multi-­year commitment of time and resources. Besides the time and effort required to computerize this massive data set, difficult questions about how to code, consolidate, and interpret the data entries would need to be addressed and investigated before the results were usable (e.g. see Cohn 2009: 14–46, 98–124). Such cannot be done here, but it is a valuable project that future scholars should undertake. In the absence of such an empirical exercise, the existing published cross-­ tabulations of the U.S. immigration and German emigration data are employed to construct yearly time series of passenger characteristics. They are not disaggregated enough, however, to isolate the characteristics of interest. For example, the full family structure by ethnicity by year, the full age distribution by gender by ethnicity by year, and the occupational distribution by ethnicity by age by gender by year are not tabulated. In their place, I use the published data

400   Epilogue tabulations that track the percentage male by ethnicity by year, the percentage in particular age categories by ethnicity by year, and the occupations of arrivals by year. These data are re-­manipulated to infer changes over time in the family and adult male occupational structures of German immigrants to the U.S., as well as to compare these structures to those for other U.S. immigrants during the mass migration era and to those for German immigrants during the founding era. Among immigrant groups, the percentage male serves as an ordinal ranking of the labor yield per immigrant. It measures the current relative transfer of raw adult male labor units from one country to another. Families have balanced sex ratios. Spouses and dependent children are typically 50 percent male; 50 percent female. For example, between 1820 and 1855, out of the 988,324 immigrants landing in the U.S. who were under age 15—those likely to be dependent children—53 percent were male. Between 1872 and 1920, out of the 416,095 German emigrants who were under age 14—those likely to be dependent children—51 percent were male.3 Given that few children and single adult females migrate outside of family units, the extent that the male percentage exceeds 50 percent is determined by the share of single adult men among the migrants. Given that almost all adult male immigrants are in the labor force, changes above the 50 percent mark in the male percentage track differences, over time and across immigrant groups, in the current transfer of raw adult male labor units per immigrant from one country to another. In other words, the extent that the male percentage exceeds 50 percent measures changes in the family orientation of the immigrant stream. For adult male immigrants, rearing costs were paid in the sending country while their labor output was enjoyed by the receiving country. As such, knowing how the ratio of dependents to adult male workers changed among German immigrants to America is important for assessing the contemporaneous impact of German immigration on American labor markets and American economic growth. Figure 20.1 charts the male percentage among German immigrants to the U.S. from 1820 to 1920. From 1820 to 1850, it was relatively high—typically in the 60 to 80 percent range, but with a substantial downward trend.4 From 1850 to 1920, it was relatively low—in the 54 to 62 percent range—with a slight downward trend.5 Variation in the male percentage is not related to the magnitude of immigration, compare Figures 20.1 and 19.1. At best, localized peaks in the German male percentage in 1850, 1879, 1865, and 1902 precede by a year or two localized peaks in the flow of German immigrants in 1854, 1866–73, 1882, and 1904. Other than that coincidence, changes in family composition unrelated to migration volume drove changes in the male percentage over time. The male percentage in the founding era differed from that in the mass migration era. The male percentage was 54 percent in 1709–38—well below that in the mass migration era. However, the male percentage was 65 percent in 1785–1820—well above that in the post-­1850 mass migration era (Chapter 5). Looking only at the male percentage, the early founding era is most like the late mass migration era, and the late founding era is most like the early mass

Percentage

German immigrants in the mass migration era   401 88 86 84 82 80 78 76 74 72 70 68 66 64 62 60 58 56 54 52 50 48 46 44

German immigrants to the U.S.

All other non-German immigrants to the U.S. after 1898

German emigrants age 21–50

All other non-German European immigrants to the U.S.

German emigrants age 14–21

1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 Year

Figure 20.1 Percentage male for select immigrant groups, 1820–1920 (source: derived from Ferenczi and Willcox (1929: vol. 1, 401–42, 696, 698–9)). Notes For German immigrants to the U.S., missing data between 1856 and 1868 was filled in with the percentage male among German emigrants embarking at Hamburg. For all other non-German European immigrants to the U.S., missing data between 1856 and 1868 was filled in with a straight line extrapolation between 1855 and 1869. For German emigrants age 14–21 and age 21–50, years when the “sex unknown” and “age not stated” categories were above 10 percent of total emigration were excluded. U.S. immigration data categorizes immigrants by country of origin before 1899 and by race or people after 1898. No adjustments were made in the figure to account for this change. The “All Other Non-German European Immigrants to the U.S.” category includes non-Europeans after 1898.

migration era. In other words, from 1709 to 1920, the male percentage exhibits a long secular arc, starting relatively low in 1709, rising to a peak in the 1820s, then declining until by 1900 it had returned to where it had started in 1709. This arc spans the discontinuity that is the transition to the mass migration era. As such, the male percentage is driven by factors weakly correlated with these two distinct eras. The percentage male is inversely related to the percentage migrating in family units. For 1871 to 1912, Figure 20.1 charts the male percentage among German emigrants between ages 14 and 21 and between ages 21 and 50. Because German emigration is synonymous with German immigration to the U.S. (Figure 19.2),

402   Epilogue the immigration and emigration data in Figure 20.1 are comparable. A large share of the 14–21 age category were dependent children. Males in this category hovered close to the 50 percent mark. The 21–50 age category contains no children and few elderly dependent migrants. The females in this category are wives and single adult migrants. The male percentage for the 21–50 age category is 6 to 15 percentage points above that for the 14–21 age category. The male percentage for all German immigrants lies between the male percentage for the 14–21 and 21–50 age categories. This age structure is consistent with the inverse relationship inferred between the male percentage and the percentage migrating in family units. Immigrants under age 15 are primarily dependent children accompanying adult married migrants. This group has a balanced sex ratio. Therefore, changes in the percentage of immigrants under age 15 reflect changes in the percentage migrating in family units and inversely track changes in the overall male percentage. This inference is further demonstrated in Figure 20.2, which charts the percentage of immigrants under age 15 out of all U.S. immigrants from 1825 to 1920, and the percentage of immigrants under age 14 among all German immigrants to the U.S. from 1899 to 1920. Overlaying Figures 20.2 and 20.1 shows that the percentage male and percentage under age 15 vary inversely. When the percentage under age 15 is relatively high, the male percentage is relative low, and vice versa. Thus, the male percentage inversely reflects the percentage migrating under age 15, which is comparable to inversely reflecting the percentage migrating in family units. Figure 20.2 also charts the percentage of German emigrants under age 11 embarking at Hamburg between 1854 and 1869, and the percentage under age 14 among all German emigrants between 1884 and 1910. These percentages are consistently above the percentage under age 15 for all immigrants to the U.S. Given the inverse relationship between the percentage male and the percentage under age 15, this pattern is also consistent with the male percentage among German immigrants between 1882 and 1917 being below that for non-­German immigrants in Figure 20.1. In addition, Figure 20.2 shows that for 1884–1910 the percentage of German emigrants between ages 21 and 50 varies inversely with the percentage under age 14. This pattern indicates that changes in the under age 14 category are not simply displacements from the 14–20 age category. In other words, changes in the under age 14 category are not reflecting changes in the age structure of dependent children. Therefore, changes in the percentage under age 14 reflects, inversely, changes in the share of single adult emigrants. Both these facts, differences in the percentage male and in the percentage under age 15, point to the same inference, namely that after 1882 the percentage migrating in family units was higher among German immigrants compared with that among other U.S. immigrants (Köllman and Marschalck 1973: 537–8). They also support the inference that among German immigrants to the U.S. the percentage migrating in family units was relatively low in the first half and relatively high in the second half of the nineteenth century.

Percentage

German immigrants in the mass migration era   403 56 54 52 50 48 46 44 42 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6

German emigrants ages 21 to 50 German emigrants under age 11 (Hamburg embarkations only)

German emigrants under age 14

German immigrants to the U.S. under age 14

All immigrants to the U.S. under age 15 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 Year

Figure 20.2 Percentage of immigrants in selected age groups by ethnic category, 1820–1920 (source: derived from Ferenczi and Willcox (1929: vol. 1, 397–8, 444–8, 696, 698–9, 709)). Notes Excludes yearly observations when “age not stated” exceeds 10 percent of the relevant total. For immigrants to the U.S., passenger arrivals are counted before 1868 and immigrant aliens are counted after 1867. In addition, the “Under Age 15” category changes to being under age 14 for the years 1899–1917 and then under age 16 for the years 1918–20. The other age categories reported in the source are not included in the figure because they are not informative, e.g., categories such as ages 15 to 40, ages 14 to 44, over age 44, and over age 49. These categories either mix dependent children with independent adults or capture only a very small (older) portion of the independent adult migrating population.

By contrast, direct measures of the percentage of Germans migrating in family units post-­1820 tell a different story. Köllman and Marschalck (1973: 530) report that 64 percent from Osnabrück between 1832 and 1846, and 72 percent from Baden between 1840 and 1855, emigrated in family units.6 Walker (1964: 186) reports that 57 percent between 1879 and 1887 emigrated in family units. Willcox (1931: vol. 2, 360) reports that 43 percent between 1891 and 1920 emigrated in family units.7 By comparison, from 1709–38 between 74 and 93 percent, and from 1785–1820 approximately 59 percent, migrated in family units (Chapter 5). The percentage male is relatively high in 1785–1840—in the 63+ percent range, and relatively low in 1709–38 and 1850–1920—in the 53 to 62 percent

404   Epilogue range. The direct evidence on the share migrating in family units should inversely track this pattern, but it does not. The family share is relatively high in 1709–38, as expected given the relatively low male percentage in those years. However, it is also relatively high in 1832–55 when the percentage male is also relatively high. The family share is relatively low in 1785–1820, as expected given the relatively high male percentage in those years. However, it is also relatively low in 1850–1920 when the percentage male is also relatively low. As such, the post-­1820 direct evidence on the share migrating in family units presented in the literature is of questionable accuracy. This evidence is derived from emigration records, often from small unrepresentative samples, that may not account for illicit emigration—those moving without legal permission.8 Therefore, the post-­1820 percentage male in Figure 20.1 is a better measure for inferring relative changes in the share migrating in family units. The percentage-­ male measure in Figure 20.1 indicates that the share of Germans migrating in family units was relatively low in the first half and relatively high in the second half of the nineteenth century—the opposite of that reported in the literature based on direct evidence samples.9 One caveat must be addressed. The above inference assumes that the share of females among single adult migrants was relatively constant and insignificant. While evidence on this share is only available before 1820 and after 1881, a substantial change in this share occurred between these dates. As such, changes in the single adult female percentage may account for some of the movement in the male percentage shown in Figure 20.1. This change is an important difference in immigrant characteristics between the founding era and the latter part of the mass migration era. Willcox (1931: vol. 2, 361) reports that, among single adult German emigrants between 1881 and 1920, females accounted for 32 percent.10 By contrast, during the founding era this percentage was 28 and 16 percent in 1709–38 and 1785–1820, respectively (Chapter 5). Changes in this percentage, along with changes in the share of family units among the immigrants, drove changes in the male percentage. The female percentage among single adult German immigrants declined over the founding era helping to increase the overall male percentage. This percentage returned to a higher level by the end of the nineteenth century, helping to decrease the overall male percentage from what it had been before 1840. The difference in the female percentage among single adult German immigrants between the founding era and the late mass migration era may be due to changes in employment opportunities in America. During the founding era the dominant employment experience for single adult female German immigrants was indentured servitude (Chapters 9–11). Besides often involving arduous work on isolated farms, servitude potentially exposed single adult females to sexual abuse and other acts of exploitation by shippers and American masters, more than for adult male servants.11 As such, the prospects of servitude may have discouraged single adult females more than it discouraged single adult males from emigrating in the founding era.

