Aging and Generational Relations Over the Life Course: A Historical and Cross-cultural Perspective 3110138751, 9783110138757

548 20 13MB

English Pages 533 [545] Year 1996

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Aging and Generational Relations Over the Life Course: A Historical and Cross-cultural Perspective
 3110138751, 9783110138757

Table of contents :
Introduction: Aging and Generational Relations Over the Life Course
Kinship and Care for the Aged in Traditional Rural Iberia
Household Patterns of the Elderly and the Proximity of Children in a Nineteenth Century City; Verviers, Belgium, 1831–1846
“The Life Stairs”: Aging, Generational Relations, and Small Commodity Production in Central Europe
Aging in a Never-Empty Nest: The Elasticity of the Stem Family
The Family, State Support and Generational Relations in Rural Ireland at the Turn of the Twentieth Century
Relations Between Older Adults and Their Adult Children in a Nineteenth-Century Italian Town
Retirement, Inheritance, and Generational Relations: Life-Course Analysis in Historic Eastern Europe
Gender, Rural-Urban and Socio-Economic Differences in Coresidence of the Elderly with Adult Children: The Case of Sweden 1860–1940
Asymmetry in Intergenerational Family Relationships in Italy
Equity Between Generations in Aging Societies: The Problem of Assessing Public Policies
Residence and Family Support Systems for Widows In Nineteenth- and Early Twentieth-Century Texas
Living Arrangements of the Elderly in America: 1880–1980
The Generation in the Middle: Cohort Comparisons in Assistance to Aging Parents in an American Community
Fathers and Sons in Rural America: Occupational Choice and Intergenerational Ties Across the Life Course
The Well-Being of Aging Americans With Very Old Parents
Exchanges Within Black American Three-Generation Families: The Family Environment Context Model
The Demography of Family Care for the Elderly
Types of Supports for the Aged and Their Providers in Taiwan
Familial Support and the Life Course of Thai Elderly and Their Children
Intergenerational Support In Sri Lanka: The Elderly and Their Children
Some Inter- and Intracohort Comparisons of Generational Interactions. From the Acquisition of Parental Roles through Postparenthood in Japan, 1914–1958
Generational Relations and Their Changes As They Affect the Status of Older People in Japan
Generational Relations: A Future Perspective

Citation preview

Aging and Generational Relations Over The Life Course

Aging and Generational Relations Over The Life Course A Historical and Cross-Cultural Perspective Edited by Tamara Κ. Hareven University of Delaware Family Research Symposium Essays

w DE

G

Walter de Gruyter · Berlin · New York 1996

Tamara Κ. Hareven, Unidel Professor of Family Studies and History, Departments of Individual and Family Studies, and History, University of Delaware, Newark, Delaware, USA

© Printed on acid-free paper which falls within the guidelines of the ANSI to ensure permanence and durability. Library of Congress Cataloging-in-Publication-Data Aging and generational relations over the life course : a historical and cross-cultural perspective / edited by Tamara Κ. Hareven. Includes bibliographical references. ISBN 3-11-013875-1 (acid-free paper) 1. Aged-Social conditions-Cross-cultural studies. 2. Aging-History—Cross-cultural studies. 3. Intergenerational relations—History—Cross-Cultural studies. I. Hareven, Tamara Κ. HQ1061.A425 1996 305.26—dc20 95-45282 CIP

Die Deutsche Bibliothek - Cataloging-in-Publication Data Aging and generational relations over the life course : a historical and cross cultural perspective ; University of Delaware Family Research Symposium essays / ed. by Tamara Κ. Hareven. - Berlin ; New York : de Gruyter, 1995 ISBN 3-11-013875-1 NE: Hareven, Tamara K. [Hrsg.]: Family Research Symposium ; University of Delaware

© Copyright 1995 by Walter de Gruyter & Co., D-10785 Berlin All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or any information storage and retrieval system, without permission in writing from the publisher. Printing: WB-Druck, Rieden. - Binding: D. Mikolai, Berlin. - Cover Design: Johannes Rother, Berlin. Printed in Germany.

Contents Introduction: Aging and Generational Relations Over the Life Course Tamara Κ Hareven

1

Kinship and Care for the Aged in Traditional Rural Iberia Stanley Brandes

13

Household Patterns of the Elderly and the Proximity of Children in a Nineteenth Century City; Verviers, Belgium, 1831-1846 George Alter, Lisa Cliggett, Alex Urbiel

30

"The Life Stairs": Aging, Generational Relations, and Small Commodity Production in Central Europe Josef Ehmer

53

Aging in a Never-Empty Nest: The Elasticity of the Stem Family Antoinette Fauve-Chamoux

75

The Family, State Support and Generational Relations in Rural Ireland at the Turn of the Twentieth Century Timothy W. Guinnane 100 Relations Between Older Adults and Their Adult Children in a Nineteenth-Century Italian Town David I. Kertzer and Dennis P. Hogan 120 Retirement, Inheritance, and Generational Relations: Life-Course Analysis in Historic Eastern Europe Andrejs Plakans 140 Gender, Rural-Urban and Socio-Economic Differences in Coresidence of the Elderly with Adult Children: The Case of Sweden 1860-1940 Lars-Göran Tedebrand 158 Asymmetry in Intergenerational Family Relationships in Italy Marzio Barbagli

191

Equity Between Generations in Aging Societies: The Problem of Assessing Public Policies Anne-Marie Guillemard 208

VI

Contents

Residence and Family Support Systems for Widows In Nineteenth- and Early Twentieth-Century Texas Jane Zachritz and Myron P. Gutmann 225 Living Arrangements of the Elderly in America: 1880-1980 Steven Ruggles

254

The Generation in the Middle: Cohort Comparisons in Assistance to Aging Parents in an American Community Tamara Κ. Hareven and Kathleen J. Adams 272 Fathers and Sons in Rural America: Occupational Choice and Intergenerational Ties Across the Life Course Glen H. Elder, Jr., Elizabeth B. Robertson and Rand D. Conger 294 The Well-Being of Aging Americans With Very Old Parents Dennis R Hogan, David J. Eggebeen and Sean M. Snaith

327

Exchanges Within Black American Three-Generation Families: The Family Environment Context Model James S. Jackson, Rukmalie Jayakody and Toni C. Antonucci 347 The Demography of Family Care for the Elderly Douglas A. Wolf, Beth J. Soldo and Vicki Freedman

378

Types of Supports for the Aged and Their Providers in Taiwan Albert I. Hermalin, Mary Beth Ofstedal and Ming-Cheng Chang

400

Familial Support and the Life Course of Thai Elderly and Their Children John Knodel, Napaporn Chayovan and Siriwan Siriboon

438

Intergenerational Support In Sri Lanka: The Elderly and Their Children Peter Uhlenberg

462

Some Inter- and Intracohort Comparisons of Generational Interactions. From the Acquisition of Parental Roles through Postparenthood in Japan, 1914-1958 Kanji Masaoka and Sumiko Fujimi 483 Generational Relations and Their Changes As They Affect the Status of Older People in Japan Kiyomi Morioka 511 Generational Relations: A Future Perspective Matilda White Riley and John W. Riley, Jr.

526

Acknowledgements This volume resulted from the International Conference on Aging and Generational Relations Over the Life Course which I organized and directed at the University of Delaware, October, 1992, with a grant from the National Institute on Aging (NIA). I would like to express my gratitude both to the NIA for their generous support and to the University of Delaware for hosting the conference. Special appreciation is extended to Dr. Marion Hyson, who was Acting Chair of the Department of Individual and Family Studies in 1992, and to Dr. John Cavanaugh, who subsequently, as Chair of the Department, provided various supports for the editing of the volume. I am grateful to Judy Wilson for the outstanding organizational support she provided for the conference, and to Martha Dimes Toher for copy editing. Since we were responsible for producing camera-ready copy of the manuscripts, the formidable task of editing and formatting was carried out by Linda Granger, Gladys Llewellen, and especially by Catherine Raphael, who successfully completed the final stages. I would also like to thank the contributors for their patience with the long editing process, and to Andrejs Plakans, Bianca Ralle at Walter de Gruyter and Richard Koffler at Aldine de Gruyter, for their continuing support and encouragement.

Tamara Κ. Hareven Newark, Delaware

Introduction: Aging and Generational Relations Over the Life Course Tamara Κ. Hareven University of Delaware

This volume addresses one of the crucial social issues of our times: the supports that older people receive from their adult children and from other kin. As the concentration of elderly people in the population increases and as life continues to be extended, the problems of generational assistance in the later years of life will become an even more central issue. It is not surprising, therefore, that the relations between the generations recently have become an important topic in social science investigations, prompted particularly by the increasing interest in the problems of an aging society. This interest in generational relations has been heightened further over the past two decades by an intense concern among social scientists and the public with the family. Since the family is a major arena in which generational relations are acted out, one needs to examine these issues within familial, as well as social-structural and institutional contexts. The study of generational relations thus requires a multidimensional approach: externally, an examination of the interaction of individuals and families with social institutions and with the historical processes affecting them; and internally, an examination of these relations within the family and the wider kin group. The essays in this volume help redress some of the misconceptions and myths that have clouded our understanding of generational supports and provide a picture that is more accurate historically and cross-culturally.

Misconceptions and Myths About Generational Relations A main misconception about generational relations in later life involves who actually provides supports for the elderly and what form these supports take. On the one hand, family members assume that the public sector carries major responsibilities of care for the aged; on the other hand, the state assumes that the family is responsible for some of these areas. This confusion in the assignment of responsibilities often means that elderly people land between the family and the welfare state without receiving proper supports from either. A major myth about the past is that a golden age once existed during which members of the older generation coresided with their children and other kin, and elderly people were

2

T. Hareven

secure in receiving supports from their family members. The myth further assumes that industrialization and urbanization eroded kinship ties and diminished generational supports. Historical research over the past three decades, including the findings presented in the essays in this volume, has helped challenge these myths. In reality, the dominant form of household structure in Western Europe and in the United States in the preindustrial period, as well as following industrialization, was nuclear: coresidence of three generations was not the dominant pattern in Western Europe and in the United States. Three generations rarely lived together in the same household, although elderly parents often resided in a separate household nearby and were engaged in a variety of mutual supports with their adult children. Coresidence was more prevalent, however, in later life, primarily when elderly parents were too frail to maintain a separate residence (Laslett, 1977; Hareven, 1991b; Chudacoff & Hareven, 1978; Smith, 1981.) Contrary to prevailing myths, elderly people in Western society have never experienced a golden age. Both in preindustrial Europe and in colonial America, aging parents entered into contracts with inheriting sons in order to secure basic supports in old age. In his chapter on Iberia, Stanley Brandes emphasizes the sense of insecurity that prompted such arrangements. Generational assistance and relations with wider kin were based on voluntary reciprocity, not on rigidly enforced customs or laws. In order to understand how these patterns changed over time, it is necessary to take into account the cultural values governing reciprocal relations among kin in different time periods and among various groups within the same culture. A full understanding of generational relations in the later years of life has also been handicapped by researchers' reliance on surveys and census schedules that are limited to one point in time. On the other hand, the life-course perspective that informs the essays in this volume, provides a dynamic approach to understanding earlier life antecedents of generational supports. Recent research on the life course has contributed a historical perspective on the phenomenological, social-structural, and behavioral aspects of intergenerational relations. As some of these essays show, relationships of mutual support among the generations were formed early in life and were carried into old age after being modified by various events over the life course.

A Life-Course Perspective A life course perspective provides both a developmental and historical framework for the study of intergenerational relations. It enables us to understand how patterns of assistance and support networks were formed over the life course and were carried over into the later years; how they were shaped by historical circumstances and by people's cultural traditions; and what strategies individuals and families followed in order to secure future supports for later life. Relations of mutual support are formed over the life course and are reshaped by historical circumstances, such as migration, wars, and the decline or collapse of local economies.

Introduction: Aging and Generational Relations Over the Life Course

3

In its emphasis on interaction with historical time, the life-course approach provides an understanding of the location of various cohorts in their respective historical contexts. Specifically, it enables scholars and policy makers to examine the historical circumstances that have affected the lives of the members of different cohorts, and that help explain the differences in the historical experiences shaping their respective life histories. Patterns of providing support and expectations for receiving support in old age are part of a continuing interaction among parents, children, and other kin over their lives as they move through historical time. Relations of mutual support are formed over life and are reshaped by historical circumstances, such as migration, wars, and the decline or collapse of local economies. Rather than viewing old people as a homogeneous group, a life-course perspective illuminates how their problems, needs, and patterns of adaptation were shaped by their earlier life experiences and by the historical conditions affecting them. As Tamara Hareven and Kathleen Adams point out in their chapter, earlier life-course experience affected by historical circumstances guided the preferences of members of different cohorts for the types of sources of support for elderly parents and for their involvement in parental care. Underlying the life-course approach are three major dimensions, all of which revolve around timing: Individual timing of life transitions in relation to external historical events; the synchronization of individual life transitions with collective familial ones; and the impact of earlier life events, as shaped by historical circumstances on subsequent ones. The timing of transitions involves the balancing and timing of an individual's entry into and exit from different family work and community roles (education, family, work, and community) over the life course. It addresses the question: How did people time and sequence their work life and educational transitions in the context of changing historical conditions? In all these areas, the pace and definition of "timing" hinge upon the social and cultural contexts in which transitions occurred. While age is an important determinant of the timing of transitions, it is not the only significant variable; changes in family status and in accompanying roles are often as important as age (Hareven, 1991; Hareven & Masaoka, 1988). The synchronization of individual life transitions with collective family transitions involve most notably the juggling of multiple family- and work-related roles over the life course. Individuals engage in a variety of familial configurations that change over the life course and vary under different historical conditions. The task of synchronizing individual transitions with familial ones can generate tensions and conflicts, especially when individual goals are at odds with the needs and dictates of the family as a collective unit. As evidenced in several of the chapters, the timing of adult children's individual transitions often conflicted with the demands and needs of aging parents. For example, Hareven and Adams found that aging parents discouraged at least one caretaking daughter from leaving the parental home and marrying in order to assume continued support in old age. Dennis Hogan and David Eggebeen demonstrate the interlocking of generational transitions in the later years of life: The death of an aging parent enabled caretaking children who were themselves "old" to begin providing for their own old age and for their adult children or grandchildren.

4

T. Hareven

Another key feature of the life course approach is the cumulative impact of earlier life events on subsequent ones. The "early" or "delayed" timing of certain transitions affects the pace of later ones. Events experienced earlier in life may continue to influence an individual's or a family's life path in different forms throughout their lives. For example, Glen Elder has documented the negative impact the Great Depression had on the cohort of young men and women who encountered it in their transition to adulthood. Delayed timing in education or early commencement of the work life also affected subsequent delays and disorderliness in the careers of the Depression cohort (Elder, 1974). Historical forces thus play a crucial role in this complex, cumulative pattern of individual and familial life trajectories. Such forces have a direct impact on the life course of individuals at the time when they encounter them, and continue to affect them indirectly through life. Thus, the social experiences of each cohort are shaped not only by the historical events and conditions encountered at a certain point in life, but also by the historical processes that shaped earlier life transitions. The impact of historical forces on the life course may continue over several generations. One generation transmits to the next the effect of the historical circumstances that shaped its life history. Elder and Hareven (1992) have found that for the same age cohorts in two different communities, delays or irregularities in the parents' timing of their work and family careers, resulting from the Great Depression, affected their children's timing of life transitions. The children thus experienced the impact of historical events on two levels — directly, through their own encounter with these events, and indirectly, in the ripple effects of these events across the generations. A life-course perspective provides a framework for understanding variability in the patterns of support in later life, as well as differences in the expectations of the recipients and the caregivers who are influenced by their respective social and cultural milieux. Patterns of generational assistance are shaped by values and experiences that evolve or are modified over the entire life course. For example, in the United States, ethnic values of premigration culture call for a more exclusive dependence on filial and kin assistance than do more contemporary attitudes, which advocate reliance on supports available from government programs and community agencies. Such differences in values are expressed in the caregiving practices and attitudes of successive cohorts. The earlier life-course experiences of each cohort, as shaped by historical events, also affect the availability of resources for their members and their modes of assistance and coping abilities in later life. The life-course approach helps clarify the distinction between "generation" and "cohort," which have been frequently confused in the gerontological literature and provides insight into the interrelationship of these two concepts. "Generation" designates kin relationships (for example, parents and children or grandparents and grandchildren); it may encompass an age span, often as wide as thirty years. A "cohort" consists of a more specific age group that has shared a common historical experience. Most important, a cohort is defined by its interaction with the historical events that affect the subsequent life course developments of that group. A generation may consist of several cohorts, each of which has encountered different historical experiences that have affected its life course. The chapter by Kanji Masaoka and

Introduction: Aging and Generational Relations Over the Life Course

5

Sumiko Fujimi demonstrates the interrelationship between cohort and generation. By comparing generational role transitions and mutual supports among five cohorts in Tokyo and Fukushima, they identify changes over time as well as community differences within each generation. In Hareven and Adams's comparison of patterns of assistance of two cohorts of adult children to aging parents in a New England community, the distinction between cohort and generation emerges with particular clarity: In families with large numbers of children, siblings in the same family belonged to two different cohorts, with different historical experiences and attitudes toward generational assistance. As Anne-Marie Guillemard points out, several studies have confused generations and age groups when assessing intergenerational equity. She argues that measuring intergenerational equity requires longitudinal analysis that follows contributions from successive generations over the entire life course. The essays in this volume revolve around the following interrelated themes: interaction of early life transitions with later ones and their impact on old age; coresidence among generations; life course antecedents in shaping intergenerational supports in the later years of life; assistance to aging parents from children and other kin who do not coreside; the institutional context affecting generational assistance and reciprocities in the later years of life; and, the impact of social change on generational relations.

Interaction of Early Life Transitions With Later Ones and Their Impact On Old Age Generational relations in later life are interconnected with experiences and transitions from earlier life. Thus, we view "old age" as part of an overall process of generational interaction in the context of changing historical events, rather than as an isolated stage or experience. Various stages and transitions in the life course are interlocked across generations. Within a familial setting, the life transitions of the younger generation are intertwined with those of the older generation. For example, the timing of leaving home and marriage in early adulthood is interrelated with the timing of the older generation's transitions into retirement or with inheritance. This is precisely where generations are interdependent. The strategies that parents and children followed in determining exchanges and transfers in their reciprocal interactions over the life course represent, therefore, an important theoretical and empirical theme. For example, how did parents control their children's timing of leaving home and marriage? As Antoinette Fauve-Chamoux points out in her chapter, for residents of the villages in the French Pyrenees, inheritance determined the timing of the heir's succession. But, parental strategies took different forms for the mate selection and the timing of each child's marriage, depending on whether it was the inheriting son or a noninheriting son or daughter who was marrying. Similarly, in their chapter, David Kertzer and Dennis Hogan emphasize the significance of inheritance in regulating coresidence and generational supports in an Italian sharecropping village. They find significant differences between wage earners and sharecroppers in the use of

6

T. Hare ven

inheritance in exchange for supports in old age and their impact on the familial and economic status of women.

Coresidence The household arrangements of older people have been a continuing theme in historical research on the family, as well as in the essays included here. As a large body of scholarship has demonstrated, nuclear household patterns have predominated since the preindustrial period in the United States and western Europe. The modal pattern has been one where the younger generation establishes a separate household. Except for some variation over the life course, elderly couples and aging widows have attempted to maintain separate households for as long as possible. The prevalent residential arrangement of older people and their children has been one of "intimacy from a distance." This pattern was modified, however, in the later years of life. Coresidence with a child, single or married, was the major solution if aging parents were unable to live separately (Laslett, 1977; Laslett & Wall, 1972; Hareven, 1971). As the chapter by George Alter, Lisa Cliggett and Alex Urbiel shows, in nineteenth-century Belgium, coresidence with adult children tended to increase in the later years of life. Similarly, Timothy Guinnane has found high rates of coresidence of elderly parents with adult children in Ireland. But whenever possible, parents preferred to coreside with an unmarried child. Brandes has found in northern Spain and Portugal a similar preference among elderly parents for the autonomy of their households. Coresidence occurred, however, when the parents — especially an aged widow or widower — were too frail to live alone. Under such circumstances, the elderly parent resided in the households of various children through a "rotation" system. In regions with impartible inheritance, the inheriting son coresided with the aging parents and was expected contractually to be their main supporter. Whenever possible, older widows continued to lead their own households by taking related as well as unrelated individuals, such as boarders and lodgers into the household. Among several ethnic groups in the United States, parents assured themselves in advance of having an unmarried child remain at home, as a security for support in later life. It was a common practice to discourage the youngest daughter from leaving home and getting married, so that she would be able to support her aging parents. Jane Zachritz and Myron Gutmann found that in nineteenth-century Texas widowhood in later life was a main determinant in leading elderly women to coreside with their children. As would be expected, in the Asian countries discussed in this book there was a higher rate of coresidence of elderly parents with adult children than in the West. Coresidence with at least one adult child has been the customary pattern for aged parents in Taiwan, Thailand, and Japan — a pattern typical for East Asia. John Knodel and his collaborators found that in Thailand, the majority of men and women over 65 coreside with one child. There are, however, considerable variations over the life course. Older children move out and set up their own households, but typically one unmarried child coresides with aging

Introduction: Aging and Generational Relations Over the Life Course

7

parents, especially when they begin to need assistance. This practice is more pronounced for aged widows, for whom coresidence with a child is the major form of support. Even within Asia, however, significant variations occur: Albert Hermalin and his collaborators found in Taiwan significantly lower rates of coresidence among the mainland Chinese than among the Taiwanese. In Thailand, older children are less likely than younger children to coreside with their parents at the very point in their parents' lives when support from a coresident child was most needed. The needs of elderly parents, especially widowed mothers, are met through coresidence with a younger child. Several of the authors found that in East Asia supports from coresiding children are more secure than those from children who live separately, even if the latter are in proximity. As Peter Uhlenberg points out, in Sri Lanka coresidence seems the only guarantee that elderly parents will receive support services from adult children. Supports from noncoresident children to aging parents primarily take the form of financial assistance — a pattern found also in Thailand. In Japan, the rates of coresidence are lower than in the other Asian countries discussed here: As the chapter by Kiyomi Morioka shows, among people older than 65, about 60% coreside with a child. Morioka anticipates, however, that this type of generational coresidence will continue to decline in Japan. Both the United States and Japan have seen a decline in the rates of coresidence of aging parents with adult children. Steven Ruggles argues that in the United States, the decline since the turn of the century in the coresidence of people older than 65 with kin may be as significant a change as the earlier major demographic transitions. Morioka found that in Japan the proportion of elderly people over 65 coresiding with an adult child has declined steadily between 1960 and 1980. He attributes this change to the disintegration of the traditional ie system following World War II, and to the increasing determination of the younger generation to live separately from their parents. Morioka argues that this change in the traditional patterns of coresidence raises serious questions about the future consistency and continuity of supports for the aged. An examination of coresidence among the generations raises several questions related to household headship and to the nature of generational supports. When a household record in a census lists a parent as being the head of the household and an adult child as residing in the household, who in reality heads the household, and what are the dynamics of flow of assets and assistance within such a household? It is difficult to answer these questions from cross-sectional data; nor can this type of data explain what the dynamics are in this pattern of coresidence. Did the son become the head of the household after his father retired or became too old or frail to support himself and manage the family's affairs? Or did the parents move into the son's household? Under what circumstances did the older generation coreside with adult children or other kin? And under what circumstances did aging parents reside separately and interact with kin in various forms of assistance outside the household? Rates of coresidence recorded in cross-sectional data may reflect a life-course pattern in which elderly parents who do not coreside with their children at the time of a census or survey may do so later when they become more dependent. Morioka's data bear out such a hypothesis, since the rates of coresidence with their children of chronically ill or

8

T. Hareven

terminally ill elderly are higher than those of the overall group of people older than 65. Similarly, Hogan, Eggebeen, and Snaith noted that cross-sectional data may obscure considerable variation in patterns of coresidence over the life course. They found in the National Survey of Families and Households that only seven percent of Americans 55 and older with a surviving parent had the parent living with them at the time of the survey; but by their late fifties, one-quarter of persons had an aging parent living with them at some point.

Life Course Antecedents In Shaping Intergenerational Supports In The Later Years of Life Generational supports in old age are part of a life course continuum of reciprocal relations between the generations. Morioka identifies specific phases in generational exchanges over the life course. Early in the life course, generational assistance is extended from parents to their children; in the later phases, the children, especially the inheriting child in the stem family, care for aging parents. Analyzing the role transitions of parents and children over the life course of five cohorts, Masaoka and Fujimi document considerable variations among cohorts as well as community differences within the general patterns discussed by Morioka. How was the caretaking child selected? How was the caretaking relationship formed? And what kind of negotiations and bargaining were involved? As seen from the historical essays in this book, especially those by Kertzer and Hogan, Ehmer, Fauve-Chamoux, and Brandes, inheritance played a major role in designating the caretaker for aging parents. In societies where a stem family system dominated, as Fauve-Chamoux and Morioka point out respectively for the Pyrenees and Japan, the main caretaking responsibility for aging parents fell upon the inheriting son, who continued to coreside with his parents after marriage, while the noninheriting children left home and maintained contact only from a distance. In the northern Iberian Peninsula, too, as Brandes points out, in areas with impartible inheritance, the inheriting son was also the caretaker for aging parents. In areas with partible inheritance, patterns of caretaking with or without coresidence took different forms. In societies with partible inheritance and in landless societies, selection of the caretaking child varied by economic circumstances, life-course antecedents, and cultural prescriptions. As Ehmer shows, in central Europe, caretaking patterns for aging parents differed between landholding families and those who engaged in small commodity production in the household, because of the difference in inheritance practices. In all these circumstances, designation of a caretaker for parents in their old age was an important life course "imperative" that often necessitated advance provisions, such as inheritance contracts or other strategies. Among landless families in rural and urban society, the selection of the caretaker was contingent on specific arrangements between parents and a caretaking child that were formed over life. Other means to select the caretaking child were, therefore, at work: As

Introduction: Aging and Generational Relations Over the Life Course

9

Hareven and Adams found, among several ethnic groups in the United States at the turn of the century, the "parent keeper" was designated by the parents earlier in life. Parents often discouraged the youngest daughter from leaving home or marrying because she was designated to be the main "parent-keeper. " Similarly, in Sri Lanka, Uhlenberg found a strategy of reciprocity between parents' providing assistance to their children early in life and their receiving support from their children in old age. These types of strong generational reciprocities are also evident in the patterns discussed in this book by Morioka for Japan, Knodel and his collaborators for Thailand, and Hermalin and his collaborators for Taiwan. The forging or dissolving of generational ties that takes place over the life course has important consequences for relations later in life. As Elder and Robertson demonstrate, in rural Iowa the migration of sons from the farm in order to pursue occupational careers in the city put a strain on generational relations. When sons continued in farm careers and resided in the same communities with their fathers, they were more consistently engaged in caring for and supporting their aged parents than those sons who had moved to the city. One of the persistent questions underlying the chapters in this book and the discussion in the conference involves the impact of demographic factors and the availability of "caretakers" for aging parents. Does higher fertility guarantee a couple or an individual surviving into "old age" a larger number of caregiving children in old age? Or, as is evident in nineteenth-century American society, regardless of how many children a couple had, did only one child carry the major role of parent-keeper? (Chudacoff & Hareven, 1978; Smith, 1981). Alter and his co-authors conclude that in nineteenth-century Belgium, high fertility made a larger number of supporters available to aging parents. Caretaking of aging parents was related not only to the number of available children, but also to inheritance systems, the cultural values governing generational supports, and the importance of gender and birth order of the children in shouldering these responsibilities.

Assistance To Aging Parents From Children and Other Kin Who Do Not Coreside Related to the designation of children as caretakers is the question of the respective roles of adult children and other kin in carrying such responsibilities in later life. When children served as the primary caretakers, how was their assistance augmented by that of other kin, and, if it was, by what other kin? For example, were older widows more likely to receive assistance from their siblings than from their children? Or did siblings step in only when children were not available? How consistent or continuous the support from nonresident children or other kin to aging parents has been in the United States is still widely open to future research. Earlier studies have documented visiting patterns and telephone communication rather than regular caretaking. More recent research has emphasized the existence of various supports from adult children to aging parents, even if they were not residing in the same household.