German immigrants in the mass migration era   405 By the late nineteenth century, employment opportunities in America for single adult females, especially in large cities, were more varied, and that employment was less arduous, than those available in the eighteenth century. These jobs also placed females in less vulnerable positions in terms of shippers and employers exploiting them with impunity, sexually or otherwise, compared with that in the eighteenth century.12 These changes lowered the costs and increased the benefits of emigrating for single adult females, relative to single adult males, and so led to an increased share of females among single adult immigrants by the end of the nineteenth century. Figure 20.1 also shows that the male percentage among German immigrants from 1820 to 1842 was slightly higher than that for other U.S. immigrants, who were primarily British and Irish (Figure 19.3). From 1842 to 1854 it was substantially higher, but from 1854 to 1880 it was about the same. These patterns imply a slightly lower percentage of families and single adult females among German immigrants than among British and Irish immigrants between 1820 and 1842, a substantially lower percentage between 1842 and 1854, and about the same percentage from 1854 to 1880. This pattern is the opposite of that in the founding era, where a much higher percentage of families and females were among German immigrants than among their British and Irish counterparts (Chapters 5, 9, and 10). The transition to the mass migration era affected the relative gender and family composition of German immigrants to a different degree than it affected British and Irish immigrants to America. Across this transition, Germans shifted away from female and family-­unit migration relatively more than did the British and Irish. After 1880, the male percentage among German immigrants fell considerably below that for other U.S. immigrants, who were after 1890 primarily from eastern Europe and Italy (Figure 19.3). After 1880, a higher percentage of families and single adult females comprised German immigrants than their counterparts from eastern Europe and Italy. However, regardless of the ethnic composition of the non-­German immigrant category, the percentage male both in levels and in short-­run movements between German and other U.S. immigrants is more similar than different. Conditions affecting all European immigrants, i.e. events in America, must have been important in determining the gender and family composition of the immigrants. In conclusion, between 1709 and 1920, the years with the highest raw adult male labor units per immigrant being added to the American labor force by German immigration—the years with the lowest dependence ratios or female and family oriented migrations—were from 1785 to 1840. This period encompassed both the rapid westward expansion and the initial industrialization of the United States. The uncertainty and wildness of these early opportunities in industry and in the trans-­Appalachian wilderness may have disproportionately attracted German immigrants who were single adult males compared with before or after this period. This was also the period when unskilled wages in America appear to have exceeded those in Germany by the most in the nineteenth century, thereby attracting a disproportionate share of young less-­skilled single adult

406   Epilogue male immigrants.13 As such, German immigrants contributed relatively more raw labor units per immigrant to the current American labor force and current U.S. economic growth during these formative years of accelerating growth, 1785 to 1850, than in periods before or after. Emigration was not only the transfer of raw labor power from one country to another, but the transfer of human capital—investments in occupational and educational skills—from one country to another. The resources and time sacrificed to make these investments were expended in the sending country, and the productive output generated by these investments were enjoyed in the receiving country. Emigration was a net transfer of investment in human capital to America, and so contributed to U.S. economic growth and development. In the eighteenth and nineteenth centuries, adult males were the primary migrants for which this transfer was relevant. Relative to Europe, America was land abundant and capital scarce. The relative returns to capital in America, including human capital, were higher than in Europe. Because emigration is a self-­selection process, those with more human capital have a greater economic incentive to migrate. Therefore, adult male emigrants should be more skilled occupationally and have more education than those they left behind, other things equal. These relative factor scarcities held through the eighteenth and nineteenth centuries. Thus, the incentive for Europeans who were more skilled and educated to emigrate to America persisted throughout this period. During the founding era, adult male German immigrants to America were relatively skilled and literate, more so than British and Irish immigrants, and more so than the American-­born population. For example, between 1793 and 1820, 24 percent were farmers and only 8 percent were laborers or had unlisted occupations. The rest, 68 percent, were spread across numerous skilled trades (the industry occupational category), from baker, blacksmith, and butcher to tailor, tanner, and turner, to name just six of the 77 occupations reported (Chapters 2, 5, 6, and 7). Whether the rise of mass migration changed the occupational distribution among German immigrants is difficult to deduce. The dramatic fall in transportation costs brought the expense of migration within the means of relatively more of the poor and unskilled. Yet, throughout the nineteenth century, the relative payoffs of migrating still favored the skilled and educated. In addition, through the middle of the nineteenth century, the expansion of the American frontier with its massive amounts of cheap farmland, especially after the American Civil War with the passage of the Homestead Act, coupled with rising rural populations and land scarcity in Germany, may have attracted relatively more Germans who wanted to stay in farming to leave for America. Empirical investigation, therefore, is required to determine whether occupational distributions changed across eras. The yearly evidence published on the occupations reported for Germans leaving Europe and arriving in America during the mass migration era makes comparisons with the evidence from the founding era difficult to interpret. The

German immigrants in the mass migration era   407 yearly evidence published for 1820–1920 reports occupations for the entire immigrant population, including women and children. Dependent women and children typically did not have occupations. They were often lumped into the labor, servant, and unlisted occupational categories. Some single adult females were listed in the industry occupational category as weavers, milliners, and seamstresses, and in the professions occupational category as teachers, artists, and actresses. In this evidence, changes in the percentage of immigrants in particular occupational categories may reflect changes in the ratio of adult males to dependent women and children rather than changes in the occupational makeup of those most likely making the migration decision, namely adult males. Several assumptions are required to isolate adult males in the yearly evidence published for 1820–1920. Male immigrants under age 15 and all female immigrants are assumed to be in the laborer, servant, and unlisted occupational category.14 Their numbers are subtracted from the count in that category (the numerator for that category) as well as subtracted from the total count of immigrant occupations (the denominator for all occupational categories). The exception is for adult female immigrants in the 1820–54 evidence who were listed in the industry and professions categories. These females are instead removed from the numerator of their respective categories. These procedures produce a numerator that contains only the occupations of adult male immigrants in that category and an overall denominator that contains only all adult male immigrants. Adult males were both the primary actors making the migration decision and the most relevant group transferring human capital investments from Europe to America. Figure 20.3 charts the occupational distribution per year for adult male immigrants landing in the U.S. Only the percentage in the agriculture, commerce, and industry occupational categories are reported. These categories are the largest, most important, and least susceptible to measurement error. Three separate data sources are strung together to provide the yearly time series shown in Figure 20.3. For 1820–54, the occupational distribution for all adult male U.S. immigrants and the percentage of U.S. immigrants who were German are reported. Comparing the percentage German with the occupational percentages allows for inferences about the occupational distribution of adult male German immigrants. For 1855–70, no published evidence could be found to calculation the yearly adult male German immigrant occupational distribution. For 1899–1920, the yearly occupational distribution of adult male German immigrants to the U.S. are reported. Because females with occupations in agriculture, commerce, and industry could not be directly removed from the calculation, as was done with the 1820–54 evidence, these categories are biased high for these years. The evidence from 1820–54, however, shows that females were never listed in the agriculture and commerce categories. Thus, if 1899–1920 is like 1820–54, these two categories are not biased. In 1820–54, females show up in the industry category. Thus, that category for 1899–1920 is likely biased high, both absolutely and relative to the agricultural and commerce categories. For 1871–98, the yearly occupational distribution of adult male German emigrants in agriculture is reported. Because German emigration is largely

80 75 70 65 60 55 Percentage

50 45

All U.S. immigrants

German emigrants

?

Percentage of U.S. immigrants who were German

German immigration to the US

Industry

40 35 30 25 20

Industry

Agriculture

15 10 5 0

Commerce

Commerce

1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 Year

Figure 20.3 Occupational distributions among adult male immigrants to the U.S., 1820–1920 (sources: derived from Bromwell (1969); Ferenczi and Willcox (1929: vol. 1, 399, 432–48, 452, 698); Köllman and Marschalck (1973: 537, 539)). Notes For 1820–55, the categories in Ferenczi and Willcox (1929: vol. 1, 399) were mapped into the occupations reported in Bromwell (1969). Commerce is composed of merchants, mariners, and clerks. Agriculture is composed of farmers. With the exception of four females in 1829, no females appear in the Commerce and Agriculture categories. Females reported as seamstresses, milliners, and weavers do appear in the Industry category. Females and female occupations were removed from the calculation. Males under age 15 were treated as being in the servant, laborer, or “occupation not stated” categories. They and their occupations were removed from the calculation. For 1871–98, the percentage in Agriculture reported in Köllman and Marschalck (1973: 537, 539) for Germans embarking at Hamburg was multiplied by total German emigration and then divided by the total number of German male emigrants over age 13 (Ferenczi and Willcox 1929: vol. 1, 698). No females, nor any males under age 14, are assumed to be in the Agriculture category. To what extent this category includes agricultural laborers as well as farmers is unclear. Only the Agricultural category is reported because the other categories in this source do not align with those reported in 1820–55 and 1899–1920. In addition, the method used results in biased-high (excessive) percentages in the other occupational categories that cannot be reconciled with the data. For 1899–1920, the numbers reported for German immigrants to the U.S. in the Agriculture, Commerce, and Industry categories are assumed to have no females in them. These numbers are divided by the number of German male immigrants minus one-half the number of German immigrants under age 14 (Ferenczi and Willcox 1929: vol. 1, 432–48, 452). A balanced sex ratio under age 14 is assumed for this group.