10

T. Hareven

The authors in this volume agree that the most consistent and reliable form of caretaking of aging parents is best achieved if they coreside with an adult child. Uhlenberg found that in Sri Lanka, supports to aging parents by noncoresident children are limited to financial contributions rather than to caretaking. In Thailand, Knodel and his collaborators concluded that the type of supports provided by children who do not coreside depend on the children's proximity to aging parents. Children living closer to their parents are more likely to provide food and clothing, while children living at a greater distance more commonly provide money. Morioka is skeptical as to how reliable and effective the caregiving of aging parents in Japan would be without coresidence. In the region of Emilia Romagna, in Italy, Marzio Barbagli finds that, on the other hand, supports to aging parents from adult children who do not coreside are persistently widespread. The relationships are, however, patrilineal rather than matrilineal: The assistance comes from sisters-in-law and daughters-in-law rather than from sisters and daughters. In American black families, as James Jackson, Rukmalie Jayakody and Toni Antonucci point out, the support relations are predominantly matrilineal, but these supports are subject to a complex pattern of interaction of various factors: income, educational levels, proximity in residence, and the environment. Such complex models have not been applied, however, to the white American population or to European populations. The role of kin — other than children who do not coreside with elderly people — in providing assistance still awaits systematic investigation. The extent to which kin are available to provide supports to aging relatives is still an open question. Through their modeling of demographic determinants of kin availability to assist aging relatives, Douglas Wolf, Beth Soldo and Vicki Freedman have concluded that kinship networks are determined not only by demographic factors, but also by the familial characteristics of network members and by their familial position vis a vis aging relatives. For example, whether siblings are married or not and what kind of obligations they carry toward the parents of their spouses helps determine their availability. The most important conclusion reached in their chapter is that generational assistance must be studied in a familial, rather than an individual, context. Configurations of siblings and the marshaling of assistance and division of responsibility among them are significant factors. Since the data used by these authors is demographic, the general question concerning kin availability still needs to be addressed: to what extent is the presence of kin proof of their actual involvement in care for the elderly? Along these lines, John and Matilda Riley emphasize the significance of "latent" kin relations or of surrogate kin networks as a future avenue for generational assistance. Major changes in the timing of life transitions or, in what Anne-Marie Guillemard calls the "deinstitutionalization of the life course" have also had a significant impact on reciprocities among the generations. As she points out, the erratic timing in exit from the work life, in recent years, and the deinstitutionalization of retirement, have rendered the ability of generations to assist each other less predictable.

Introduction: Aging and Generational Relations Over the Life Course

11

The Impact of Social Change What was the role of the welfare state and of charitable or public welfare services and institutions, such as homes for the aged and nursing homes, in influencing generational coresidence and assistance? As Guinnane shows in Ireland, the 1908 Pensions Act did not bring about the system of generational coresidence; it modified, however, the existing system by providing an increase in elderly parents' coresidence with unmarried children and a corresponding decline in their coresidence with married children. For Sweden, Lars Goran Tedebrand attributes the decline in coresidence among generations over the twentieth century to the development of the welfare state. Similarly, Morioka highlights the significance of the Japanese social security policy and social welfare services both in substituting for an earlier intensive kin support system for the aged and in providing "an effective buffer system to keep the coresidence functioning positively," in those cases where coresidence is still surviving. Hareven and Adams demonstrate that the willingness of elderly people and of their adult children to accept help from the welfare state and external agencies has varied by their ethnic background and their cohort experience. Later cohorts, who are less kin-oriented and who have had a greater exposure to bureaucratic agencies over their lives, were more inclined to view public agencies as sources of support than earlier cohorts, who relied more exclusively on kin assistance over their lives. All the issues discussed above need to be interpreted in the larger context of social and economic change. Did industrialization and urbanization lead to the dramatic restructuring of generational relations and mutual assistance? A comparison of the patterns in Europe and the United States with those in East Asia raises important questions about global change: Is the trend toward the residential separation of older people from their children in Asian countries part of a "convergence" process with the West? Are Asian countries recapitulating at a more rapid pace the historical transitions that have already occurred in Western Europe and the United States? Even if this were the case, are similarities between Asian and Western societies a disguise for much more profound cultural differences? For example, do those elderly parents now living in a nuclear household in Japan and Taiwan experience an interaction with adult children similar to that of contemporary Americans living in nuclear families? Cultural traditions have played a major role in shaping patterns of generational assistance and the expectations of generations from each other. What may appear on the surface to be an "isolated" nuclear family by no means suggests that elderly parents are isolated from assistance from their children and other kin. This is precisely an area where comparative research is needed. Such comparisons need to employ, however, the same variables on types of supports and services across several societies. But even such systematic comparisons require sensitivity toward the internal differences within each society and differences in the cultural meaning of family relations and obligations for support across societies. Within Taiwan, for example, Hermalin and his associates found major differences in supports to aged parents between Mainland Chinese and Taiwanese. Within the United States, one needs to examine such differences among ethnic groups and social classes more systematically.

12

T. Hareven

When addressing these questions, it is also important to keep in mind that patterns of change that appear on the surface to be similar may not really be the same. It is not sufficient to compare changes in Asia and Western Europe without understanding the difference in the starting points of these changes in the societies examined. This issue in itself provides a future agenda.

References Chudacoff, H. & Hareven, T. K. (1978). Family transitions to old age. In T. K. Hareven (Ed.), Transitions: The Family and the Life Course in Historical Perspective, 217-43. New York: Academic Press. Elder, G. H. (1974). Children of the Great Depression. Chicago: Chicago University Press. Elder, G. H. & Hareven, T. K. (1992). Rising above life's disadvantages: From the Great Depression to global war. In J. Modell, G. H. Elder, Jr. & R. Parke (Eds.), Children in Time and Place, 47-72. New York: Cambridge University Press. Hareven, T. K. (1971). The history of the family as an interdisciplinary field. Journal of Interdisciplinary History, 2 (Autumn), 339-494. Hareven, T. K. (1991b). The history of the family and the complexity of social change. American Historical Review, 96(1), 95-124. Hareven, T. K. & Masaoka, K. (1988). Turning points and transitions: Perceptions of the life course. Journal of Family History, 13(3), 271-289. Laslett, P. (1977). Family life and illicit love in earlier generations. Cambridge: Cambridge University Press. Laslett, P. & Wall, R. (Eds.) (1972). Household and Family in Past Time. Cambridge: Cambridge University Press. Smith, D. S. (1981). Historical change in the household structure of the elderly in economically developed societies. In J. G. March, R. W. Fogel, E. Hatfield, S. B. Kiesler, & E. Shanas (Eds.), Aging: Stability and Change in the Family, 91-111. New York: Academic Press.

Kinship and Care for the Aged in Traditional Rural Iberia Stanley Brandes University of California, Berkeley

Interest and Emotion in Iberian Care for the Aged In 1880, while excavating in the northwestern Portuguese province of Minho, archaeologist F. Martins Sarmento heard a curious cautionary tale. He was told that long ago sons would cart their aged fathers to a deserted mountain, popularly known as O Picoto do Pai — roughly, "Papa's Peak" — where they would be given a blanket and a loaf of bread and left to die. The legend states that one day an old man was taken to Papa's Peak by his son. About to suffer abandonment, the old man counseled the son to leave him only half a blanket and take the other half for himself, to use when it was his turn. Startled and confused by this advice, the son asked whether he would also eventually die on the mountain. "Well, what else?" responded the old man. "I brought my father here, you bring me here, and your son will treat you the same. " Upon hearing this response, the son hurriedly put the old man back in the cart and carried him home. In the Minho, villagers believed that this event provided a turning point in the history of human relations. The tale has become transformed over time into a myth that describes how patricide came to disappear, an event that the people take to represent the beginning of civilized life (reported in Bouza-Brey, 1982, pp. 81-82). This legend, as it turns out, has a long history and wide distribution (Aarne & Thompson, 1961, Tale Type 980A; Ralston, 1982).1 A variant is found in the medieval tale of Handlyng Synne, as told by George Homans in his account of thirteenth-century rural England (cited in Behar, 1986, p. 93). Present-day versions have been reported from Spanish Galicia (Bouza-Brey, 1982, p. 83) and the Old Castillan province of Valladolid (Diaz Viana, personal communication, 1990), where Luisa González, a seventy-two year old informant, stated that she learned it from her father when she was a child. Legends, like all folklore, provide more than mere entertainment. They reflect both the anxieties and resolutions to those anxieties that are built into the social order. We may therefore surmise that Luisa González's father was moved to tell this tale by his own sense of insecurity about the future. He no doubt hoped it would inspire his daughter, out of enlightened self-interest, to take care of him in old age. With specific regard to the Portuguese version of this tale, it is striking that the old father is saved from abandonment and death neither from a sense of guilt nor from filial love. It is rather from what at best we might call outright selfishness, at worst cowardly fear, that the grown son decides to rescue his elderly father and take him home. The story portrays three generations of men — father, son, and grandson — whom we may presume

14

S. Brandes

to be interrelated by close affective ties. They are also motivated by the most basic kind of material interest: the struggle to live. As a kind of social drama, the tale places in relief a general message forcefully articulated by Hans Medick and David Sabean: "Material interest cannot be excluded from consideration in analyzing the family in those contexts where it fails to be explicitly articulated." As those authors point out, "the practical experience of family life does not segregate the emotional and the material into separate spheres, but is shaped by both at once, and they have to be grasped in their systematic interconnection" (Medick & Sabean, 1984, p. 11). When analyzing intergenerational relations within the family, it is impossible to overlook that people who love one another also calculate more or less openly the tangible benefits to be derived from their kinship ties. Because of the intertwining of interest and emotion in the study of Iberian aging and intergenerational relations, the ethnographic evidence seems to place parents and children in adversarial roles. According to my own field observations, as well as the social science literature on rural Spain and Portugal, parents and children do frequently think of one another in what might be called exploitative terms. They know that health and wealth confer material advantages and that these advantages can help them to secure the maximum in goods and services from their kinsmen. The published record certainly reveals a kind of jockeying between parents and children for actual or potential benefits. Especially where poverty is evident and modern social security systems are absent — conditions that prevailed until the 1970s throughout most of the Iberian countryside — parents and children necessarily view one another as exploitable resources. In Spain and Portugal, as elsewhere (e.g., Bulatao & Lee, 1983; Caldwell, 1982; Fawcett, 1972, Mamdani, 1972), even the decision to have children must be understood partly in such crude economic terms. Nonetheless, it is good to remember that in traditional rural Iberia all the members of a single family — parents, children, and grandchildren — depend on one another to maintain the family's good name. In the context of the small village or town where most Spaniards and Portuguese until recently lived out their lives, what has always counted most is reputation. It is this supreme value that forms the basis of the well-known honor/shame complex (Gilmore, 1987; Peristiany, 1965) throughout the Mediterranean region, including Spain and Portugal. One's honor — that is, the esteem and respect that one commands — rests largely on family reputation, so that the behavior of parents and grandparents automatically determines the perceived trustworthiness, and hence to a great extent the economic prospects, of children and grandchildren (Brandes, 1987). Above all, one's position in the marriage market, the most salient avenue to mobility in stratified, state societies (Ortner, 1978) is at stake in a family's reputation (Brandes, 1987). In traditional rural Iberia, proper comportment — including proper care for the elderly — is the most precious asset that members of a family can give one another. Their collective self-interest in maintaining a good name and sensitivity to the opinions of those around them, therefore provide the broad context within which we must understand aging and intergenerational relations within the family. An individual can jockey within his or her family for power and resources.

Kinship and Care for the Aged in Traditional Rural Iberia

15

But inevitably there are limits to such behavior that cannot be transgressed without risking the ruin of one's entire family.

Shared Care for the Aged In the 1970s, while carrying out fieldwork in the tiny village of Becedas, in Old Castile, my family and I lived next door to a widow in her sixties named Consuelo Hernández. Tía Consuelo, as we called her, was born and raised in Becedas. She gave birth to five children, three sons and two daughters, although only the sons survived early childhood. She lived in her own house, adjacent to that of one of her sons. She was supported by small monthly payments from her three married sons, who worked the land and managed the village oven that she formally owned. She also received a small widow's pension from the government. Like most widows, Tía Consuelo feared being alone at night, so her grandchildren took turns sleeping with her. Her mornings usually were spent at housework and helping village women who came to bake cookies at the oven. In the afternoon she would knit or darn in the company of her sisters, both also widowed, or sit outdoors with female neighbors. Tía Consuelo's circumstances and outlook were probably typical of the aging Castilian peasant woman of her day. She maintained residential independence and an ongoing economic arrangement with her sons that would assure her a small but steady income. She also took for granted the centrality of motherhood both in her own experience and that of her family and society. One of her favorite proverbs supported this view: "May nobody cry for someone who has a mother" [El que tenga madre, qué no le llore nadie]. At the same time, she was certain that in general, parents remain more devoted to children than children do to parents. Again, her culture provided her proverbial confirmation of this position. She would reiterate time and again, "a mother's breast is worth more than anyone else's affection" [Vale más la teta del madre que el cariño de nadie] and "one mother is [enough] for a hundred children, but a hundred children aren't [enough] for one mother" [Una madre es para cien hijos, pero cien hijos no son para una madre]. When I asked Tía Consuelo to interpret this last proverb she explained that although a mother has within her sufficient love to be distributed among a hundred children, a hundred children are inadequate to give a single mother all the love she deserves. Mothers, according to Tía Consuelo, also have enough power to convey their boundless love; hence, "two breasts can do more than a hundred carts" [Pueden más dos tetas que cien carretas]. To my knowledge, Tía Consuelo never explicitly articulated a direct connection between her perspective on motherhood and her economic arrangement with her sons. Nonetheless, her viewpoint is one that we repeatedly encounter in the ethnographic record from Spain and Portugal. The literature indicates that, much as parents love and care for their children, Iberian peasants openly express skepticism about the degree to which their children will reciprocate those feelings. Hence, parents have always relied upon property and inheritance as a guarantee that they would not be abandoned in old age. To be sure, few Spanish or Portuguese peasants would deny that there are some truly devoted children who care for elderly parents out of a natural, spontaneous sense of love

16

S. Brandes

and duty. However, most assume that it is above all the expectation of material reward that motivates the young to look after the aged. "Relations between parents and children ought ideally to be unforced and disinterested, " reported anthropologist Ruth Behar (1986, p. 94), "when in fact [these relations] are often contractual and tainted with self-interest." In the Leonese community that Behar studied, as throughout the Iberian Peninsula, parents and children are bound to one another through the mutual expectation that property will be given in return for old age care. A Catalan peasant put it this way: "Bringing up an hereu (or male heir to the estate), the parents always have [their future] assured. It seems that he is an insurance, an unequivocal source of support for the father and mother" (Barrera, 1986, p. 201). Moreover, as a double guarantee against abandonment, Iberian parents generally retain legal rights to their houses and land until death. In the Portuguese Minho, for example, people warn that "whoever gives away his property before dying, deserves to be beaten up" (Pina-Cabral, 1986, p. 68), while far to the east, in Spanish Aragón, a widow on her deathbed recommended to a friend, "don't divide your things before dying; my children's dissatisfaction, not my sickness, is killing me" (Lisón-Tolosana, 1966, p. 163). There is no doubt that throughout rural Spain and Portugal, property has been the most powerful lever in achieving a secure old age. It is therefore impossible to understand care for the aged without considering inheritance. Both partible and impartible inheritance systems exist in the Iberian Peninsula, and each has produced a distinct pattern of old age care. In regions associated with partible inheritance — roughly the southernmost two-thirds of the Peninsula, including Andalusia and the two Castiles — sons and daughters inherit equally, and a high value is placed on absolute equality of rights and duties among the entire sibling set. Consequently, all brothers and sisters assume equal responsibility for failing parents. Although precise arrangements vary regionally, the pattern that I observed two decades ago in Becedas (Brandes, 1975, pp. 107-130; Brandes, 1984 pp. 32-35) was and still is widely practiced. In Becedas, virtually everyone strives to establish and maintain the nuclear family household, by far the preferred residential type. The formula advanced by Leopold Rosenmayr and Eva Köckeis to describe residential preferences of the elderly in Austria and the United States — that is, "intimacy — but at a distance" (Rosenmayr & Köckeis, 1963, p. 418) — perfectly describes the preference of most people in Becedas. It is, as those authors state, "this wish to maintain some distance but not to be isolated" (Rosemayr & Köckeis, 1963, p. 418; italics in the original) that Becedas villagers consider ideal. Like Austrians and Americans, people in Becedas believe that "the most satisfactory arrangement is to have their own home in the vicinity of their children" (Rosemayr & Köckeis, 1963, p. 419; italics in the original), but to live independent of them. "The married person wants a house" [El casa'o, casa quiere], states a Becedas proverb. It is considered almost inevitable that coresident relatives will cause problems between husband and wife that would not occur if the couple were living alone. Any form of joint or extended household in Becedas is assumed to be a compromise, necessitated by considerations of health or wealth. For this reason, in Becedas a customary practice prevails that enables newlyweds

Kinship and Care for the Aged in Traditional Rural Iberia

17

to accumulate enough capital so that they avoid having to live with parents (Brandes, 1973). Likewise, an elderly husband and wife whose children have married and left home continue to reside in their own independent house for as long as circumstances allow. When their strength fails, however, the grown children are mobilized to keep the parental household viable by moving there periodically. Typically, in Becedas, married sons and daughters, together with their respective families, temporarily live with the elderly parents on a rotating basis — por meses, or "by months," as the system is known — for anywhere between one and four months at a turn. Thus, a married son, together with his wife and children, actually pack their movables, including beds and mattresses, and bring them to the house of his parents for the established period of coresidence. (Anyone who happens upon Becedas is likely to encounter villagers carting furniture and other belongings through the streets.) During this time their own house remains locked and uninhabited. Afterward they move back home, and another sibling with his or her family takes over. When all the brothers and sisters have completed an equivalent turn, the cycle is renewed. During the period that a married couple lives with elderly parents, the daughter or daughter-in-law maintains the house and does the cooking. However, expenses are divided according to the size of each coresident nuclear family. When Alberto, his wife Marta, and their two children moved in with Marta's parents, the parents paid one-third the living costs (mainly for food), and Alberto and Marta, representing four members of the six-member joint household, paid two-thirds. Gifts and special holiday foods were carefully calculated on a separate basis. On one occasion, while my family and I were visiting Alberto and Marta for dinner at the parents' home, Marta's father-in-law produced an unopened bottle of cognac. Since we were invited by Alberto and Marta, we were sure that they would be held responsible for the cost of the liquor, which they could ill afford. Hence, we politely refused. We were later told that after we left, the father-in-law insisted that Alberto should pay for the cognac. The cost of the unopened bottle became a source of contention between the two men, though in the end the father-in-law, responsible for the purchase in the first place, was held accountable. As long as both aging parents are alive and well, the parental household is maintained intact. One Becedas villager said that to ask an old couple to move out of their house would be like "asking them to die. " However, as soon as the elderly husband or wife does die, the viability of the independent household is threatened. At this point, the widow or widower may be forced to shut the family home and move from one married child's house to another on a rotating basis. In this case, the same principle of sibling equality applies, and all married brothers and sisters take equal turns caring for the surviving parent. Meanwhile, the parental house remains closed. This system of rotating care for the elderly parents assures that the preferred nuclear family residence need be disrupted only temporarily. Whether a married couple takes an old mother or father into their own home or shares a household with married parents whose health is failing, the joint residential arrangement occupies only a portion of the year. Because care for the elderly is assumed to be the price that siblings pay for their equal inheritance, brothers and sisters, at least ideally are supposed to share the burden of

18

S. Brandes

parental care with almost compulsive equality. The people of Becedas go so far as to apply the rotational system even in cases where family members have migrated. Hence, widows or widowers may travel to Madrid (the destination of most Becedas migrants), where migrant children take a turn at caring for them. Long-distance displacement such as this usually occurs during the winter months, when Becedas, a mountain town, may become unbearably cold, creating a hardship for the elderly. In situations where a sibling lives very far away or is unmarried, or where the parents are unable to travel, the migrant brother or sister may be absolved from responsibility for parental care. To compensate the other siblings, he or she must provide the ones who do share the caretaking role with a monthly stipend or some other form of payment. To an outsider like myself, the Becedas rotational system seems at first eminently fair, brilliantly designed to avoid jealousies and resentment among siblings. In fact, despite the appearance of fairness and justice, this system of care by which the elderly are bandied around from one house to another often produces a good deal of unhappiness. For one thing, siblings do not always feel that the burden of parental care is divided equally. In one family, a sister and brother and their respective families take turns living at the home of their elderly parents. A third sibling — a sister — lives far away in northeastern Spain and is unable to share this responsibility. In compensation, the migrant sister allows her brother to work her village lands. He divides the profits from these lands with her elderly father. The sister who resides in Becedas receives nothing, which remains a sore point with her and her husband. In Becedas, married couples typically complain about even the brief periods of coresidence with elderly parents, which they sometimes seem to tolerate only on the promise of quick relief. As for the elderly themselves, they are all too aware that no single child is willing to sacrifice enough on their behalf to care for them on a long-term basis in a truly nurturant, loving, disinterested way. Even if their own child would wish to take on the burden alone (an improbable occurrence), it is believed that the son-in-law or daughter-in-law would invariably block such an arrangement. The people of Becedas apparently are not alone in recognizing the limitations of this system of parental care. Villagers in Ruth Behar's community of Santa Mara del Monte, located north of Becedas in the province of León, believe that "to pass around a mother in this way is viewed as a pathetic sign of discord between the generations" (Behar, 1986, p. 91). In Andalusia, much further to the south, the Becedas style of rotation used to be routine but is so no longer. Widowers in particular were said to be so miserable at being unceremoniously tossed from one house to another that they were sometimes driven to suicide (Luque Baena, 1974, pp. 122-123). I am still haunted by the memory of a funeral that took place in Becedas on the day, in April 1969, that my family and I settled in town. The deceased was an elderly widower who had been living with his children por meses, a suicide victim very possibly driven to self-destruction by the feeling that he was unwanted. This kind of violent death would not have been an unusual occurrence in traditional rural Spain. Spanish anthropologist Enrique Luque Baena (1974, p. 122) reported from his Andalusian informants that "it was frequent, at least through the first quarter of this century, that an elderly widower took refuge in suicide." What is more, Luque Baena (1974, p. 123) states that "the suicide of the elderly widower, obligated to resign himself

Kinship and Care for the Aged in Traditional Rural Iberia

19

to isolation in successive homes where he could contribute nothing and could only serve to disturb the cohesion of the nuclear family, seems to have been in former times behavior that people wanted and expected." 2 In the Zamoran community of Bermillo, about an hour's drive north of Becedas, a depressed widower declared to the Peruvian poet José María Arguedas, who resided during the 1960s in that town, "I already am worth very little, almost nothing. It's better to go to the next world when you reach this state" (quoted in Arguedas, 1987, p. 159). In Becedas itself, widowers are considered to be at a particular disadvantage. Devoid of virtually any homemaking skills, they are immediately forced into a situation of dependency on their children. The state of their health is irrelevant, for even if they are able-bodied they have been trained from early childhood to avoid even the most minimal household tasks. During the period of my residency in the village, there was only one grown man who lived by himself — a middle-aged shepherd whose transhumance kept him in Becedas about half of each year. Villagers singled out this man's unusual lifestyle — citing particularly that he slept in the same room with animals and lived among unmentionable filth — to illustrate why men should not reside alone. Widows, on the other hand, can and usually do maintain domestic independence. Only when their health fails must they begin the monthly rotation from one child's house to another. Even widows who live alone, however, rarely spend the night unaccompanied. As suggested above by Tía Consuelo's experience, it is a general principle in Becedas that grandchildren, including both boys and girls, should take turns — a month at a time — sleeping at a widowed grandmother's house, where they share her bed. One additional gender difference is significant in the villagers' attitudes towards rotational care: Elderly people typically prefer to share a house with married daughters rather than with married sons. Women in Becedas, as perhaps in most parts of the world (Chodorow, 1978, pp. 160-167), are perceived to be the primary caretakers. There is no doubt that married women in Becedas remain more closely attached to their parents than do married men, who, upon marriage, seem almost automatically to substitute dependence upon their mothers for dependence upon their wives (Brandes, 1975, pp. 107-120). For this reason in Becedas it is said that "daughter who you marry off, son who you gain; and if it is the son [who marries], son lost" [Hija que casas, hijo que ganas; y si es el hijo, hijo que pierdes]. The situation in Becedas is replicated throughout Iberia and, in fact, many other parts of the world, where, in the telling words of Janet Finch and Dulcie Groves (1983, pp. 4, 5-6), there exists an implicit "designation of women as the 'natural' carers... .[The] designation of women as carers is a social construct at the level of ideology, which has also concrete material implications." Men, even sons, are simply not expected to be able to care for elderly parents in the way that women can and should. At the same time, it is not just any woman, but specifically the daughter, whom people in Becedas deem the most capable of caring for a weak or ailing mother and father. One married daughter, who lives with her mother and father on a rotational basis, claims that daughters are motivated to look after elderly parents out of "affection" (cariño), whereas daughters-in-law provide care out of "obligation" (obligación). With few exceptions, mothers-in-law and daughters-in law, each with distinctive ideas about household

S. Brandes

20

maintenance, and perhaps also competing for the attention and love of the husband/son, seem tense in one another's company. Mothers-in-law, in particular, complain about ill treatment — real or imagined — that they receive from daughters-in-law. Hence, several elderly women reiterated to me the traditional refrain, Your daughter is right by your side [lit., by your blouse], Your daughter-in-law further away, She closes the door on you; And your daughter opens it. [La hija al pie de la camisa, Y la nuera más afuera, Te cierre la puerta; Y la hija te la abre.3 In fact, even sons-in-law are said to cause problems. "How good my son-in-law would be if he were far away!" [¡Largo es mi yerno, si fuera bueno!], states a popular village proverb. The Becedas rotational system, although placing equal responsibility for parental care upon all children, in fact proves burdensome to all parties involved. The system is difficult for elderly parents, who feel unwanted and for whom coresidence is in any event a necessary and unwanted departure from nuclear family independence; for daughters, upon whom the main burden of caretaking falls; and for daughters-in-law, for whom the arrangement is an unpleasant if temporary intrusion into their normal family routine. Because much of their day is spent away from the home in the fields and they are otherwise insulated against the caretaking role, sons and sons-in-law find the system less intrusive and burdensome than do women. Widowers, who constitute perhaps the most pitied, dependent, and unhappy segment of Iberian rural society, are from an emotional point of view the most severely affected of all by household rotation.

Undivided Responsibility for Parental Care In Becedas, only one circumstance predictably alters the rotational principle whereby married children care for aged parents: the presence of grown, never-married children at home. When children remain single, a relatively frequent occurrence (Brandes, 1976), they usually assume full responsibility for their mother and father. In return, they can expect to receive the parental house, in which case all other property is divided equally among their siblings. This solution resembles the norm in areas of impartible inheritance — that is, where families select one major heir. Impartible inheritance systems prevail throughout most of northern Iberia, including the Portuguese Minho as well as Spanish Galicia, Asturias, Catalonia, and the Basque country. Throughout Spain's northwest — in Galicia and Asturias, for example — the impartible inheritance system known as a millora (the Gallego term) or la mejora (the Spanish term)

Kinship and Care for the Aged in Traditional Rural Iberia

21

has persisted for centuries. A single heir — in Gallego, o millorado or, in Spanish, el mejorado (roughly, "favored beneficiary") — receives the family house and land, while other heirs split the remaining assets equally. The favored heir, in return, assumes responsibility for living with and caring for the parents until they die. The same is true in northern Portugal, where a proverb proclaims, "Quern é teu herdeiro? Quem te limpa o traseiro" — that is, "Who is your heir? The one who cleans your butt" (O'Neill, 1983 p. 68). For this reason, although male primogeniture is preferred, parents tend to select from among their children a major heir who can hope to satisfy two of their primary aspirations: first to maintain the economic viability and continuity of the family house, and, second, to provide them willing care and companionship in old age. Hence, not only sex and birth order but also the degree of compatibility between parents and adult children seems to be important in the selection of a favored heir. From a comparative standpoint, in areas where impartible inheritance prevails, parents have generally proven no more trusting of their children's good will than is true under partible inheritance systems. For instance, in historic northern and central Europe (Gaunt, 1983), as well as in nineteenth-century rural America (Lee-Whiting, 1985, pp. 69-73), parents assured themselves of a secure future by requiring their major heirs to sign inheritance contracts, guaranteeing maintenance in old age in return for farmland, animals, and the family house. Among the European peasantry, cautionary tales were common. Said historian David Gaunt (1983, p. 260), "a classic example [from Sweden] tells of a farmer who had turned over his property to his grown son and daughter-in-law and was being treated badly. Eventually he hit upon an ingenious plan and told the young couple that he hadn't given them all his wealth, but that he had more hidden away in a chest which he would leave to them when he died. From that moment his treatment became much better and continued to be good until death. After burial the son and his wife rushed to open the chest, but the only thing they found was a long-handled club on which a verse was written: He who gives so that he must beg Ought to be clubbed until he lies out flat. Although this proverb derives from sources published in the 1930s (Granberg, 1934; Trotzig, 1938), Gaunt (1983, pp. 259-260) also cited evidence from early nineteenthcentury Germany where, in some cities, a large club was hung from the town gates with the following inscription: He who has made himself dependent on his children for bread and suffers from want He shall be knocked dead by this club. Contemporary Iberian peasants are apparently affected by the same sentiment. Portuguese anthropologist Joäo de Pina-Cabral (1986. p. 68) stated that in the Alto Minho, one hears the saying, "Whoever gives away his property before death, deserves to

S. Brandes

22

be beaten up." A Spanish version of this apparently widespread proverbial expression, which I collected in Andalusia in 1976, declares: He who turns over his wealth before death Deserves a blow on the forehead with a hammer.