German immigrants in the mass migration era   409 synonymous with German immigration to the U.S. (Figure 19.2), this evidence is comparable to the immigration evidence from before 1871 and after 1898. This 1871–98 emigration evidence, however, is problematic. The sum across all occupational categories exceeds 100 percent. Only the agriculture category aligns with the start of the 1899–1920 immigration evidence in Figure 20.3. For these reasons, only the agriculture category is reported for 1871–98. While this particular occupational category appears to be the least biased, the extent that agricultural laborers were consistently placed in this category rather than in the laborer/worker category is unclear. The peak in the percentage in agriculture in 1874 is suspiciously excessive and may be due to some measurement anomaly. Given how the evidence in Figure 20.3 is constructed, interpretative guesswork is required. From 1820 to 1854, the share of adult male immigrants arriving in the U.S. with agricultural occupations rose from 15 to almost 40 percent. This rise was accompanied by the rise in the share of immigrants who were German, from under 5 to almost 50 percent. In addition, the yearly variance in the share with agricultural occupations tracks the yearly variance in the share who were German.15 These two patterns indicate that between 1820 and 1854 the share of adult male German immigrants with agricultural occupations must exceed that reported for all adult male immigrants by a considerable margin. German immigration drove the occupational share in agriculture. By inference, the share of adult male German immigrants arriving with agricultural occupations was likely in the 40 to 50 percent range throughout 1820–54, which is comparable to, if not higher than, that in 1876–90 and above that in 1890–1920.16 Setting aside the 1874 spike as anomalous, from 1820 to 1920 the percentage of adult male German immigrants in agriculture was in the 30 to 50 percent range, except during 1890 to 1905 and 1915 to 1918 when it dipped below the 30 percent range. Agriculture was likely the single largest occupational category among adult male German immigrants in the mass migration era. This evidence indicates that considerably more agricultural sector workers were among German immigrants during the mass migration era than during the late founding era (Cohn 2009: 109; Chapter 5).17 After 1835, the fall in transportation costs and the availability of cheap land in the American interior may have altered the occupational distribution of German immigrants toward farmers and other agricultural workers compared with that during the late founding era. The commerce category exhibits the opposite pattern. For 1820–54, the share of adult male immigrants in commerce moves conversely with the share who were German. By inference, the share of adult male German immigrants in commerce must be less than that for all adult male immigrants. For adult male German immigrants, the share in commerce likely stayed in the 5 to 12 percent range throughout 1820–54, which is also its range in 1899–1914. It only exceeded that range during World War I. Throughout 1820–1920, commerce was likely the least important occupational category with the exception of the professions.18

410   Epilogue For 1820–54, the industry category is unrelated to the trend and variance in the share of Germans among the immigrants. By inference, it is likely that this category was second in importance to agriculture among adult male German immigrants, being somewhere in the 15 to 35 percent range (Cohn 1995: 395; 2009: 109). After 1898 and before World War I, the industry category accounts for roughly a third of adult male German immigrants (see Figure 20.3). This is a biased high measure because females in this category could not be removed from the calculation after 1898. Industry was still likely the single largest occupational category between 1899 and 1906—exceeded again by agriculture from 1906 to 1914. The evidence in Figure 20.3, while a bit speculative, suggests that during the mass migration era the share of adult male German immigrants arriving in the U.S. who were farmers or agricultural sector workers was relatively high, in the 40 to 50 percent range from 1820 to 1890, higher than in the late founder era. Agriculture was the single most important occupational category from the 1830s until 1890, when it was likely eclipsed by the industry category until 1906. Over the nineteenth century, the percentage from the agriculture sector declined, but probably not by a lot until after 1890. The industrial category likely rose somewhat throughout the century. From 1830 to 1880, German immigrants were far more likely to have occupations in the agriculture category, slightly more likely to have occupations in the industry category, and far less likely to have occupations in the commerce, laborers, servant, and unlisted categories than were their primary immigrant competitors, the British and Irish (Cohn 1995: 394–5; 2009: 109; Figure 19.3). Immigrants did not necessarily continue in the same occupations they had at arrival for the rest of their working lives in America. Occupational constraints and payoffs differed between continents and changed over time. For example, immigrant agricultural workers may have found the upfront capital cost of starting a farm in America constraining after exhausting their savings to pay for the expense of migrating. This may have channeled them into other pursuits post-­ arrival. In addition, knowledge of farming in Germany may not have been that valuable when applied to farming on the American plains (Gutmann, et al. 2004). Immigrants were also typically young when they arrived. With age, progress up the occupational ladder in America could lead to more movement into the white collar or professions category post-­arrival than were in this category at arrival (Bade 1980: 364–5; Kamphoefner 1995). Published materials from U.S. decennial censuses from 1850 to 1890 and again in 1910 provide the occupations of the gainfully employed by place of birth. The occupational distribution of German-­born Americans can be taken from these sources and compared with that of the American-­born and with that of other foreign-­born residents. These occupational distributions represent the accumulation of German immigrants still living up to that census. As such, they lag behind contemporaneous movements in the yearly occupational distribution of current immigrants, i.e. those shown in Figure 20.3.

German immigrants in the mass migration era   411 The published census materials do not group occupations by category consistently. These categories also do not consistently match those used in the immigration evidence in Figure 20.3, nor do they always separate the data by gender. While these categories are organized here to align across sources and over time as best as can be, care must be taken when making occupational comparisons over time between censuses and with immigrant data sources. Because the censuses only count occupations for the gainfully employed, the male-­only portion of the census data is a reasonable match, in terms of measuring the same demographic segment of a population, to the adult male immigrant occupational data used in Figure 20.3. Nevertheless, all comparisons that follow should be considered rough approximations at best. Table 20.1 presents the occupational distribution reported in the U.S. decennial censuses from 1850 to 1890 and again in 1910 for German-­born residents by gender. Among German-­born males, 22 percent were farmers from 1850 to 1870, whereas 30 percent were in agriculture from 1880 to 1910. While the percentage in agriculture increased in the second half of the nineteenth century, the percentages are below that implied by the German immigration evidence for 1830 through 1890 in Figure 20.3. The farmer percentages in the 1850–1910 censuses are closer to that for German immigrants at arrival in the late founding period (Chapter 5) compared with that in the mass migration era. The comparison suggests that during the mass migration era some exit out of agricultural occupations by adult male German immigrants, by as much as 10 to 20 percentage points, occurred between arrival and some years after arrival. By contrast, among German-­born males in the 1850 to 1870 censuses, the percentage in the white-­collar/professions and in the skilled and manufacturing/ mechanical (industry) categories exceeded that in the adult male German immigration data by at least 5 to 15 percentage points (as inferred from Figure 20.3). From 1880 to 1910, the percentages in these two categories appear similar between the census and the immigration data. The portion of German-­born males in the trade and transportation category in the 1880 and 1890 censuses is comparable to that in the commerce category among adult male German immigrants in 1899–1920. During the mass migration era, compared with occupations at arrival, some shift away from agriculture and into industry and white collar professions occurred with time spent in America. Table 20.1 also shows that for German-­born residents who were gainfully employed in the 1880 and 1890 censuses only 7 to 11 percent were female. Among these females, 62 percent were laborers or domestic servants and 23 to 28 percent were in manufacturing (industry). These data support the assumptions and procedures used to construct the adult male occupational distributions displayed in Figure 20.3. Most adult female immigrants were without occupations and so were lumped into the unlisted category. The rest were likely to show up in the laborer and, to a lesser extent, the industry categories. This implies that of the major occupational categories constructed for adult male immigrants, those shown in Figure 20.3, agriculture and commerce are the least biased and most comparable across sources and over time.

Males only White collar Skilled Farmers Unskilled Both sexes Agriculture Professional and personal services; laborers Trade and transportation Manufacturing, mining, and, mechanical industries Both sexes Agriculture, mining, and fisheries Professional services Domestic and personal service; laborers Trade and transportation Manufacturing and mechanical industries Males only Agriculture Professional and personal services; laborers Trade and transportation Manufacturing, mining, and, mechanical industries Males only Agriculture, mining, and fisheries Professional services Domestic and personal service; laborers Trade and transportation

Occupation 10.0% 44.3 21.9 23.8

1850 census 13.2% 40.8 21.0 25.0

1860 census 14.8% 39.7 21.8 23.7 (836,418) 26.8% 22.9 13.4 36.9

1870 census

(954,759) 30.5% 17.8 15.5 36.2

(1,033,190) 28.4% 21.2 14.8 35.6

1880 census

(1,337,660) 29.9% 2.1 17.7 14.9

(1,500,617) 27.5% 2.1 22.4 13.9 34.1

1890 census

1910 census

Table 20.1 Percentage distribution of reported occupations in the U.S. census by German-born residents who were gainfully employed, 1850–1910

(78,431) 3.2% 62.5 6.2 28.1 (162,957) 7.8% 2.2 60.9 5.7 23.4

35.4

(1,112,898) 2.4% 29.9 5.7 19.8 16.9 24.5

Notes Numbers in parentheses are the German-born totals as reported in that census. For the 1910 estimate, males over age 9 are counted. Only those classified as the gainfully employed are counted. Shifting occupational categorization and grouping across the censuses and sources required separate row reporting for some censuses. For the complete occupational listings in each category see the notes in the sources cited above.

Sources: Derived from Cohn (2009: 175); Hutchinson (1956: 79, 82–4, 98, 101–5, 121–2, 124–31); Thomas (1954: 148–9).