[El que da los bienes antes de la muerte Pa' qué le den con un martillo en el frente.] Virtually wherever we look, aging parents deem the love and affection of their children insufficient to assure themselves a secure future. Control over property is their means to avoid loneliness, poverty, and abandonment. In areas where a single child is responsible for aging parents, it is not only the parents who express insecurity about the future. Major heirs throughout Iberia have also felt vulnerable to the whims of their fathers and mothers. Caroline Brettell (1989, p.28) stated that in northwestern Portugal, where impartible inheritance systems prevail, "property was a form of old age security and it was dispensed in such a way as to guarantee the care of the older generation. The rights and obligations between parents and children were clearly negotiated and the fact that transactions often constituted promises rather than actual transfers meant that parents could maintain control over at least some of their children throughout their lifetime." In fact, there are cases on record in which dissatisfied elders very occasionally, but decidedly, have exercised their right to disinherit the originally favored child. Disinheritance occurred, for example, in 1862 in the Galician community of Seone de Caurel. In that year, the widow Vicenta Garcia Novo, in a supplement to her will, declared that her son, Manuel, whom she had designated as principal heir the year before, be denied the major inheritance. To justify this extreme measure, the testament states that Manuel "has not fulfilled nor fulfills the obligations of a good son, (he) does not feed her, provide her with shoes or clothing, much less aid in her afflictions, looking on her with harshness and little respect. " Therefore, he proved himself unworthy of "the gift of the mejora," which Vicenta bequeathed instead to her eldest child, a married daughter (Bauer, 1983, p. 36). Rainer Bauer (1983, p. 36), who has studied a number of such cases in Seone, observed that "most commonly, parents and their major heirs broke over whether the heir was permitted to emigrate for a time, over his marriage, or over the question of how the casa ['estate'] was to be administered. Parents in these instances attempted to manipulate the millora, to determine the behavior of a favored child.... When this child proved intransigent..., he was denied the millora." Where inheritance contracts prevail in Iberia, as for example in the Basque country, parents and children might hope to avoid at least the economic repercussions, if not the emotional unpleasantness, of altered sentiments. Anthropologist William Douglass reports that in the Basque village of Murélaga, transfer of the elders' property to the heir occurs when the heir marries. The document, prepared by a notary public and signed in his presence by both couples, stipulates not only that the parents relinquish the family estate to the newly married heir and spouse, but also that the two couples should coreside. This contract, according to Douglass, also "directs them to live amicably and enjoins the young

Kinship and Care for the Aged in Traditional Rural Iberia

23

couple to treat the old couple with respect and affection. " It may specify that "the donors have no obligation to continue working..." It may stipulate that "the old couple receive a daily cash stipend for their personal use. Commonly, the donors reserve the right to one-half the cash profits accruing to the farm" (Douglass, 1969, pp. 101-102). In cases where these obligations are not met, or where the parents and children fail to get along, the contract stipulates which of the two couples will leave the family estate, the manner in which the movable property is to be distributed, and provisions for the financial support of the elders (Douglass, 1969, p. 102). As in northern Europe and America in centuries past, contracts such as these leave little to the enduring good will and affection of aging parents and their children. To establish the basis for a financially and emotionally stable future, peasant couples in areas of impartible inheritance usually had some choice as to which child would become heir and caretaker (e.g., Bauer, 1987, pp. 174-176; CaroBaroja, 1968, p. 164; Douglass, 1969; O'Neill, 1983, p. 65). Childless couples, however, have had to resort to a kind of pseudo-adoption to provide for their future security. The most frequent solution throughout Spain and Portugal is for barren couples to take into their home a young boy or girl — usually a niece or nephew with scant inheritance prospects — and raise the child as their own. In such cases, there are no formal adoption procedures; everybody knows the child's natural parentage. But over time, this niece or nephew becomes the couple's future heir and caretaker. This arrangement occurred during the period of my fieldwork in Becedas. Likewise, Rainer Bauer (1983, p. 55) noted that in Seone de Caurel a seventy-six year old Seone spinster named her grandniece the beneficiary of her estate, as she said, "so that she takes care of me. " Carefully worded wills are, and long have been, associated with such adoptions and related caretaking arrangements. Caroline Brettell, writing about the northern Portuguese parish of Lanheses, located in the same general region as Becedas and Seone, reported on a document dating from 1792 in which a childless couple left their estate to a niece. However, the cautious couple made sure to include the following: that "this niece, because we love her and have received many services from her which we hope she will continue to provide, we leave her our house in the lugar of Rocha and the lands, stables, vines, threshing floors, and outbuildings...on the condition that she behaves well and that she does not marry without our consent, and that she remains with us until we die, because if she shows ingratitude, this legacy will not have effect" (quoted in Brettell, 1986, p. 45). Occasionally, an unrelated child may even take precedence over one's natural children. Among the vaqueiros de Alzada (Cátedra Tomás, 1988, p. 171), a distinctive ethnic group located in rural Asturias, a widower took an unrelated boy into his house as heredeiro — the term by which a nonfilial heir is known in Asturian (to distinguish from el mejorado) — when all his own children emigrated. The children themselves had no hope of ever returning to the parental home, and the widower aspired to the continuity of his estate and to old age security. Like others in this situation, he managed to leave his property intact to the heredeiro, despite the illegality of an act that in essence resulted in filial disinheritance. Among vaqueiros who find themselves without an heredeiro, there is still one last resort for old age care. In case of permanent sickness or incapacitation, a widow

24

S. Brandes

or widower can fall upon the mercy of a willing neighbor. The neighbor takes the ailing man or woman into the home and cares for him or tier until death. In return, the caretaker becomes sole heir to the deceased's house and land. As with heirs who were actual kin, adopted heirs in this whole area of northwestern Iberia (including Galicia, Asturias, and northern Portugal) could occasionally be disinherited. Brettell encountered one such instance in Lanheses, in which a couple who had been contracted to care for an elderly spinster was deprived of their promised inheritance because their caretaking did not satisfy the donor, who had therefore been forced to depend upon the "Christianity" of the community. The spinster eventually left her property to the Church (Brettell, 1986, p. 45).

Recent Changes in Aging and Intergenerational Relations Traditionally, in areas of impartible inheritance, the favored heir was an object of envy. Privileged not only by the economic security that house and land represented, but also by the prestige that automatically came with managing the undivided estate, the heir could be counted upon gladly to assume responsibility, not only for the property, but also for the elderly parents. As Susan Harding (1984. p. 163) observed in a community in the Pyrenees, "no matter how modest the casa, its heir had been the high-ranking, most fortunate, sibling." The heir's never-married brothers and sisters could always opt to live at the parental house, and share in that farming enterprise, but under such circumstances they would forever remain subordinate to the heir. At the same time, the heir's ultimate subordination to the parents was assured, given that parents could sanction filial ingratitude by disinheritance. Another, more suitable child was almost always available as a replacement. Although the historical and ethnographic records are discontinuous, and it is difficult to measure to the degree to which actual behavior conformed to cultural ideals and expectations, it seems that some sort of coresidence between aging parents and their grown children endured as long as nonfarm work was scarce. Marked changes in residential patterns and intergenerational relations have occurred with the expansion of the Spanish and Portuguese economies after World War II, and particularly since the 1960s. For the past several decades, emigration to cities and foreign countries has resulted in considerably greater affluence and higher living standards than could be achieved by remaining on the parental estate. Nowadays, states Carmen, a Gallega from the maritime province of La Corufta, whose life history was collected in the 1970s by Hans and Judith-Maria Buechler (1981, p. 163), "so many old people have been left by their children with nothing to eat. [The parents] stayed because they had to. " A Gallego journalist observed what is evident to anyone who drives through the countryside: hamlets constituted almost exclusively of old people, some with only "five inhabitants or less" (Mandianes Castro, 1989). One old Gallego peasant stated that "everyone labels me a hermit, but nobody explains why they flee from me" (quoted in Mandianes Castro, 1989 ); another lonely man lamented, "when you don't have

Kinship and Care for the Aged in Traditional Rural Iberia

25

anyone around to tell about your troubles, your blood becomes poisoned and you begin to go crazy from turning things over and over in your mind" (Mandianes Castro, 1989). Throughout rural Iberia, children try to avoid rural life, and heirs are no longer considered privileged siblings. In Susan Harding's Pyreneean community, the postwar changes have been so dramatic that already by 1970 "the heir had lost special status and was acquiring an aura of being the unfortunate, disadvantaged, socially inferior sibling" (Harding, 1984, p. 163). About the same time, a Gallego reported to Carmelo Lisón-Tolosana (1974, p.5) that "we believe that the millorado [or favored heir] is the jeopardized one." The current situation in Galicia (Bauer, 1983, p. 269), seems to be typical: "Young adults who still reside in their natal casas...remain at home by default of their siblings. The decline of the agropastoral enterprise and the waning of...indigenous culture have largely converted the privileges of the millora into burdens. " In fact, in Seone de Caurel heirs agree to stay on only if they gain control over their parents' social security payments, a welfare arrangement unavailable to independent Spanish farmers as recently as three decades ago (Bauer, 1983, p. 48). The same is true in Andalusia, where at least one anthropologist (Luque Baena, 1974, p. 124) reported that social security pensions have made it more attractive to care for elderly mothers and fathers; hence a child who needs additional income often will assume full caretaking responsibility for the parents. Then, too, we find that an heir's siblings, rather than being overcome with envy, are likely to compensate him or her with loans, cash gifts, or even their minor portions of the estate, in order to assure that the parents receive quality care (Bauer, pp. 58-59). In addition to extra monetary compensation, a grown child who stays on the parental estate wields considerably more power over household decision making than used to be the case. Sociologist Víctor Pérez Díaz, who carried out fieldwork in the Tierra de Campos in the 1960s, claimed that the rapid mechanization of agriculture in that part of Spain derived mainly from pressure coming from children. In that period, the only way to keep a son and his family at home was the "immediate purchase" of a tractor (Pérez Díaz, 1966, p. 85), and that in fact was the course that most aging parents took. Pérez Díaz wisely recognized the parental ambivalence prevalent in the Tierra de Campos in those days. In their role as parents, farmers favored the maximum economic well-being of their children, leading inevitably to the children's emigration from the countryside. As modernizing and aging entrepreneurs, however, farmers yearned for children to stay with them. The degree of influence and control that parents could exercise over their children's future, however, decreased markedly with time, so that as children grew older they became less and less willing to accommodate to traditional rural conditions (Pérez Díaz, 1966). Either those conditions would change radically or the children would leave—and they might leave anyway. The presence of viable alternatives to rural life has seriously affected the status and security of the aged themselves. Julio Caro Baroja (1968, pp. 164-65) wrote of a family in his native Basque community of Vera de Bidasoa wherein none of the three children wanted to stay at the parental home. The Buechlers' Gallegan informant, Carmen, declared that children used to coreside with their elderly parents "because.. .at least they had a house to live in. Often the parents would treat them badly, order them around and give them no

26

S. Brandes

freedom. But they had nowhere else to go. Now it's the other way around. The parents are quiet, humble; they give their children whatever they can so that later, one of them at least will stay to help him in their old age" (Buechler & Buechler, 1981, p. 163). At the same time, social security has provided a measure of financial support for older, formerly poverty-stricken peasants with landholdings insufficient or only barely sufficient to maintain themselves. Coresidence with grown children, albeit medically desirable at times, is no longer economically essential. Carmen declared that prior to the 1970s, when social security became available to independent farmers, "the old people used to drag themselves out to work until they died, and if they had a son or a married daughter they would have to beg them for every cigarette they smoked. They had nothing of their own. Now see how they live. They aren't exactly free of problems, but at their age, they always have something. Some old folks in the village have hens, a pig, or a small piece of land to plant their potatoes. They get along. And older people don't eat very much anyway" (Buechler & Buechler, 1981, p. 183). The advent of social security has meant that the services of one's children in terms of health care are more important to the elderly than whatever financial benefit, in the form of farm labor, those children might provide. Increasingly, however, old age homes are a common recourse. Even where these residences do not provide a permanent home, married children have begun to rely on them to intern their ailing parents while they go on vacation in August (Nogueria, 1989). In Madrid in 1989, there were 13,000 beds in old age residences. Of these, 5,000 were publicly financed, although the waiting list for these accommodations has become so long that it takes several years to be admitted. In Spain as a whole, it is estimated that only 2.5 percent of people over 65 are institutionalized (Nogueria, 1989), although the numbers are rapidly increasing. Said one commentator, "things have changed markedly and rapidly. From the shrinking size of houses to female employment, diverse factors combine to make it difficult to take care of the aged at home, above all if they are incapacitated" (Nogueria, 1989, p. 3). Throughout Spain and Portugal, care for the elderly has changed with changing material circumstances. Above all, traditional power relations between the generations have been inverted. But the desire to maintain nuclear family independence, as well as the relative economic strength of parents and their children, still determines patterns of old age care. Neither traditionally nor contemporaneously has the prestige of the elderly or the benevolence of the young proven sufficient motivation for caretaking. Now, as before, enlightened self-interest remains the most certain protector of the aged.

Notes Earlier versions of this paper were presented at the Annual Meeting of the Social Science History Association, Washington, D.C., November 16-19, 1990, and to the Bay Area Demography Group on October 3, 1991. I am grateful to commentators from those meetings, as well as to members of my University of California at Berkeley graduate seminar on Mediterranean Europe, Fall 1990, for providing me constructive criticisms of the presentation.

Kinship and Care for the Aged in Traditional Rural iberia

27

1

A Japanese version of this legend, called ubasute-yama (glossed approximately "abandon old woman on the mountain"), is recounted in the movie Narayama Bashiko (The Song of Narayama). Andrejs Plakans (1989, pp. 177-178) reported that "a northern European folktale tells of a grandfather being pulled on a sled toward a nearby forest by his middle-aged son and young grandson, where the old man is to be abandoned. The grandson is heard to warn his father not to leave the sled behind, because he, the grandson, would need it one day to do the same thing with his father, the man now pulling the sled. "

2

On the islands of Formentera and Ibiza, the suicide rate is higher for widows than widowers. Psychologist Claudio Alarco von Perfall (1981, p. 269) explains this pattern simply by reference to demographic balance: "female longevity, the greater number of women and the Civil War. "

3

From Andalusia, in the south of Spain, I learned a proverb that reinforces the same point of view: "Mother-in-law and daughter-in-law [do not fit together] even within a threshing floor; and mother and daughter [fit together] within a shirt" [Una suegra y una nuera, ni en una era; y una madre y una hija en una camisa].

References Aarne, Α., & Thompson, S. (1961). The types of the folktale: A classification and bibliography. Helsinki: Academia Scientiarum Fenilica. Alarco von Perfall, C. (1981). Cultura y personalidad en Ibiza. Madrid: Editoral Nacional. Arguedas, J. M. (1987). Las comunidades de España y del Perú. Lima: Ministerio de Agricultura, Pesca y Alimentación. Barrera, A. (1986). Primogenitura y herencia en la Catalunya Velia: Norma ideal y pràtica actual. In Los Pirineos: Estudios de antropología social e historia (pp. 177-215). Madrid: Universidad Complutense, Casa de Velázquez. Bauer, R. L. (1983). Family and property in a Spanish Galician community. Unpublished doctoral dissertation, Stanford University. —. (1987). Inheritance and inequality in a Spanish Galician community, 1840-1935. Ethnohistory, 34, 171-193. —. (1986). Santa Maria del Monte: The presence of the past in a Spanish village. Princeton: Princeton University Press. Bouza-Brey, F. (1982). Referencia a una eutanasia familiar primitiva en el folklore gallego-portugués. In T. I. Vigo, Etnografia y folklore de Galicia (pp. 81-91). Spain: Edicions Xerais. Brandes, S. (1973). Wedding ritual and social structure in a Castilian peasant village. Anthropological Quarterly, 46, 65-74. —. (1975). Migration, kinship, and community: Tradition and transition in a Spanish village. New York: Academic Press. —. (1976). La soltería, or why people remain single in rural Spain. Journal of Anthropological Research, 32, 205-233. —. (1984) Nombres que enganyen: Cine problèmes en la interpretació de dades censáis en l'Espanya rural. [Tricky numbers: five problems in the interpretation of rural Spanish census data]. Quaderne (Barcelona), 5,28-43. —. (1987). Reflections on honor and shame in the Mediterranean. In Honor and Shame and the Unity of the Mediterranean, edited by D. D. Gilmore (pp. 121-134). Washington, DC: American Anthropological Association Special Publication No. 22. Brettell, C. B. (1986). Men who migrate, women who wait: Population and history in a Portuguese parish. Princeton: Princeton University Press.

28

S. Brandes

—. (1989 Property transmission, life course transitions and family relations: A Portuguese perspective on a Mediterranean problem. Unpublished ms. prepared for the conference, The Historical Roots of the Western Family, Bellagio, Italy, May 22-26. Buechler, H. C., & Buechler, J. M. (1981) Carmen: The autobiography of a Spanish Galician woman. Cambridge, MA: Schenkman. Bulatao, R. Α., & Lee, R. D. (Eds.)· (1983). Determinants of fertility in developing countries. 2 vols. New York: Academic Press. Caldwell, J. D. (1982). Theory of fertility decline. New York: Academic Press. Caro Baroja, J. (1968). Estudios sobre la vida tradicional española. Madrid: Ediciones Península. Cátedra Tomás, M. (1988). La muerte y otros mundos. Barcelona: Jcar Universidad. Chodorow, N. (1978). The reproduction of mothering. Berkeley: University of California Press. Douglass, W. A. (1969). Death in Murélaga: Funerary ritual in a Spanish Basque village. Seattle: University of Washington Press. Fawcett, J. T. (1972). The satisfactions and costs of children: Theories, concepts, methods. Honolulu: East-West Center. Finch, J., & Groves, D. (1983) A labour of love: Women, work and caring. London: Routledge and Kegan Paul. Gaunt, D. (1983) The property and kin relationships of retired farmers in northern and central Europe. In Family forms in historic Europe, edited by R. Wall (pp. 249-279). Cambridge: Cambridge University Press. Gilmore, D. D. (Ed.). (1987) Honor and shame and the unity of the Mediterranean. Washington, DC: American Anthropological Association Special Publication No. 22. Granberg, G. (1934) Släktklubban. Budklavan, 13,1-13. Harding, S. (1984) Remaking Ibieca: Rural life in Aragon under Franco. Chapel Hill: University of North Carolina Press. Lee-Whiting, B. (1985) Harvest of stones: The German settlement in Renfrew County. Toronto: University of Toronto Press. Lisón-Tolosana, C. (1966) Belmonte de los Caballeros: A sociological study of a Spanish town. Oxford: Clarendon. —. (1974) Antropología cultural de Galicia. 2nd Ed. Madrid: Siglo XXI. Luque Baena, E. (1974) Estudio antropológico social de un pueblo del sur. Madrid: Tecnos. Mamdani, M. (1972) The myth of population control: Family, caste and class in an Indian Village. New York: Monthly Review Press. Mandianes Castro, M. (1989) El drama de la Galicia rural (y II). Diario 16 de Galicia, August 4. Medick, Η., & Sabean, D. W. (1984) Interest and emotion in family and kinship studies: A critique of social history and anthropology. " In Interest and emotion: Essays on the study of the family and kinship, edited by H. Medick and D. W. Sabean (pp. 9-27). Cambridge: Cambridge University Press. Nogueria, C. (1989) Los abuelos no son para el verano. Diario 16, Dossier de a semana, August 6, pp. 1-3. O'Neill, B. J. (1983). Dying and inheriting in rural Trás-os-Montes. In Death in Portugal: Studies in Portuguese anthropology and modern history, edited by R. Feijó, Η. Martins, & J. de Pina-Cabral (pp. 44-74). Oxford: JASO. Ortner, S. (1978) The virgin and the state. Feminist Studies, 4,19-37. Pérez Díaz, V. (1966) Estructura social del campo y éodo rural: Estudio de un pueblo de Castilla. Madrid: Editorial Tecnos. Peristiany, J. G. (Ed.). (1965) Honour and shame: The values of Mediterranean society. London: Weidenfeld and Nicolson. Pina-Cabral, J. (1986) Sons of Adam, daughters of Eve: The peasant worldview of the Alto Minho. Oxford: Clarendon.

Kinship and Care for the Aged in Traditional Rural Iberia

29

Plakans, A. (1989) Stepping down in former times: A comparative assessment of retirement in traditional Europe. In Age structuring in comparative perspective edited by D. I. Kertzer & K. W. Schaie (pp. 175-196). Hillsdale, New Jersey: Lawrence Erlbaum Associates. Ralston, W. R. S. (1982) Cinderella. In Cinderella: A Casebook edited by A. Dundes (pp. 30-56). New York: Garland. Rosenmayr, L., & Köckeis, E. (1963) Propositions for a sociological theory of ageing and the family. International Social Science Journal, 15,410-437. Trotzig, D. (1938) Ätteklubban. Folkminnen och folktankar, 25,64-69.

Household Patterns of the Elderly and the Proximity of Children in a Nineteenth-Century City, Verviers, Belgium, 1831-1846 George Alter Lisa Cliggett Alex Urbiel Indiana University

A previous generation of sociologists viewed the twentieth-century decline of household size and the growing residential isolation of the elderly with some alarm. They asked whether this represented an abandonment of the traditional European family system, in which families were assumed to have been large and extended or multigenerational. Subsequent research, however, has rejected both sides of this concern. Old people in the later twentieth century are not as isolated from their families as had been feared, and families in the past were not as large or complex as had been assumed. 1 The attack on the historical component of this conventional wisdom was led by Peter Laslett. Laslett and colleagues from a number of European countries were able to show the predominance of small, nuclear family households in Western European history (Laslett, 1972). Indeed, the size and structure of households in England have been remarkably constant for at least three centuries. Laslett's approach has been central to the historical demography of the European family for more than two decades, and it suffers from some well-known limitations. We intend not to review this entire debate here, but rather to focus on two aspects that are relevant to the study of the elderly. First, Laslett and his associates based their work on the study of the household, defined by coresidence, not on the family as defined by kinship. Clearly, kin outside the household can provide important support, and Laslett has always acknowledged this limitation. However, historical sources of information about the family are very limited, and most of this research is based upon censuses and other nominative lists. It is very difficult to reconstruct contacts with kin living outside the household. Second, discussions of historical changes in household structure have not included a theoretical model of the position of the elderly in European history. Laslett's early work was primarily an empirical critique of prevailing sociological theories, and was not motivated by a theoretical structure of its own. Increasingly, however, Laslett has come to rely upon the concept of the "European marriage pattern" as an explanation of the predominance of the nuclear family. Recognition of a distinctive European marriage pattern began with John Hajnal's empirical finding of a West European zone of late marriage and high proportions never marrying (1965, 1983). Hajnal's explanation for this pattern describes a social system in which couples are not able to marry until they are economically independent. The younger generation must postpone marriage while they

Household Patterns of the Elderly

31

save toward the purchase of a farm, or wait for the older generation to pass on their inheritance. Hajnal's description of the European marriage pattern is similar in spirit to Malthus's "preventive restraint." As a theory, the European marriage pattern does not provide an explanation of social support for the elderly. In contrast to Frederic Leplay's "stem family," or models of extended family systems that emphasize intergenerational relations, the European marriage pattern focuses attention on household formation among the younger generation. Some writers have noted the potential for intergenerational conflict in this system, where children cannot begin their own households until their parents retire or die (Shorter, 1977, pp. 32-34). Laslett recently has argued that the European family system could not make sufficient provisions for old age and that community and state institutions were necessary to supplement it.

Retirement in the past, therefore, seems to have been secured by a miscellany of expedients rather than by a set of recognized usages and institutions.... The general conclusion to be drawn is that the coresidential family group is very difficult to adapt to the eventualities of the individual life course, and providing for old age seems to be beyond its capacities; (Laslett, 1991, p. 125)

For this reason, Laslett argues that old age security could not have been a significant factor in maintaining the high birth rate in the past. We suggest here an alternative view of family support for the elderly in European history, which seems to be more consistent with evidence from a nineteenth-century Belgian city. We have attempted to identify and locate noncoresident children of elderly persons, so that we can describe their household patterns as choices. This addition helps to overcome an important weakness of most earlier research, which is limited to the coresident group. Our data suggest that coresidence with children was a very important source of support for the elderly, but that most of that support came from unmarried children. Our emphasis on the importance of support from children does not run counter to the main lines of research on role of marriage in the history of European household structure. On the contrary, we propose a different perspective on the European marriage pattern, which emphasizes the advantages for parents of children's marrying late. In the next section we attempt to view the European marriage pattern from the perspective of the elderly, and conclude that parents gained substantial economic advantages from keeping unmarried children in their households. The system did result in a large number of unmarried persons without children to rely upon, but old age security was greatly enhanced for those who did marry by having a large family.

32

G. Alter, L. Cliggett, and A. Urbiel

Late Marriage and the Economic Position Elderly Parents Although the European marriage pattern usually is presented in terms of economic circumstances effect on household formation among the younger generation, it is useful to consider their effect on the parental generation. Late marriage and high proportions never marrying created a large pool of unmarried children who were available to support elderly parents. Studies of nineteenth-century family budget surveys emphasize the importance of children's income in the nineteenth-century. An 1853 survey of Belgian families with four children estimated that 22% of total family income came from the earnings of children (Alter, 1984). The importance of children's earnings is also apparent in the multinational family budget survey conducted by the U.S. Commissioner of Labor in 1889-90 (see Haines, 1979). Income from children was much greater than the wage earnings of wives, who could only work part-time and who generally seem to have withdrawn from the labor force in the last half of the nineteenth-century (Alter, 1984). The contributions of children were especially important in later stages of the family life cycle because they compensated for the early decline of earnings by adult males. The peak in the age-earnings profiles of U.S. workers in the 1889-90 U.S. Commissioner of Labor survey occurred as young as 32 in some industries and usually before age 40 (Rotella & Alter, 1990). Leboutte (1990) showed similar patterns for Belgian mine workers, whether they worked below ground or on the surface. However, total family incomes did not decline in the households of older workers. The additional income contributed by children more than compensated for the lower earnings potential of male heads of household (Haines, 1979; Rotella & Alter, 1990). Children usually entered the labor force between ages 12 and 14, and wage rates rose rapidly with age. Total family incomes were highest when the head of household was age 50 to 60 and several unmarried, adult children were contributing wages. It is likely that most, if not all, of the earnings of children were pooled for the benefit of the family as a whole. Young children often worked as assistants to their parents or other kin (Smelser, 1959, pp. 187-193). In the early nineteenth-century, workers were often paid as teams, usually on a piece rate, and fathers were frequently in a position to collect the wages of adult sons and daughters. As their children entered the labor force, parents could expect to be relatively prosperous until their children went off to form their own households. A representative of the Société de Saint Francois Régis of Verviers, an organization established to combat illegitimacy by encouraging marriages, told a Belgian commission in 1886: "In the working class, often egoism, the desire to conserve the earnings of children, prevents parents from consenting to the marriage of their children" (Belgium, Commission du travail, 1887, Vol. 1, p. 1024). Just as reliance upon children could offset the reduced earning power of the male head of household, it could also serve as a strategy for support in old age. In a population with an average age at marriage above age 25, and between 10 and 20% never marrying, most parents could count on having at least one unmarried child after they had passed age 65. This reasoning suggests a logic linking high fertility with support in old age that differs

Household Patterns of the Elderly

33

from the usual account. It draws attention to the benefits from never-marrying children, rather than the average returns from all children, and to the timing of children during the parents' life course. Under this strategy, parents did not count on old age support from all of their children, but they did hope to have at least one child who would take care of them. Economic transfers from most children may have been small or zero, as long as substantial support came from this caregiver. In a large family there was a greater likelihood that at least one child would never marry and be available for this role. The timing of the first and last child played a particularly important role in the income stream of the parents. Couples preferred to have at least one child before the husband reached age 30, so that this child entered the labor force when the husband's income was beginning to decline. They also wanted the last child to be born as late as possible, so that this child's contribution would continue into their old age. It is possible to incorporate the timing of grandchildren into this strategy as well. As the parent passed age 70, the first set of grandchildren entered the labor force. At this point the oldest children in the middle generation were experiencing their highest total family incomes, and they were in a better position to share the burden of an elderly parent with their younger siblings. In this model, high fertility was encouraged by two aspects that are unique to highparity births. First, high-parity births occurred at older parental ages, and high-parity children contributed to the parental household when parents needed them most. Second, high-parity births occurred at a point in the family life cycle when total family income was highest, because their older siblings were already in the labor force. Thus, parents could use the earnings of early children to finance the births of later children, and the earnings of later children to finance their own old age.