Manufacturing and mechanical industries Females only Agriculture Professional and personal services; laborers Trade and transportation Manufacturing, mining, and, mechanical industries Females only Agriculture, mining, and fisheries Professional services Domestic and personal service; laborers Trade and transportation Manufacturing and mechanical industries Males only Professionals Proprietors, farmers, and managers Clerks and kindred workers Skilled workers and foremen Semi-skilled workers Unskilled workers

414   Epilogue Table 20.2 compares the occupational distributions of the German-­born in the 1850–1910 censuses with that of the U.S.-born and the non-­German foreign-­ born. Among males from 1850 through 1880, German-­born residents were far less likely to be farmers than the U.S.-born. The German-­born gained some ground on the U.S.-born by the 1890 census, and by the 1910 census the two groups appear similar in terms of the percentage in farming. By contrast, from 1850 through 1910, German-­born residents consistently have a higher percentage in farming than other foreign-­born residents. German immigrants appear more prone to pursue farming in America than their immigrant competitors throughout the mass migration era. This observation is consistent with German immigrants having a relatively high percentage in the agriculture category at arrival compared with other immigrants throughout the mass migration era (see Figure 20.3). Relatively more pursued farming in America, but not as many as those who arrived with agricultural backgrounds. At least to 1890, being American-­born was an important determinant to pursuing and maintaining an agriculture occupation. Table 20.2 also shows that, among males, the German-­born were far more likely to be employed in the skilled/manufacturing (industry) category than were the U.S.-born, and slightly more likely than were the other foreign-­born. This pattern is consistent both with the importance of the industry category among German immigrants throughout the nineteenth century (Figure 20.3) and with some net transition into this category after arrival. From 1850 through 1890, German-­born males were slightly more likely to be in the unskilled/laborer category than were U.S.-born males and slightly less likely to be in this category than other foreign-­born males. In 1910, however, German-­born males were considerably less likely to be in this category than were U.S.-born males and far less likely to be in this category than other foreign-­born males. The differences among males in the other occupational categories are too small to matter. Similar patterns hold among females in the 1880 and 1890 censuses, with the exception that German-­born females in the 1890 census are far less likely to be in manufacturing (industry) than both their U.S.-born and foreign-­born counterparts, the opposite pattern as among German-­born males in that census. In general, throughout the mass migration era, the process of migration and then residing in America for Germans favored occupational pursuits in farming and industry over those involving unskilled labor both absolutely and relative to other U.S. immigrants. In the nineteenth century, near universal literacy and free public schooling were achieved in the U.S. and Germany. As such, an analysis of the determinants of German immigrant literacy and of the choices made by German-­immigrant parents regarding the education of their children—similar to that done in Chapters 6–8—lacks meaning. For example, among the 18,054 German and Swiss male immigrants employed in 1868–9 by the Labor Exchange in New York City, 97 percent were literate. Among the 1,825 female non-­English speaking immigrants employed (87 percent of whom were German), 86 percent were literate.

–0.4 1.0 21.0 5.2 –29.9 4.4 9.4 –10.6

Males only White collar ( 1) (2) Skilled (1) (2) Farmers (1) (2) Unskilled (1) (2)

Male and female combined Agriculture ( 1) (2) Professional and personal services; laborers ( 1) (2) Trade and transportation ( 1) (2) Manufacturing, mining, and, mechanical industries ( 1) (2) Male and female combined Agriculture, mining, and fisheries ( 1) (2) Professional services ( 1) (2) Domestic and personal service; laborers ( 1) (2) Trade and transportation ( 1) (2)

1850 Census

Occupation 0.3 3.2 18.8 2.2 –22.0 3.8 2.9 –9.2

1860 Census

–27.3 5.7 3.9 –11.1 4.6 1.8 18.8 3.6

1.0 4.0 17.5 –1.8 –15.7 5.3 –2.8 –7.5

1870 Census

–20.9 7.3 –0.9 –10.4 5.1 2.4 16.8 0.8

1880 Census

–13.5 2.7 –3.4 –0.2 8.8 –6.4 –3.1 –0.1

1890 Census

continued

1910 Census

Table 20.2 Percentage point differences between German-born occupational distributions and that for (1) the U.S.-born and for (2) the other foreign-born in the U.S. censuses

Females only Agriculture ( 1) (2) Professional and personal services; laborers ( 1) (2) Trade and transportation ( 1) (2)

–23.5 2.0 13.4 –0.4 4.2 3.2

–23.1 5.9 0.7 –8.3 4.3 1.4 18.1 1.0

1880 Census

1890 Census

–16.5 0.8 –2.1 –0.1 7.4 –5.0 –3.6 –1.1 14.7 5.4

1870 Census

Males only Agriculture ( 1) (2) Professional and personal services; laborers ( 1) (2) Trade and transportation ( 1) (2) Manufacturing, mining, and, mechanical industries ( 1) (2) Males only Agriculture, mining, and fisheries ( 1) (2) Professional services ( 1) (2) Domestic and personal service; laborers ( 1) (2) Trade and transportation ( 1) (2) Manufacturing and mechanical industries ( 1) (2)

1860 Census 11.2 4.0

1850 Census

Manufacturing and mechanical industries ( 1) (2)

Occupation

Table 20.2 continued 1910 Census

5.9 –4.8 –2.2 4.4 –10.9 –0.4 28.6 1.9 –3.1 1.5 –12.3 –7.4 –0.9 0.4 0.4 13.2 –4.6 0.4 6.7 2.2 7.3 0.9 –9.8 –17.9

Notes See the notes to Table 20.1. The percentages reported in Table 20.1 were subtracted from the comparable percentages for (1) U.S.-born and for (2) all other non-German foreign-born. A positive number means the German-born were more concentrated in that occupation compared with that for these other groups. A negative number means less concentrated. The number size measures the percentage point gap between the German-born occupational distribution and that for these other groups. For the complete occupational listings in each category see the notes in the sources cited above. For the first four occupation rows, the United States occupational percentages reported were used for (1) the U.S.-born, and the unweighted averages of British and Irish-born occupational percentages were used for (2) all other non-German foreign-born (see Cohn 2009: 175).

Sources: Derived from Cohn (2009: 175); Hutchinson (1956: 79, 82–4, 98, 101–5, 121–2, 124–31); Thomas (1954: 148–9).

Males only Professionals (1) (2) Proprietors, farmers, and managers ( 1) (2) Clerks and kindred workers ( 1) (2) Skilled workers and foremen ( 1) (2) Semi-skilled workers ( 1) (2) Unskilled workers ( 1) (2)

Females only Agriculture, mining, and fisheries ( 1) (2) Professional services ( 1) (2) Domestic and personal service; laborers ( 1) (2) Trade and transportation ( 1) (2) Manufacturing and mechanical industries ( 1) (2)

Manufacturing, mining, and, mechanical industries ( 1) (2)

418   Epilogue By comparison, among the 20,972 British and Irish male immigrants employed, 72 percent were literate. Among the 23,315 British and Irish female immigrants employed, 67 percent were literate. Those employed by the Labor Exchange were likely the poorest and most disadvantaged of the immigrants. Nevertheless, relatively high literacy rates were observed. However, as in the colonial period, literacy rates were higher for German than for English-­speaking immigrants (Chapter 6; Kapp 1870: 116–17). During the founding era, the education of German-­immigrant children in America occurred in the private sector. Being constrained by geography and cost, it was far from universal. Children were taught by parents and employers. Some were sent to fee-­paying church or private schools that were not necessarily German-­language based. A substantial portion of German-­immigrant children paid for their own education as part of their immigrant servant contracts. They agreed to work longer in exchange for employer-­provided education. American employers, who were not necessarily of German heritage, directly instructed their German immigrant servant children in the employers’ homes or sent them to nearby private fee-­paying schools. Most of this education was limited. It amounted to being taught to read and write informally by an employer or parent, or to what could be gained from a year or less of formal schooling. Only occasionally did German immigrant parents explicitly contract with their children’s employers to have education provided in the German language (Chapters 6, 7, 8, 11, and 14). The size of German-­American communities and the rise of free public schooling in nineteenth-­century America changed how the children of German immigrants acquired education. It also changed the kind of education available to these children. The absolute size of German-­American communities, their relative density per square mile in many rural locations and big city enclaves, meant that cost effective German-­language-based parochial schools were viable. Affiliated with German-­Lutheran, German-­Catholic, or German-­evangelical churches, these parochial schools partially served the German-­American communities’ educational demands. Parochial school enrollments rivaled public school enrollments in some locales (Kamphoefner 1987:173–6). The relative size of German-­immigrant communities also gave German-­ Americans political clout. This clout was used to reshape public school systems in areas where German-­Americans were concentrated. Beginning in the 1840s, and then sweeping across a substantial swath of the north-­central U.S., a German-­English bilingual public school system arose. By the close of the nineteenth century, almost a quarter million students studied German in public elementary schools each year. The U.S. entry into World War I, the anti-­German sentiment engendered by that war, and the decline in German immigration between 1890 and 1920, brought bilingual public schooling in German largely to an end (Conzen 2004: 110–11; Fessler 2004).19 The primary method of bilingual education used in the public schools in the nineteenth century was a two-­way immersion program. English-­speaking and German-­speaking children were placed in the same classroom. Part of the day’s

German immigrants in the mass migration era   419 instruction was in German and part in English. Non-­German enrollment in these programs was high. For example, in Cleveland between 1871 and 1895, English-­ heritage students accounted for roughly 35 percent of the enrollment in the German–English bilingual program. Overall, German–­English bilingual programs accounted for a third of total school enrollments in Cleveland. Similar patterns appear in other mid-­western cities. Only in Ohio, however, did state law require German-­English bilingual programs in public schools. On the other hand, restrictive English-­language-only education laws, such as those attempted or passed in Illinois and Wisconsin in 1889, were defeated or rescinded by the power of the German-­American voting bloc. Elsewhere, politics on the local or district level determined the presence of German–English bilingual education in the public school system (Fessler 2004). The integration and acculturation of German immigrants and their offspring into American society differed during the mass migration era compared with that during the founding era. Part of this difference was due to modernization both of the economy and of political institutions, namely the rise of free public schooling, political suffrage with a widening franchise, and the inland transportation and communication revolutions. Part of this difference was also due to the fall in transatlantic migration costs that triggered the mass migration of Germans to the U.S., made German immigrant servitude unviable after 1835, and shifted the occupational distribution of German immigrants toward agriculture. While the German-­born were dispersed geographically and occupationally more widely across America in the mass migration era than during the founding era, and more widely than that for other non-­English speaking immigrant groups, large concentrations or enclaves still existed within this larger dispersion. Ethnic networks, chain-­migration, and congregational cohesion persisted on the local level.20 At the end of the mass migration era, America was in many ways as much German America as it was English or Irish America, just as during the founding era Pennsylvania became in many ways as much German Pennsylvania as it was English Pennsylvania. Large concentrations still mattered. Just as Lancaster County was far more German than the rest of Pennsylvania in the founding era, so the upper mid-­west and central plains states were far more German than the rest of the U.S. at the end of mass migration era. At the same time Germans were also Americanized, learning to use the vote to influence their environment, especially for public schooling, and adapting to the unique occupational opportunities America provided. Immigrant acculturation is a two-­way street. It was certainly so for German immigrants in America. Even a book written by a third-­ generation German-­American, such as this one, who knows little about his German heritage is in some ways still very German: long, dense, earnest, plodding, and occasionally insightful.