Data from Nineteenth-Century Verviers An unusual set of documents allows us to study the details of household structure and dynamics in Verviers, Belgium. From 1806 to 1846, city officials compiled a nominative list of the population every year. These Relevés des habitons (Inventories of inhabitants) list name, occupation, place of birth, and marital status, as well as place of origin and date of arrival for those not born in Verviers. Children under age 12 are not listed by name, but are tallied with their parents. Periodically, the registers were indexed by name, and after 1828 many marginal notes were added to show deaths, out-migration, and locations in preceding and following registers. The registers refer to the population at the beginning of the year; however, internal evidence suggests that they usually took one or two months to compile. The Relevés des habitans are very similar to annual censuses, but the crossreferences to adjacent years add a continuity over time similar to population registers. In 1846, Verviers began using continuous population registers, which were introduced throughout Belgium in that year (Alter, 1988). Verviers has several advantages as a site for such a study. During the eighteenth century the city had become a major center for the production of woollen textiles. In early years this industry was organized along classic "protoindustrial" lines. Most of the

34

G. Alter, L. Cliggett, and A. Urbiel

spinning and weaving took place in peasant cottages in the surrounding rural areas, and finishing operations were performed in water-powered mills in the city. Verviers was the first place on the European continent to employ mechanical spinning machinery, and after 1800 the industry was rapidly transformed by steam-powered factories. In the first half of the nineteenth-century, both spinning and weaving operations were being drawn from the countryside into the city. Although most weaving was still done by hand, but it was increasingly organized into large workshops and factories (Lebrun, 1948; Desama, 1985; Gutmann, 1988). Although it had a distinctly urban character, Verviers remained a small city, which makes it much easier to study. Between 1831 and 1846 the population rose from 18,834 to 23,339, and peaked around 45,000 later in the century. It would be much more difficult to trace individuals in a larger city. In addition, the economies in all textile cities present certain common features, such as a predominance of female workers, that are familiar from previous studies in other places (Anderson, 1971; Tilly, 1979; Hareven, 1982). We drew data on the elderly in Verviers from the annual registers in both cross-section and longitudinal samples. We constructed a cross-sectional view of 1831 by selecting all households that included a person age 55 or older in that year. We then constructed a longitudinal sample by focusing on persons who were 60 to 65 in 1831. These individuals were followed forward in time through successive registers to 1846. Thus, the longitudinal sample consists of up to 16 observations on each individual. However, attrition from mortality, migration, and recording problems in the register reduces this sample from 661 individuals in 1831 to 87 in 1846. For each year of observation, we collected information on the entire household in which the sample person resides. The rich documentation available in the annual registers for Verviers has enabled us to extend our view of residential choices to include kin living in different households. We have attempted to identify all of the children of our sample individuals and to locate those children who were living in Verviers. Children who were coresiding with a person in the longitudinal sample were followed when they left the parental household. We also traced our sample backward in time to the register for 1821 to identify children who had left home before 1831. We have added information about the households of noncoresident children to the data base. We have also constructed map coordinates for all of the houses in the city, so that we can calculate the physical distance between these children and their parents. The high rate of attrition in a longitudinal sample, especially one focused on the elderly, raises a difficult question about counting the multiple observations on each individuals. If each observation is weighted equally, a person who was located in every year from 1831 to 1846 will be counted 16 times more often than one who died in 1831. Household structure changes very slowly, however, and multiple observations of the same individual cannot be considered independent. One approach is to weight individuals inversely to the number of times that they are observed. These weights can be constructed so that each individual will be counted equally. Unfortunately, this kind of weighting has the effect of emphasizing data from younger ages. Since attrition occurs as the sample moves forward in time, individuals who leave the sample are observed at younger ages only. An

Household Patterns of the Elderly

35

alternative approach would attach higher weights to older ages to compensate for attrition. We have attempted to steer a course between the dangers of giving too much weight to individuals who are observed more often, on one hand, and overweighting the younger ages, on the other. The observations on each individual were grouped into four age groups (60-64, 65-69, 70-74, and 75 and over), and weights were assigned inversely to the number of observations in each interval of ages. Thus, if an individual is observed twice in an age group, each observation is assigned a weight of one-half. The weights for each individual sum to one in each age group. 2 The unweighted and weighted numbers of observations in the longitudinal sample are presented in Table 1.

O l d A g e in V e r v i e r s , 1831 Tables 2 , 3 , and 4 provide an overall picture of the households of the elderly in Verviers in 1831 from the cross-sectional sample. Tables 2 and 3 show the percentages of persons who were coresiding with other types of individuals. These categories are not exclusive, so a person residing with a child might also live with other kin or nonkin. Several patterns are noteworthy in this table. First, we note that the percentages who lived alone were much lower than present patterns in developed countries. Solitary residence increased dramatically during the twentieth century (Kobrin, 1976). Most observers attribute this increase to rising incomes and the impact of pensions and social security programs. Steven Ruggles (1991), however, has recently disputed this interpretation, arguing for changes in preferences about coresidence with children. Second, elderly males were more likely than females to live with spouses. Half of the men age 70 to 74 were still living with wives, but only one woman in four resided with her husband. This reflects the higher mortality of males, the age differences between spouses, and higher proportions of women who never married. Third, the proportions of persons who resided with children were high, and these proportions remained high until after age 75. Overall, more than two thirds of persons 55 or older resided with at least one child. There is no clear difference in age patterns of coresidence with children between men and women. Widows, however, were more likely than widowers to reside with children. At younger ages coresident children are mostly unmarried. Married children become more common at older ages, but the proportion of women who resided with married children reaches only 35 % among the oldest women. There is also a strong difference between men and women in coresidence with married children. Women, especially widowed women, were the most likely to live with a married child. Fourth, the proportion of persons who lived with nonkin increased dramatically at older ages. In part, this reflects the residents in Verviers's municipal old age home, the Hospice des vieillards, which opened in 1668 (Verviers, Le Centre Public d'Aide Sociale de Verviers, 1989). This institution accommodated about 60 persons most of whom were above age 70. However, old people who were not in the old age home also frequently lived with nonkin. In contrast, other types of kin were not very important. About 10%

36

G. Alter, L. Cliggett, and A. Urbiel

of the elderly lived with siblings or other kin, when grandchildren and sons- and daughtersin-law are excluded. Table 4 looks at household structure in a different way. This table cross-classifies the person in the sample by relationship to head of household (head/spouse, parent, other) and by the presence of children in the household. Household headship helps to identify the direction of support in cases of coresidence with children. 3 It was not unusual in Verviers for newly married couples to live with parents. This arrangement was more common when the new couple was young or had borne a child before the marriage occurred, which suggests that they were receiving assistance from their parents (Alter, 1988, pp. 131-137). There seems to have been a strong preference among the elderly for retaining household headship. Most elderly persons, male and female, who lived with children continued to be listed as the head of household. After age 75 male and female patterns tend to diverge. The oldest women show a tendency to move into the households of married children, but this option remained uncommon for males.

Household Patterns in the Longitudinal Sample Table 5 can be used to compare the household patterns in the longitudinal sample to those in the 1831 cross-section (Table 4). The two samples are very similar in the proportions of persons who headed their own households, with and without children present. Two differences appear after age 70, however. First, persons in the longitudinal sample were less likely to live in households headed by others, primarily because few of the people in the longitudinal sample entered the old age home. We are doing a special study of the origins of persons in the old age home which may explain this difference. Second, older women were less likely to live with married children in the longitudinal sample than in the cross-section. It is possible that the cross-section includes old widows who moved into Verviers to live with children, a process that would not be captured by the longitudinal sample. Table 6 reveals the strong effect of family size on coresidence with children. The "number of children identified" is our closest approximation to total family size. We include all children who were found in the 16 years of the longitudinal sample and in the registers of the 1820s. Parents with very large families were nearly assured of having at least one child to live with, and these coresident children usually were unmarried. We also see the difference between older males and females that emerged in earlier tables. Fathers and mothers with smaller families were equally likely to live with unmarried children. But mothers with one or two children were much more likely to live with married children. Among parents with only one child, mothers were 13% more likely to live with a child than fathers, because mothers were more likely to live with married children. In Table 7 we add two other dimensions, household headship and marital status. Persons who had ever been married tended to retain household headship, and more than two thirds of currently married couples had at least one child living with them. Widows with one or two children were the only group that tended to move into the household of

Household Patterns of the Elderly

37

a married child. Daniel Scott Smith (1979) found similar patterns in the U.S. census of 1900, which reports the number of surviving children. However, U.S. widows seem to have been more likely to live with married children than were widows in early nineteenthcentury Verviers. This difference may lie in the earlier marriage and higher proportions marrying in the United States. In Table 8 we examine the spatial distances between parents and their closest child in a different household. This table considers whether parents who did not have children living in the same household were more likely to have children living nearby. In most cases, children in the "Locations unknown" category probably had migrated out of the city. Those who did live in the city were not likely to be far away. Verviers was a small place, and the greatest distance between any parent and child in the data set was only 1.2 kilometers. Nevertheless, there was a slight tendency for parents without coresident children to have children living in different households at the same address. 4 Table 9 considers the choices that parents made to live among different types of children. In this table we classify parents by the various combinations of married and unmarried children that we have identified. The issue that concerns us is whether parents showed a preference for living with unmarried children. Overall, they showed a clear preference for residence with unmarried children, but this effect is significantly weaker among widows. Elderly married couples almost always headed their own households, and their households almost always included an unmarried child when one was available. When the only child was married, however, the situation was more complex. Elderly married men usually (77%) maintained households separate from their married children. But about half of the elderly married women lived with a married child, and the head of household was as likely to be the parent as the child. This difference between married men and women may be due to differences in the health of their spouses. Since women were usually younger than their husbands, an older woman was more likely to have a spouse in poor health. Elderly widows and widowers display a somewhat different pattern. Here again, when we could find only unmarried children, they almost always lived in the household of the older generation. Widows and widowers, however, were likely to move into the household of a child, when married children were the only alternative. Finally, when both unmarried and married children were available, coresidence was the norm but household headship was less certain. The widowers headed their own households, which included a child 58% of the time. Widows, however, were only slightly more likely to continue to head a household that included a child (46.5%) than to move in with a married child (40.3%). In Table 10 we use the longitudinal sample in a different way to examine the relationship between household patterns and mortality and migration. This table shows the proportion of the sample who died (or migrated) in the next year by household composition at the beginning of the year. The surprising finding in this table is that parents who lived in households headed by a married child were more likely to die: While the overall probability of death was 6.3%, 9.9% of those living in a household headed by a married child died in the next year. The statistical significance of this finding was checked with a logistic regression that controlled for age, year, sex, and marital status. The logistic

38

G. Alter, L. Cliggett, and A. Urbiel

regression results indicate that a sample person living in the household of a married child was 2.2 times more likely to die. We interpret these higher death rates as evidence that these parents were in poor health before they moved in with their married children. Although parents preferred to maintain independent households, some of them apparently sought assistance from their married children shortly before their deaths.

Conclusions These data suggest that high fertility was a good strategy to provide for support in old age in nineteenth-century Europe. Late marriage and high proportions never marrying created a large pool of unmarried children who shared the households of their aging parents. Parents with more than two surviving children could be quite confident that at least one child would continue to live with them until they reached at least age 75. This does not mean that married children were unavailable for support, they were called upon in key situations. Parents tended to live with a married child when they had fewer surviving children, when the parent was widowed, or when their health deteriorated. However, when faced with a choice between living with an unmarried child or moving into the household of a married child, most parents chose to use the support of an unmarried child to remain heads of their own household. If the European marriage pattern worked to the benefit of parents, this was little comfort for those without children. Those who never married could seek the help of siblings or other kin, but the majority lived with nonkin or alone (See Table 3). In many cases the unmarried elderly and those without children must have been forced to rely upon the community for support. It is not surprising, therefore, that Verviers had both a municipal old age home and outdoor relief for the poor. It remains to be determined whether the needs of the elderly actually contributed to the late marriage and high proportions unmarried in the European past. Did parents put pressure on some children not to marry so that they could continue to draw upon their support?5 Oral histories collected by Tamara Hareven sometimes mention pressures placed upon a child who was selected to care for an aging parent (Hareven & Adams, 1991). The role of children in providing old age support may help to explain the persistence of late marriage in nineteenth-century cities like Verviers. Despite the predictions of pessimists, industrialization and urbanization in nineteenth-century Europe did not lead to extremely early marriages. Since factory workers had little hope of saving enough money to buy a farm or business, and their parents had little property to bequeath them, the motivations for late marriage proposed by Hajnal do not seem to be relevant. Perhaps the support demanded by parents from adult children played a role in continuing the European marriage pattern.

Household Patterns of the Elderly

39

Notes The research reported in this paper has been supported by a grant from the National Institute on Aging, PHS AG09331. The authors would also like to thank participants in the Duke Population Studies Seminar and the RAND Conference on Economic and Demographic Aspects of Intergenerational Relations, March 20-22,1992 for their comments. 1

Compare the references to the past in Shanas (1961, p. 1) to those in Friis, Townsend, and Shanas (1968, p. 4).

2

For example, a person who enters the sample at age 61 in 1831 and is followed until age 76 in 1846 will be observed 16 times: 4 times at ages 60-64, 5 times at ages 65-69, 5 times at ages 70-74, and twice at ages 75 and older. This person would be weighted as one-fourth for ages 61 to 64, one-fifth for ages 65 to 74, and one-half at ages 75 and 76. The total weight from all 16 observations would be four. In contrast, a person who was observed only at ages 61 and 62 would be assigned weights of one-half for each of those ages, but would have a total weight of only one.