420   Epilogue

Notes   1 See Chapters 6 and 7; Baines (1991: 8, 39, 42, 54–61); Cohn (2009: 190–208); Hatton and Williamson (2005: 75–99); Wegge (2002); Willcox (1931: vol. 2, 363–71).   2 See Chapters 5, 9, 10, and 15; Galenson (1981: 23–33); Grubb (1992: 176–7; 2000: 109–13).   3 Derived from Bromwell (1969: 22, 26, 30, 34, 38, 42, 46, 50, 54, 58, 62, 66, 70, 74, 78, 82, 86, 90, 94, 98, 102, 106, 110, 114, 118, 122, 126, 130, 134, 138, 142, 146, 150, 154, 158, 162, 166, 170); Ferenczi and Willcox (1929: vol. 1, 698–9).   4 For 1820 through 1855, 53 percent of U.S. immigrants under age 15 were male. However, in 1821–4 and 1833, 77 percent of U.S. immigrants under age 15 were male. Outside these five years, this percentage exceeded 55 percent only twice, in 1832 and 1840, and then it did not go above 58 percent. For the five years 1821–4 and 1833, the high male percentage among dependent children is anomalous. That only married couples with predominantly male children chose to migrate in these five years, when they did not chose to do so in other years, stretches credulity. There is no differential incentive or payoff in these years that can explain this anomaly. A more likely explanation is that female children were systematically undercounted in these years. Grubb (1990) shows that the U.S. customs passenger statistics suffer from substantial undercounting in the early 1820s. Female children, the “least important” migrants, may have suffered from this systematic undercounting disproportionately. If so, then the high overall male percentage depicted in Figure 20.1 in 1821–4 and 1833 may be just measurement error, namely the systematic undercounting of female children. Correcting for this undercount by inflating the under-­age-15 female percentage to its long run average of 47 percent only lowers the overall male percentage in 1821–4 from 77.0 to 76.5 percent, and in 1833 from 77 to 70 percent. This adjustment, while lowering the overall male percentage depicted in Figure 20.1 for these five years, does not lower it enough to change the fact that the male percentage is still at its 1709–1920 peak in these years. As such, this adjustment does not change the conclusions reached in the text. Derived from Bromwell (1969); Ferenczi and Willcox (1929: vol. 1, 401–42).   5 For German emigrants, in no year between 1872 and 1920 did the percentage male among dependent children—those under age 14—exceed 53 percent. Derived from Ferenczi and Willcox (1929: vol. 1, 698–9).   6 Walker (1964: 160) reports that over 80 percent of emigrants from Baden were in family units in the 1840s, whereas under 70 percent were in family units in the early 1850s.   7 See also Köllman and Marschalck (1973: 544).   8 See the discussions in Ferenczi and Willcox (1929: vol. 1, 686–91); Wegge (2002: 387–9; 2010); Willcox (1931: vol. 2, 330–8).   9 For example, Baines (1991: 39), relying on this direct evidence, concluded, “in the earlier part of the nineteenth century the typical emigrants were a young family . . . There were roughly equal numbers of males and females. By the later nineteenth century the typical emigrants were young single adults and males outnumbered females by about two to one.” This is the opposite of what the percentage male and the percentage under age 15 reported in Figures 20.1 and 20.2 indicates, at least for German immigrants to America. 10 The post-­1880 German emigration data is likely more comprehensive, representative, and less biased than the pre-­1870 emigration data used in the literature. Even if the division between family and single migrants in this data is off by some magnitude, the gender division within the single emigrant category may still be reasonably accurate. 11 See the firsthand accounts of the treatment of single adult German females immigrating to America in 1773 by John Frederick Whitehead and Johann Carl Büttner in Klepp, Grubb, and Pfaelzer de Ortiz (2006). 12 For discussions of German immigrant adult female employment opportunities in the

German immigrants in the mass migration era   421 late mass migration era, see Blaschke (1995); Bretting (1995); Lintelman (1995); Mageean (1995); Sinke (1995); Wehner (1995). 13 Compare the evidence in Chapter 14; Adams (1982); Hatton and Williamson (2005: 54–7); Wegge (2002: 377, 385–6); Williamson (1995: 148–57). 14 Derived from Bromwell (1969). In addition, Wegge (2002: 387) found that between 1852 and 1857 among 1,400 German female emigrants with occupational identifiers, over 90 percent were classified as laborers or servants. 15 A similar association appears between the share of Irish immigrants and the percentage of adult male immigrants in the laborer, servant, and unlisted occupational category, and between the share of British and percentage of adult male immigrants in the commerce occupational category (not shown in Figure 20.3). By inference, these patterns indicate that share of adult male German immigrants in the labor, servant, and unlisted category and in the commerce category were relatively low. Adult male German immigrants were primarily in agriculture and industry in this period. See also Cohn (1995: 394–5; 2009: 109); Wegge (2002: 387). 16 See Cohn (1995: 394–5; 2009: 109); Kamphoefner (1987: 42–51); Köllman and Marschalck (1973: 530–1). 17 By contrast, Wegge (2002: 387) reports that between 1852 and 1857 among 8,929 male emigrants from Hesse only 14.4 percent were listed as farmers, 23 percent were listed as laborer and servants, 7 percent were in the commerce category, and the rest, 55.6 percent, were in industry. Not all regions of Germany may have contributed proportionately to the occupational distribution of German immigrants to America. 18 The professions never account for more than 5 percent of the occupations, and often for under 1 percent. For example, see Browell (1969). 19 I would be remiss to not include some personal anecdotal evidence here. My grandparents, Carl Heinrich Wilhelm Schmale and Anna Marie Louise Knefelkamp, though born in America of German-­immigrant parents, spoke German at home as adults on their farm in central Nebraska. Because of the anti-­German sentiments engendered by World War I, and because their oldest child, born in 1919, had difficulty with English when she started school, my grandparents decided to speak only English at home to the younger children. As such, my mother, born in 1923, retained only a few words of the German language. The side benefit was that my grandparents could then argue with each other in German without their children knowing what the disagreements were about. German language and culture were largely lost in my parents’ generation. 20 For examples of local-­area studies on the integration and acculturation of German immigrants and their offspring into U.S. society and economy during the late nineteenth and early twentieth century, see the articles in Helbich and Kamphoefner (2004); Hoerder and Nagler (1995); Kamphoefner (1987); Reichmann, Rippley, and Nagler (1995).

References Adams, D. R. Jr. (1982) “The standard of living during American industrialization: evidence from the Brandywine region, 1800–1860,” Journal of Economic History, 42: 903–17. Bade, K. J. (1980) “German emigration to the United States and continental immigration to Germany in the late nineteenth and early twentieth centuries,” Central European History, 13: 348–77. Baines, D. (1991) Emigration from Europe, 1815–1930, New York: Cambridge University Press. Bentley, E. P. (1999) Passenger arrivals at the port of New York, 1820–1829: from customs passenger lists, Baltimore: Genealogical Publishing Company. Blaschke, M. (1995) “Communicating the old and the new: German immigrant women

422   Epilogue and their press in comparative perspective around 1900,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Bretting, A. (1995) “Women’s lives in New York City, 1890–1910: the women’s pages of the middle-­class New York Staats-­Zeitung and the socialist New York Volkszeitung,” in E. Reichman, L. J. Rippley, and J. Nagler (eds.) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­American Center. Bromwell, W. J. (1969) History of immigration to the United States, New York: Arno Press. Carter, S. B., Gartner, S. S., Haines, M. R., Olmstead, A. L., Sutch, R., and Wright, G. (2006) Historical statistics of the United States: earliest times to the present, millennial edition, volumes 1–5. New York: Cambridge University Press. Cohn, R. L. (1995) “Occupational evidence on the causes of immigration to the United States, 1836–1853,” Explorations in Economic History, 32: 383–408. Cohn, R. L. (2009) Mass migration under sail: European immigration to the antebellum United States, New York: Cambridge University Press. Conzen, K. N. (2004), “Immigrant religion and the public sphere: the German catholic milieu in America,” in W. Helbick and W. D. Kamphoefner (eds.) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German- American Studies. Ferenczi, I. and Willcox, W. F. (1929) International migrations, volume I, New York: National Bureau of Economic Research. Fessler, P. (2004) “The political and pedagogical in bilingual education: yesterday and today,” in W. Helbick and W. D. Kamphoefner (eds.) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German-­American Studies. Galenson, D. W. (1981) White servitude in colonial America, New York: Cambridge University Press. Glazier, I. A. (2002–2004) Germans to America, series II: lists of passengers arriving at U.S. ports in the 1840s, volumes 1–7, Wilmington, DE: Scholarly Resources Inc. Glazier, I. A. and Filby, P. W., eds. (1988–2002) Germans to America: lists of passengers arriving at U.S. ports, volumes 1–67, Wilmington, DE: Scholarly Resources Inc. Grubb, F. (1990) “The reliability of U.S. immigration statistics: the case of Philadelphia, 1815–1830,” International Journal of Maritime History, 2: 29–54. Grubb, F. (1992) “The long-­run trend in the value of European immigrant servants, 1654–1831: new measurements and interpretations,” Research in Economic History, 14: 167–240. Grubb, F. (2000) “The transatlantic market for British convict labor,” Journal of Economic History, 60: 94–122. Gutmann, M., Pullum-­Pinon, S., Baker, S. G., and Burke, I. (2004) “German-­origin settlement and agricultural land use in the twentieth- century great plains,” in W. Helbick and W. D. Kamphoefner (eds.) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German-­American Studies. Hatton, T. J. and Williamson, J. G. (2005) Global migration and the world economy, Cambridge, MA: MIT Press. Helbick, W. and Kamphoefner, W. D., eds. (2004) German-­American immigration and ethnicity in comparative perspective, Madison: WI: Max Kade Institute for German-­ American Studies.

German immigrants in the mass migration era   423 Hoerder, D. and Nagler, J., eds. (1995) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Hutchinson, E. P. (1956) Immigrants and their children, 1850–1950, New York: John Wiley & Sons, Inc. Kamphoefner, W. A. (1987) The Westfalians: from Germany to Missouri, Princeton, NJ: Princeton University Press. Kamphoefner, W. A. (1995) “Paths of urbanization: St. Louis in 1860,” in E. Reichman, L. J. Rippley, and J. Nagler (eds.) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­American Center. Kapp, F. (1870) Immigration and the commissioners of emigration, New York: Nation Press. Klepp, S. E., Grubb, F., and Pfaelzer de Ortiz, A., eds. (2006) Souls for sale: two German redemptioners come to revolutionary America: the life stories of John Frederick Whitehead and Johann Carl Büttner, University Park, PA: Pennsylvania State University Press. Köllman, W. and Marschalck, P. (1973) “German emigration to the United States,” Perspectives in American History, 7: 497–554. Lintelman, J. K. (1995) “Making service serve themselves: immigrant women and domestic service in North America, 1850–1920,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Mageean, D. M. (1995) “Acculturation of immigrant women in Chicago at the turn of the century,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Reichman, E., Rippley, L. J., and Nagler, J., eds. (1995) Emigration and settlement patterns of German communities in North America, Nashville, IN: Max Kade German-­ American Center. Sinke, S. M. (1995) “The international marriage market: theoretical and historical perspectives,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Tepper, M. (1982) Passenger arrivals at the port of Baltimore, 1820–1834, Baltimore: Genealogical Publishing Company. Thomas, B. (1954) Migration and economic growth, Cambridge, UK: Cambridge University Press. Walker, M. (1964) Germany and the emigration, 1816–1885, Cambridge, MA: Harvard University Press. Wegge, S. A. (2002) “Occupational self-­selection of European emigrants: evidence from nineteenth century Hesse-­Cassel,” European Review of Economic History, 6: 365–94. Wegge, S. A. (2010) “What drives the self-­selection of emigrants? Evidence from 19th century Germany,” Unpublished working paper, economics department, College of Staten Island, City University of New York. Wehner, S. (1995) “German domestic servants in America, 1850–1914: a new look at German immigrant women’s experiences,” in D. Hoerder and J. Nagler (eds.) People in transit: German migrations in comparative perspective, 1820–1930, New York: Cambridge University Press. Willcox, W. F. (1931) International migrations, volume 2, New York:: National Bureau of Economic Research. Williamson, J. G. (1995) “The evolution of global labor markets since 1830: background evidence and hypotheses,” Explorations in Economic History, 32: 141–96.

Index

Page numbers in italics denote tables, those in bold denote figures. Abercrombie, James 69n31 adult males per ship, London vessels compared with Dutch/German departures 23 age: children in servitude by 166; effect of gender and on educational choice 150–2; and marital status of female servants 191; servitude incidence and 162 age distributions: English/German comparison 95, 96; mass migration era 402–3; and social composition 95–6 age structure of literacy: estimation 133–4; interpretation 134–7 Aix-la-Chapelle, Peace of 32 Alsace 22 American Civil War 377, 406 American Revolution 33, 90, 102, 132, 171, 184, 255, 385 Amiens, Peace of 343 Amish 8 anglicized corruption, of German names 19 Argyle 29 arithmetic (ciphering) 146, 148, 152 Austrian Succession, War of the, see War of the Austrian Succession autocorrelation 32 Bailyn, B. 116 bargaining process, Fearon’s observations 311–13 Bauer, Mary Elizabeth 223, 224 Beneset, Daniel 62 “Book A of redemptioners” 19, 303 “Book B of redemptioners” 303 “Book C of redemptioners”: example

contract from 305; health inspections 310–11; information contained 304; passenger contract 308–10; quantitative summary of the information in 306 Britannia, families arriving on 282, 284 British North America, total European immigration to 18 Büttner, Johann Carl 8, 228, 280, 294 buy-out clauses 227, 365n45 cargo space, underutilized 64 chain-migration 32, 43, 359–60, 394, 419 Chesapeake, percentage of English servants going to 198 Chesapeake and Delaware regions, tobacco and grain farming in the 173 children: arbitrage puzzles 293–6; contract analyses 295; evidence of German parents concern about the welfare of their 286–7; exemption from service contracts 166; impact of colonial relief law on 291–3; incidence of servitude in Pennsylvania 165; likelihood of entering servitude 288; and the overseers of the poor 291–5; passage fare structure 277–8; passage mortality 82; sale of into servitude, see sale of children into servitude; servitude among German immigrant 287 children in servitude: by age 166; contract prices/parameters 166 cholera 381 ciphering (arithmetic) 146, 148, 152 climate crisis, in the Rhineland (“year without a summer”) 232, 311–12, 343, 354, 357

Index   425 Coalition Wars 33 colonial entrepreneurs, reliance on the continuing arrival of bound immigrant workers 197 colonial era migrants, characteristics 398–9 colonial literacy: causes of high initial levels of 112–13; and the costs of education 122–3; creolean degeneracy in, see creolean degeneracy in colonial literacy; vs European 107–11, 132; family transmission responsibilities 116; Gallman’s findings 114; impact of German immigrant literacy on 133; impact of immigration 112; intergenerational transmission of 112–13, 115, 117–18; Kilpatrick’s evidence 114; measuring by where acquired 114; as net wealth transfer from Europe to America 121; nonlinear trend in regional 113–17; studies of (1650–1840) 108; supply and demand model of 119; supply vs demand model of growth in 117–21; testing the supply vs demand models 122–4; Tully’s findings 114 colonies: Smith on the prosperity of the 25; wealth position 26–7 comparative perspective, post-1820 German immigration, see mass migration era German immigration congregational cohesion 419 contagious diseases 79, 250, 381 contract examples 222–3, 226, 248, 313 contract length: influencing factors 269; relationship between employer occupation and 269–70 contract register, information held 303 contracting methods: buy-out clauses 227; creation of a new 225, 227; and division of risk 220–1; identification issues 224, 227–8; “indenture of redemption” 224–5, 227, 229; language 227–8; legal terms 227; and migration costs 230; regional perspective 223, 225, 228–9, 231; and skills values 221–3; tradeoffs 222 contracts: average lengths 306; registration process 313 convict immigration, statistics for 18, 72 corporal punishment 273n20 cotton, French imports 385 Craig, J.E. 132 credit price for passage, compared with cash price 248, 308

Crellius, Joseph 58 creolean degeneracy in colonial literacy: economic impact 121; explaining 120; nonlinear pattern 106, 113, 115–17; regional patterns 114; and rural-urban differentials 123, 149; social impact 121 Cressy, D. 132 Crimean War 377 De Koker, Peter 57 debarkation morbidity: affliction types 78; evidence gathering 78–9; and incubation periods 78; and passenger numbers 79, 82; per ship variables 78; percentages 248, 381; percentages (1727–54) 75; relationship between post-voyage mortality and 81 debt pooling 249–50, 321 debt-shifting: arbitrage puzzles 293–6; Britannia manifest evidence 288–90; dominant family strategy 288; impact of colonial relief law 291–3; see also sale of children into servitude defaults, loan 64, 248–50, 259, 261, 278 Delaware valley: estimated population growth (1730–75) 3; explaining the variance in German immigration 33; German immigrants’ contribution 18; new ship construction 39, 41; post-1709 increase in German immigration 5; summary of total German immigration (nonmilitary) into 21; surviving immigration records 19–20; total German immigrants (1812–23) 20; total German immigrants (1817–35) 20 Delaware valley distribution study: British/ German comparison 212; comparison of the percentage distributions of Pennsylvanian occupations with those of servant purchasers (1800 census) 216; context 197–8; distribution by the percentage of resident population by county (1771–3) 200; female servitude 212; geographic distribution of British immigrant servant purchasers (1771–3) 204, 209; geographic distribution of German servant purchasers by county (1771–3) 206; geographic distribution of German servant purchasers by county (1787–1804) 207; geographic distribution of German servant purchasers by town (1771–3) 210; geographic distribution of German servant purchasers by town (1787–1804)

426   Index Delaware valley distribution study contd 211; geographic distribution of German servant purchasers in the Philadelphia market (1745–1804) 199; geographic distribution of Irish indentured servant purchasers by county (1745–6) 202; occupational distribution 213; occupational distribution of German immigrant servant purchasers (1787–1804) 214; sources 198 dependency ratio: causes of fluctuations 93; English/German comparison 90, 91, 92; and family size 93 Dick, John 58, 64 Diffenderffer, F.R. 212 disappearance of immigrant servitude: and the American Revolution 342; available evidence 342; chronological perspective 343–8; and the fall in passage fares 380; monitoring and contracting costs perspective 354–5; regulatory perspective 348; social and occupational perspective 353–4; supply versus demand perspective 349–52; transportation costs perspective 352–3 disappearance of organized immigrant servant markets: abolition of imprisonment for debt hypothesis 319–20; abruptness 356–8; drivers 380; hypotheses 316–17; immigrant aid society activism hypothesis 317, 321–2, 349; and migration costs 399; recession hypothesis 325; runaway hypothesis 322–5; wage labour and demand hypothesis 330, 333–4 disease, and summer arrival 34–5 distribution of immigrant servant labour in the Delaware valley, see Delaware valley distribution study division of risk, methods of acquiring transatlantic passage and 220–1 Duffy, J. 82 DuPont gunpowder mills 367n61 Dutch East India Company (VOC) 8, 230–1 dysentery 75, 78 economic theory of migration 24, 28 education: accounting for expansion in between the 1770s and ‘90s 148; expansion in the provision of 146; see also servant contract education evidence Elizabeth 249, 355 Eltis, D. 8

embarkation ports: distribution of (1727–75) 23; distribution of (1785–1804) 24; distribution of (1847–1920) 386; easier and cheaper travel to 378–9; German/Dutch comparison 23–4; mass migration era 385, 386; passage negotiations at 60; post-colonial rise in importance of German 23; regional distribution of German immigrants’ residence (1798–1808) 22; servitude incidence and distance from 169–70; temporary importance of Le Havre 385, 387 Emergency Quota Act (the Johnson Act) 376 employer occupation, relationship between contract length and 269–70 end-of-contract payments 295, 353 English Channel 29, 36–7, 178n23 English immigrants, dependency ratio, see dependency ratio English indentured servants, literacy rate 112 epidemics 75, 81–2 ethnic networks 419 European immigrant servitude, hypotheses for the ending of 316–17 (see also disappearance of organized immigrant servant markets) European migrations, comparing with German 20 eyewitness descriptions, of the sale of German immigrant redemptioner servants 311 family composition: childless couples and single adults 94; dependency ratio, see dependency ratio; distribution of children 93–4; of German and English immigrants to Pennsylvania (1727–1820) 92; German/English comparison 102; measuring 90 family groups, average return to redemptioners in 250 family status, and servitude 161 family-oriented migration: British immigrants 94; proportions 94 famine: Irish 379; in the Rhineland (“year without a summer”) 232, 311–12, 343, 354, 357 fares 64, 65, 164, 235, 248–9, 279, 344, 346 farm work, and contract length 270 Farrows, N. 326

Index   427 Fearon, Henry Bradshaw 311 Federal Passenger Act 42 Federal passenger lists 19 female education, extent of 152 female proportion of immigrant servants: pre- and post-Revolution 190–1; by region of origin 185, 187, 188, 190 female school-teachers, increasing acceptance of 149 female servants: age and marital status of 191; English 192 female servitude, decline 191 financial panics/crises 34, 325–6, 330, 377, 378 formal schooling: costs of urban compared with rural 148–9; percentage of in education stipulations 148; postRevolution increase in the availability of rural 149 founding era migrants, characteristics 399–400 Franco-Prussian War 377 Franklin, Benjamin 6, 77, 131, 279–80 freedom dues 246, 258–9, 295, 305 freedom, religious 8 freight, term analysis 277 French Revolution 33 frontier schooling, cost 116 Furly, Benjamin 4 Further Account of the Province of Pennsylvania and its Improvements. For the Satisfaction of Those that are Adventurers and Inclined to be so, A (Penn) 34 Galenson, D.W. 79 Gall, Ludwig 25, 229, 320 Gallman, R.E. 114 Gemery, H.A. 79 Georgia 2–3, 56–9, 171–2, 212, 328 German immigration statistics 3, 277, 303, 377, 383 Germany, migration within 382–3 Glazier, I.A. 390 Grubb, F. 123, 303 Hamburg 23 Hancock, John 59 Hansen, M.L. 325, 326 health inspection, an example 310–11 Hector, Charles 27, 71 Hessen 21 Holland: average voyage time to Philadelphia 74; fare structure of

German passenger trade from 277; French occupation of 21; passport regulations 57–8 Homestead Act 406 Hope & Co 56–60, 66 Howell, Samuel 62 human capital: literates’ investment in 137; location-specific 137 Hunt & Greenleafe 59 illiteracy rates: age structure of 135, 136; German/European comparison 132; of German immigrants to Philadelphia 132; urban 133 immigrant literacy: determinants of 137–8; trend and relative position of 131–3; see also literacy immigrant servant sales, percentages by region 198 immigrant servitude: disappearance of, see disappearance of immigrant servitude; impact on geographic distribution and social integration 316; post-Revolution data 342–3; prohibition 348; ways to measure the importance of 342 immigrant-per-ton ratios 38; see also passenger-per-ton ratios impartible inheritance 24, 94 incidence of servitude in Pennsylvania: all immigrants (1771–3) 167; children (1785–1804) 165; by family size (1785–1804) 163; see also servitude incidence indenture: bearer of transportation risks under 220–1, 244; Christopher Sauer’s advice 229; concept analysis 244; guaranteed fixed contract at embarkation under 221; vs redemption 220–1, 244–7, 255; and the verification of skills 221 “indenture of redemption,” example 227 industrial development, and mass emigration from Europe to the US 379 industrialization, German 379 infectious diseases 79, 82, 83 informal instruction, cost advantage in the countryside 149 interwar years, surge in emigration during 64 investment, literacy 120 Irish potato famine 379 Jones, A.H. 26 Kalm, Peter 5

428   Index Keeling, D. 380 Keith, William 5 Kilpatrick, W.H. 114 kinship networks 354, 357, 359–60 Kramer, John 59 labour yield per immigrant, measuring 400 Lemon, J.T. 212 “lemons problem” 231–2, 238, 278 literacy: accumulation process 133; adult male 97, 107–8; age structure of, see age structure of literacy; colonial, see colonial literacy; commercial value of 119–20, 124; and costs of frontier education 97, 121; creolean degeneracy in, see creolean degeneracy in colonial literacy; determinants of immigrant 137–8; European 106–7, 132; in farming occupations 100; female lag in the acquisition of 152; and human capital investment 137; intergenerational persistence 138; as investment good 120; largest and least biased body of evidence of 96; native-born, see colonial literacy; near universal 414; observations about common patterns of growth in 107; percentage of informal provision 146, 148; population density and 107, 118, 119, 120; postadolescent acquisition 136; and the propensity to migrate 133, 136–8; rates of among German immigrants 96; religious value 127n15; signature evidence 107, 131; sources of evidence 107; sources of evidence for 96; studies of European adult male (1633–1840) 110; three studies of immigrant 112; urban vs rural 107; usual method of quantifying 107 literacy rates: English indentured servants 112; of German immigrants 277 loan defaults 64, 248–50, 259, 261, 278 maintenance costs, percentage of the redemptioner’s yearly labour value 259 malaria 80–1 married persons, likelihood of paying for passage in advance 162 married servants: average contract length 306; average contract price 307; average per-year price 334 Maryland Gazette 228 mass migration era, passage mortality 381 mass migration era German immigration: by age 402–3; British and Irish

comparisons 405, 406, 418; debarkation ports and areas of residence 389–94; dominant region of origin 385; educational perspective 418–19; embarkation ports 385, 386; emigration rates and destination percentages 383; fall in migration costs 399; and the fall in passage mortality 381; family units 403–4; females percentages 404–5; and French cotton imports 385, 387; Friedrich Kapp’s observations 379; human capital transfer 406; influencing factors 377–8; integration and acculturation comparisons 419n; male percentages 400–2; occupational distributions 406–7, 409–11, 414; occupational distributions (1850–1910) 412, 415; occupational distributions (adult males 1820–1920) 408; and the overcoming of the poverty trap 378; and passage fares 379–80; percentage distributions 383, 384; political perspective 418–19; regional origins 387, 388, 389; and the steamship 381–2; transition from founding era to 399 masters, see buyers Melchior, Leonard 292 Menard, R.R. 79 Mennonites 8 methods of acquiring transatlantic passage, and division of risk 220 middle states, concentration of GermanAmericans in 2–3 migration, economic theory of 24, 28 miners, influx of German 105n19 minority faiths 8 Mittelberger, Gottlieb 35–8, 72–3, 116, 164, 258, 278, 280, 294 monopolization: allegations against the Hope family 56–7; allegations against the Stedman family 53, 56; literary evidence for 53–61; quantitative evidence against 61–6 Moravians 8 Morgan, E.S. 79 Mühlenberg, Frederick 29, 136, 262, 355 Mühlenberg, Henry M. 74, 89, 116, 228, 279, 325 Naas, Johannes 278, 280 Napoleonic Wars 19, 33, 317, 343, 359–60, 385, 389 native-born literacy, see colonial literacy New England, educational decline 116

Index   429 Nova Scotia 102 oath lists, Strassburger’s collection 17 oath of abjuration 17, 131 oath of allegiance 289 oaths 5, 17–18, 23, 34, 46, 78, 131, 290 occupational distributions: among buyers (1787–1831) 331; and changing economic opportunities 101; craftsmen and professionals 101; English/German comparison 98; farmers 100; low-skilled occupations 100; retail trades 100–2; servitude 102; skilled occupations 100–1, 197; southern Rhineland 101 occupational training: and contract length 268; contract specifications 145; and indigent/immigrant children’s price differences 295; pauper apprentices/ German immigrant children comparison 295; percentage provision (1745–73) 246; percentage provision (1771–3) 264; percentage provision (1787–1804) 266 ocean passage, price of 159 organized immigrant servant markets: beginning and end 316; disappearance, see disappearance of organized immigrant servant markets; last systematic body of records documenting 303 origins of German overseas emigrants: most common 21; by region (1830–1910) 388; by region (1860 and 1870 censuses) 389; by region (port of departure 1798–1808) 22; by village 22 overseers of the poor, Philadelphia 291–5 Palatinate 21, 32, 135, 138, 389 passage debts: shifting of, see debtshifting; typical amounts 286 passage fares, and contract prices (1772–1835) 350 passage mortality: and average voyage times 74; burials 80; children 82; decline 74; economic perspective 74–5; German immigration to Pennsylvania (1727–1805) 74; mass migration era 381; monthly rates 74; rates of 72; reducing the risk of 250; shippers’ precautions against 74–5; slave trade comparison 71–4, 82; sources of evidence 72; Strassburger ship lists sample 73, 74 passage prices, credit vs cash 248, 308 passenger contracts, an example 308, 309, 310

passenger-per-ton ratios 40–3, 250, 251 passport regulations, Dutch 57–8 Peace of Aix-la-Chapelle 32 Peace of Amiens 343 Penn, William 4, 34, 37 Pennsylvania: adult male literacy among German immigrants to 97; age distribution of adult German and English immigrants to (1730–1820) 95; colonial zenith of German immigration to 198; distribution of servant contracting by family status (1785–1804) 161; education and age of German immigrant servants in (1787–1804) 151; education specified in servant contracts for German immigrant children in (1771–1804) 147; English alarm at the proliferation of Germans in 5–6; failure to maintain literacy levels 116; family composition of German and English immigrants to (1727–1820) 92; founding 4; German-Americans concentrated in 2; German immigration to (1727–75) 18; German immigration to (1772–1835) 344; German immigration to (1772–5) 18; German percentage of the population 17; German preference for 59; impact of German immigration on the population of 4; incidence of servitude, see incidence of servitude in Pennsylvania; incidence of servitude among German immigrant children to (1785–1804) 165; inspection regulations 79; occupational distribution of adult male German and English immigrants (1709–1820) 98; percentage of population with German ancestry 89; percentage of servants among German immigrants to (1772–1835) 347; poor relief statute 291; quantitative evidence suggesting competition in the market 61–6; recruitment success 59; register of German immigrant servant contracts 7, 303; role of ship captains in carrying German passengers to 61; servitude statistics, see incidence of servitude in Pennsylvania; social composition, of German and English immigrants to (1709–1820) 91, 93; State laws regulating the immigration and sale of German immigrant servants 305–6; variance of German immigration to (1727–74) 30

430   Index Pennsylvania Brethren 57 Pennsylvania Gazette 152, 184, 228, 257 persecution, religious 24 Philadelphia: auction of British and German redemptioner immigrants (1771–3) 264; auction of German redemptioner immigrants (1787–1804) 266; average voyage time from Holland to 74; causes of mortality due to infectious disease (1747–75) 83; consignment of German redemptioner cargos in (1751–74) 63; crowding (passengers-per-ton) (1727–74) 40; embarkation port distributions (1727–75) 23; embarkation port distributions (1785–1804) 24; eyewitness descriptions of sale of German immigrant redemptioner servants in 311, 313; families on the Britannia (1773) 282, 284; fares for adult male Germans from Holland to (1708–1819) 65; female servants’ age and marital status (1771–1831) 191; flow of German immigrants through the port of (1727–1835) 28; German immigrants landing in (1809–16) 19; German percentage of total arrivals (1812–23) 20; illiteracy of German immigrants to (1727–75) 132; illiteracy of German immigrants to (1730–54) 135, 136; immigrant servant contracts market 160; immigrant versus resident mortality (1738–62) 76; immigrants per ship from Holland to (1727–74) 66; the market for German immigrant servants (1785–1831) 318; mortality rates 82; overall incidence of servitude among German immigrants arriving in 160; overseers of the poor 291–5; per capita distribution of new servants 201; percentage of German arrivals selling themselves into servitude 316; port of last clearance for German immigrant vessels arriving in (1727–75) 36; prominence as a port of entry 5, 18, 20–1, 203; ratio of slaves to servants 202; redemptioner servant compensation for single adult male immigrants (1771–1804) 260; runaway servants (1785–1820) 324; seasonal distribution of German immigrants arriving in (1727–1804) 34; servant auction records surviving 9; servant contracts 123; sexual composition of German and Irish

servant immigration to (1729–75) 188; shipping records 62; smallpox epidemics 81–2; Stedman family’s activities (1731–54) 54; total German immigration through the port of (1785–1804) 19; wage margins 25 Philadelphia Custom House Records 9, 19–20 Philadelphia immigrant servant sales records: composition 182–4; demographic uses 184–7; as legal protection 182; problems with 182–4; sexual composition 187–92; sexual composition of immigrant servants by nationality (1745–1831) 185; surviving documents 182; usefulness 181 Philadelphia market: distribution of contract characteristics by contract type (1745–73) 246; geographic distribution of British servant purchasers by county (1771–3) 204; geographic distribution of British servant purchasers by town of residence (1771–3) 209; geographic distribution of German servant purchasers by county (1771–3) 206; geographic distribution of German servant purchasers by county (1787–1804) 207; geographic distribution of German servant purchasers by town (1771–3) 210; geographic distribution of German servant purchasers by town (1787–1804) 211; geographic distribution of immigrant servant purchasers by county (1745–1804) 199; geographic distribution of Irish servant purchasers by county (1745–6) 202; indentured versus redemptioner servitude 245; occupational distribution of German servant purchasers (1787–1804) 214, 331; record-keeping 7; supply and demand for German servants in the (1772–1821) 351 pietist sects 8 pooling debts 249–50, 321 poor-relief: Pennsylvania’s statute 291; privatization of children’s 291 population density, and literacy 107, 118, 119, 120 population distribution, by nationality and state (1790 census) 3 ports of origin, see embarkation ports post-voyage mortality: decline 381; epidemiological perspective 80, 81–2;

Index   431 rates of 79–82, 248, 381; relationship between debarkation morbidity and 81; see also seasoning mortality printing houses, German 132 productivity: and contract length 268; regional perspective 270 Protestant populations, rising intolerance towards 32 quantitative research, pitfalls of 9 Queen Anne’s War 90 recruitment network, VOC’s 230–1 redemption: the auction 262; average rates of return (1802–19) 64, 249; and the avoidance of servitude 256; basic market model of the trade 232, 236, 240; bearer of risk 221, 245; benefits 229–30, 245–6; competition and average servant compensation 257–61; contract examples 222–3, 248; and family debt consolidation 237–8; first known reference 256; fixed cost elements 237; flexibility 221, 229–30, 246, 278; Germans’ use of 228–31; vs indenture 220–1, 244–7, 255; and the “lemons problem” 231–2, 238; origins 229, 231; and pre-embarkation debt 230; productivity and contract length 239, 268; risk and the rate of return 247–51; and search restrictions 256–7, 269; system analysis 59–60, 159, 245, 255, 258, 342–3; and the VOC 231 redemption auction: bargaining process 262; British and German immigrants 264; comparison with the auction of slaves or indentured servants 269; German immigrants 266; modelling the 262–3; relationship between contract length and employer occupation 269–70; shipboard 269 redemptioner servant compensation, for single adult male immigrants 260 registration, net benefits of 343 registry of German immigrant servant contracts, Pennsylvania government’s establishment of 7 remittance system 358–60, 380 resource endowments 26 Rhineland, climate crisis and famine (“year without a summer”) 232, 311–12, 343, 354, 357 risk: bearer of under indenture and redemption 220–1, 244; pooling 249

Rotterdam to Philadelphia market 58–60, 64, 66, 235, 256, 282, 284 Ruecastle, Robert 62 runaways 184, 322–5 Sahlin, Anna Margareta 223, 225 sale of children into servitude: evidence against 280–8; evidence for 278–80; explanations for 276, 293; selling price 288 Salinger, S.V. 187, 202 Sauer, Christopher: on debt-shifting 280; impact of letters from 27–8; on indenture and redemption 229, 256, 257; on monopolization 53, 56; occupations 97; on Pennsylvania wages 25 Scheldt, conflict over the navigation of the 23 Schofield, R.S. 132 schooling: cost of on the frontier 116; formal, see formal schooling; masters’ obligations 305 Schultz, Christopher 56 seasickness 75, 82 seasonal pattern for German immigrant arrivals 34–5, 38 seasoning mortality 35, 79–80, 82, 84 Second German Reich 376–7 secular rise in servant productivity 258 servant buyers: distribution of servants by number purchased (1817–31) 329; geographic distributions 199, 202, 204, 206, 207, 209, 210, 211, 327; occupational distributions 214, 331; social and regional perspectives 307 servant contract choice: indenture vs redemption 244–7; insurance vs flexibility 244–7 servant contract education evidence: contextual perspective 145–6; effect of age and gender on educational choice 150–2; formal schooling 146, 148; geographic distribution of educational methods 148–9; and occupational training clauses 145; Pennsylvania 147, 151; percentages 123, 146; sequence of instruction 146; types of education contracted for 146–8 servant market, single largest in North America 160 servant markets, disappearance of, see disappearance of organized immigrant servant markets servants, ratio of slaves to 202

432   Index servitude: disappearance of immigrant, see disappearance of immigrant servitude; and family status 161; impact on geographic distribution and social integration of immigrants 316; percentage of German emigrants arriving in Philadelphia selling themselves into 316; rising urbanization of 203 servitude incidence: across colonies 170–4; across national groups 166–70; and age 162; agricultural perspective 173; among German immigrant children 165; British immigrants 162, 164, 168–9, 171, 172; Caribbean colonies 171, 173–4; Carolinas 171, 173; Chesapeake region 112, 173; contextual perspective 159–60; dependent children 163–6; determinants 160–3; and distance from port of departure 169–70; evidence assessments 159–60, 168, 171, 181; and family size 162, 163, 164; and gender 161; Irish immigrants 168–9; and marital status 161–2; middle colonies 171, 173–4; northern colonies 174; pre- and post-Revolution 169, 171; Scots immigrants 169; sensitivity of to changing wealth 162; southern colonies 174; see also incidence of servitude in Pennsylvania Seven Years War 29, 32, 52, 64, 90, 184, 191, 343 sexual composition of immigrant servitude 161, 187, 191–2 ship captains, role of in carrying German passengers to Pennsylvania 61 ship names, duplication of 69n30 shipboard mortality: nineteenth century decline 381; and time at sea 381; see also passage mortality shipping services, adequacy of 38–9 shipwrecks 86n11 Shoemaker, Benjamin 62 signature literacy 107, 115, 122, 124, 132–3, 138, 146, 152 single adult male servants, average peryear price 334 single adult males, average contract length 306 skills, indenture and the verification of 221 slave labour 173, 197–8, 201, 203, 342 smallpox 74, 78, 81–2, 84 Smith, Abbot E. 56, 58, 159, 247 Smith, Adam 25 Smith, B.G. 72, 81

Smith, Charles 62 social composition, of German and English immigrants to Pennsylvania 91, 93 social composition of German immigrants: age distributions 95–6; causes of change in 93; family composition, see family composition; and literacy 96–7; occupational distribution, see occupational distributions; sources of evidence 89, 101 social control, literacy and 119, 121 social structure, and literacy 137 Spanish Succession, War of the 4 St. Andrew 56 steamships, introduction of 381–2 Stedman, John and Charles 53, 54, 56, 57, 58, 59, 62 “Strangers’ Ground” 80 Strassburger, R.B. 17, 19 summer arrival, disease effect of 34–5 Switzerland 20–2, 278 Tambora, eruption 33 tax records 201–2 “The Best Poor Man’s Country” 171 Theobald, Charles 138 Thirty Years War 8 Thomas, Governor 4, 17 Trachsler, Hans 58 transatlantic migration, cost of 159 transatlantic passage, methods of acquiring 7, 220 transportation risks, bearer of under indenture and redemption 220–1, 244 Trevor, Robert 56–7 Tully, A. 114 typhoid 75, 78 typhus 75, 78, 381 universal literacy, achievement of 144 VOC (Dutch East India Company) 8, 230–1 voyage mortality, see passage mortality; shipboard mortality voyage time, for German immigrant vessels 35–7 wage comparisons 25–6, 29, 378 Walsh, L.S. 79 War of 1812 19, 33, 37, 192, 303, 316, 319, 326, 346, 353, 357, 389 War of the Austrian Succession 29, 32, 60, 64, 184, 343

Index   433 War of the First Coalition 33 War of the Polish Succession 32 War of the Second Coalition 33 War of the Spanish Succession 4 warfare: impact on migration 32–4; in North America 29 wars: American Civil 377, 406; Crimean 377; Franco-Prussian 377; Napoleonic 19, 33, 317, 343, 359–60, 385, 389; Queen Anne’s 90; Seven Years 29, 32, 52, 64, 90, 184, 191, 343; Thirty Years 8; World War I 376–7, 384; see also American Revolution wealth of American colonists, per capita lifetime accumulated 26 West Indies, disease environment 73

Whitehead, John Frederick 8, 280, 294 will-signing, as a measure of literacy 108, 109, 115, 132; see also signature literacy Willcox, W.F. 404 Willing & Morris 62 Wokeck, M.S. 17–18, 293 Wolters, D. 56 World War I 376–7, 384 Wright, L.B. 131 Wurttemberg 21 “year without a summer” 232, 311–12, 343, 354, 357 yellow fever 80–2

Taylor & Francis

eBooks FOR

liBRARIES

Over 23,000 eBook titles in the Humanities, Social Sciences, STM and Law from some of the world's leading imprints. Choose from a range of subject packages or create your own! ~ Free MARC records ~ COUNTER-compliant usage statistics ~ Flexible purchase and pricing options

... Off-site, anytime access via Athens or referring URL ... Print or copy pages or chapters ... Full content search ~ Bookmark, highlight and annotate text

... Access to thousands of pages of quality research at the click of a button For more information, pricing enquiries or to order a free trial, contact your local online sales team. UK and Rest of World: [email protected]

US, Canada and Latin America: [email protected]

www.ebooksubscriptions.com

o

~

Taylor & Francis ~:r.m~ nylof&Frand,G""'p