3

Household headship is in part a cultural artifact of the persons who compiled the annual registers. The order of persons in the household follows a consistent pattern in which the head, usually male, and spouse are listed first, followed by their children in descending order by age, other kin, and nonkin. After a spouse's death, widows usually continue to be listed as heads of households, even when they have coresident adult sons.

4

The boundaries between households were frequently ambiguous, and sometimes a married child would be listed in the same household one year but as a separate household the next year. About 30 % of newly married couples in 1844 and 1845 were listed as separate households in the same house as a parent (Alter, 1988, pp. 136-137). The boundaries between households involve an interpretation by those who compiled the registers. However, we suspect that these boundaries may have corresponded to units of taxation.

5

In nineteenth-century Belgium, parents could easily frustrate the desires of children who wished to marry against their will. The Belgian Civil Code, derived from the Napoleonic Code, made it difficult to marry without parental consent. An adult whose parent refused to sign the marriage register was required to employ a notary to "respectfully" request the parent's permission to marry.

References Alter, G. (1984). Work and income in the family economy: Belgium, 1853 and 1891. Journal of Interdisciplinary History, 15, 255-276. —. (1988). Family and the Female Life Course: The Women of Verviers, Belgium, 1849-1880. Madison: University of Wisconsin Press. Anderson, M. (1971). Family Structure in Nineteenth-Century Lancashire. Cambridge: Cambridge University Press. Belgium, Commission du travail. (1887). Réponses au questionnaire concernante le travail industrial, vol. 1; and Procès-Verbaux des séances d'enquete concernante le travail industriel, vol. 2. Brussels: A. Lesigne. Desama, C. (1985). Population et révolution industrielle, Bibliothèque de la Faculté de Philosophie et Lettres de l'Université de Liège, Fascicule CCXLIII. Paris: Société d'Edition "Les Belles Lettres.

40

G. Alter, L. Cliggett, and A. Urbiel

Friis, H., Townsend, P., & Shanas, E. (1968). Old people in three industrial societies: An introduction. In E. Shanas, P. Townsend, D. Wederburn, H. Friis, P. Milhoj, & J. Stehouwer (Eds.), Old People in Three Industrial Societies (pp. 1-17). New York: Atherton Press. Gutmann, M. P. (1988). Toward the Modern Economy. New York: Alfred A. Knopf. Haines, M. R. (1979). Industrial work and the family life cycle, 1889-1890. In P. Uselding (Ed.), Research in Economic History, 4, 289-356. Hajnal, J. (1965). European marriage patterns in perspective. In D. V. Glass & D. E. C. Eversley (Eds.), Population and History (pp. 101-143). London: Edward Arnold. —. (1983). Two kinds of pre-industrial household formation. In Richard Wall (Ed.), Family Forms in Historic Europe (pp. 65-104). Cambridge: Cambridge University Press. Hareven, Τ. K. (1982). Family Time and Industrial Time: The Relationship Between the Family and Work in a New England Industrial Community. Cambridge: Cambridge University Press. Hareven, T. K. & Adams, K. (October 10-13, 1991). The middle generation: Cohort comparisons in assistance to aging parents in an American community. Paper presented at the Conference on Aging and Generational Relations, Newark, Delaware. Kobrin, F. E. (1976). The fall of household size and the rise of the primary individual in the United States. Demography 13, 127-138. Laslett, P. (1991). A Fresh Map of Life. Cambridge, MA: Harvard University Press. Laslett, P., with the assistance of Wall, R. (1972). Household and Family in Past Time. Cambridge: Cambridge University Press. Leboutte, R. (1990). Perception et mésure du vieillissement durant la transition démographique. Populations Agées et Revolution Grise, Chaire Quetelet 1986, Institut de Démographie, Université Catholique de Louvain (pp. 599-618). Louvain-la-Neuve: Editions Ciaco. Lebrun, P. (1948). L'industrie de la laine à Verviers pendant le XVIIIe et le début du XIXe siècle. Liège: Bibliothèque de la faculté de philosophie et lettres de l'Université de Liège, Fascicule 114. Rotella, E. & Alter, G. (August 20-24, 1990). Working class debt in the late nineteenth- century. Tenth International Economic History Congres, Leuven, Belgium,. Ruggles, S. (October 10-13, 1991). Living arrangements of the elderly in America, 1880-1980. Paper presented at the Conference on Aging and Generational Relations, Newark, Delaware. Shanas, E. (1961). Family relationships of older people. Health Information Foundation Research Series, No. 20. New York: Health Information Foundation. Shorter, E. (1977). The Making of the Modern Family. New York: Basic Books. Smelser, N. J. (1959). Social Change in the Industrial Revolution: An Application of Theory to the British Cotton Industry. Chicago: University of Chicago Press. Smith, D. S. (1979). Life course, norms, and the family system of older Americans in 1900. Journal of Family History, 4, 285-298. Tilly, L. A. (1979). Individual lives and family strategies in the French proletariat. Journal of Family History, 4, 137-152. Verviers, Le Centre Public d'Aide Sociale de Verviers. (1989). Des origines à nos jours.... Verviers.

Household Patterns of the Elderly

41

Table 1: Unweighted and Weighted Number of Observations by Age, Verviers Longitudinal Sample, 1831-1846 Unweighted

Weighted

Number

Percent

Number

Percent

60-64 65-69 70-74 75 +

1448 1730 911 416

32.1 38.4 20.2 09.2

554 460 227 127

40.5 33.6 16.6 09.3

Total

4505

100.0

1368

100.0

Age

42

G. Alter, L. Cliggett, and A. Urbiel

Table 2: Percent of Persons 55 or Older Living with Selected Types of Other Individuals, by Sex and Age, Verviers, 1831

Female

Percent living with: 1. Spouse 2. One or more children— a. Any child b. Never-married child c. Currently-married child d. Widowed child 3. Other kin— a. Excluding spouse/children b. Excluding spouse, children, sons-, daughters-in-law, grandchildren 4. Nonkin 5. Alone Number (N)

Age 55-59

Age 60-64

Age 65-69

Age 70-74

Age 75-79

Age 80-84

Age 85 +

57.3

45.7

39.7

25.7

18.9

08.2

07.0

70.7 68.3 09.2 01.3

64.2 56.8 13.6 02.8

65.7 53.9 15.2 04.0

58.8 40.5 20.9 05.4

48.9 25.6 27.8 04.4

59.2 30.6 28.6 02.0

46.5 09.3 34.9 04.7

24.0

28.1

26.0

32.4

36.7

40.8

39.5

11.3 25.6 03.2

10.5 29.3 06.2

05.9 22.5 08.3

07.4 37.2 03.4

06.7 36.7 05.6

08.2 40.8 10.2

04.7 62.8 04.7

379

324

204

148

90

49

43

Male Age 55-59

Age 60-64

Age 65-69

Age 70-74

Age 75-79

Age 80-84

Age 85 +

1. Spouse 2. One or more children— a. Any child b. Never-married child c. Currently-married child d. Widowed child 3. Other kin— a. Excluding spouse/children b. Excluding spouse, children, sons-, daughters-in-law, grandchildren 4. Nonkin 5. Alone

75.3

74.1

57.5

57.0

40.3

40.0

40.7

72.6 71.2 05.3 00.9

67.3 65.4 06.1 00.8

58.0 53.6 08.2 01.9

66.9 58.7 14.9 01.7

47.2 38.9 15.3 02.8

37.1 25.7 05.7 05.7

40.7 33.3 07.4 03.7

19.7

20.5

22.7

29.8

27.8

28.6

25.9

11.5 21.8 02.9

11.4 25.1 03.0

09.7 30.9 05.8

11.6 24.0 04.1

09.7 37.5 08.3

17.1 34.3 02.9

14.8 51.9 07.4

Number (N)

340

263

207

121

72

35

27

Percent living with:

Household Patterns of the Elderly

Os

O •«t O

•a I

•i· Ν r- e- o

o

r» o

r- m σ\ -Η

«1 -H

m en l· oo οί οό >o it ΐ ^ Ν O

«

VD O

Ο Tf « Ν ΟΟ ΟΟ f . \0 O O

W> f i 00 r^ ι—' ö o o o

ts Tt

00

O ΓΛ (N O

Tim

44

G. Alter, L. Cliggett, and A. Urbiel

Table 4: Percent of Persons 55 and Older by Type of Household, Age, and Sex, Verviers, Belgium, 1831

Female

All

ΑΠ

Age 55-59

Age 60-64

Age 65-69

Age 70-74

Age 75-79

Age 80-84

Age 85 +

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

20.0 52.4

17.9 66.2

19.1 54.9

26.0 52.0

18.2 45.3

27.8 25.6

16.3 32.7

11.6 16.3

01.1 07.4

02.1

01.2 06.5

02.0 07.8

02.0 08.1

01.1 15.6

04.1 18.4

25.6

16.0 03.1

11.3 02.4

16.4 01.9

08.3 03.9

23.0 03.4

23.3 06.7

24.5 04.1

41.9 04.7

Headno child child present Parentunmarried child married child Otherno child child present

Male

All

AU

Age 55-59

Age 60-64

Age 65-69

Age 70-74

Age 75-79

Age 80-84

Age 85 +

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

20.9 60.7

18.5 70.3

23.2 65.0

21.7 52.7

16.5 62.0

25.0 41.7

28.6 31.4

22.2 40.7

00.5 01.6

00.3 01.2

— 00.8

01.9 01.9

— 02.5

— 04.2

— 02.9

— -

14.8 01.5

08.8 00.9

09.5 01.5

19.8 01.9

16.5 02.5

27.8 01.4

34.3 02.9

37.0 -

Headno child child present Parent— unmarried child married child Otherno child child present

Household Patterns of the Elderly

45

Table 5: Percent of Persons by Type of Household and Age, Verviers Longitudinal Sample, 1831-1846

Female

All Headno child child present Parentunmarried child married child Otherno child child present

AU

Age 60-64

Age 65-69

Age 70-74

Age 75 +

100.0

100.0

100.0

100.0

100.0

26.2 51.4

21.7 53.6

26.7 50.1

31.0 52.2

35.7 44.9

00.9 06.0

01.0 06.4

00.5 07.1

01.2 04.3

01.9 03.3

13.2 02.2

14.7 02.6

12.5 03.1

11.1 00.2

12.8 01.4

Male

All Headno child child present Parentunmarried child married child Otherno child child present

All

Age 60-64

Age 65-69

Age 70-74

Age 75-79

100.0

100.0

100.0

100.0

100.0

28.7 58.0

25.0 62.3

25.3 60.8

35.3 50.0

44.8 43.6

00.8 01.7

00.2 01.4

00.9 01.4

01.3 03.3

01.8 00.6

09.7 01.2

09.8 01.3

10.4 01.0

08.9 01.3

08.2 00.9

G. Alter, L. Cliggett, and A. Urbiel

46

Table 6: Percent of Persons Living with Children by Number of Children Identified, Verviers Longitudinal Sample, 1831-1846 Number of Children Identified All

0

1

2

3

4

5

6+

0.606 0.523 0.113 0.032

-

0.757 0.524 0.217 0.016

0.797 0.714 0.156 0.042

0.822 0.768 0.088 0.051

0.910 0.843 0.183 0.081

0.955 0.955 0.015 0.072

1.000 1.000 0.108 0.073

0.616 0.590 0.039 0.019

-

0.625 0.555 0.067 0.003

0.760 0.739 0.038 0.006

0.857 0.824 0.040 0.071

0.881 0.828 0.110 0.007

0.946 0.946 0.007 0.028

0.974 0.974 0.043 0.031

Female: Any child Never-married child Currently-married child Widowed child Male: Any child Never-married child Currently-married child Widowed child

47

Household Patterns of the Elderly Table 7: Percent of Persons by Type of Household, by Number of Children Identified, Sex, and Marital Status, Verviers Longitudinal Sample, 1831-1846

Number of Children Identified All

0

1

2

3

4

5

6+

27.5 68.7

90.2 —

19.0 74.3

19.2 78.8

15.3 82.5

05.0 95.0

100.0

100.0

00.1 01.4

— —

— 06.7

00.3 00.3

02.1 00.3

09.8 —



01.3 00.0

02.2





-

21.5 48.9

52.3 -

19.9 45.4

15.6 55.7

15.0 68.6

12.2 68.7

87.2

96.8

01.9 11.6

— —

01.6 22.3

03.2 14.4

01.5 04.9

05.3 11.0

— —

-

11.7 04.4

47.7 -

04.8 06.0

04.4 06.7

05.1 04.9

— 02.7

10.9 01.8

03.2

30.6 66.3

95.5 —

33.4 60.2

24.1 73.3

13.6 82.5

15.3 84.7

06.0 94.0

03.9 95.4

00.4 00.5

— -

— 01.5

00.4 00.2

01.3 01.3

— -

— -

00.7 -

01.8 00.3

04.5 —

04.9 —

01.7 00.3

01.3







Female. married: Headno child child present Parentunmarried child married child Other no child child present Female, widowed: Headno child child present Parentunmarried child married child Otherno child child present Male, married: Headno child child present Parentunmarried child married child Otherno child child present

continued on next page.

G. Alter, L. Cliggett, and A. Urbiel

48 Table 7, continued...

Number of Children Identified All

0

1

2

3

4

5

6+

18.1 50.6

37.1 -

24.3 43.2

10.8 76.7

12.4 63.3

67.8

100.0

100.0

02.0 05.1

— -

— 11.9

— 00.6

10.7 00.8

03.2 29.0

— -



20.5 03.8

62.9 —

11.5 09.1

08.7 03.2

03.8 09.1

-

Male, widowed: Headno child child present Parentunmarried child married child Other no child child present

-

-

Household Patterns of the Elderly

49

Table 8: Percent of Persons by Distance to Nearest Nonresident Child by Number of Coresident Children by Sex, Parents with at Least One Nonresident Child, Verviers Longitudinal Sample, 1831-1846

Number of Coresident Children Female

Male

None

1+

None

1+

13.9 08.5 22.8 04.4 50.4

10.6 09.9 17.4 03.4 58.7

13.5 07.2 19.1 02.6 57.5

06.4 07.3 14.0 06.6 65.7

100%

100%

100%

100%

Closest Coresident C h i l d Same address ?

Ο)

I I

Ξ

s

α o •o

I I

1

I I

s os

s *

δ

I



68

m

§

S

S ®

Η o' — oo

1Λ o

o (Ν

68

O

νο 00* σΝ m



m

I

ι «s o

I I I

£

m O ov

I I

00

I I

Tf



S

=>

i

si 1

68

8

Ό ο

o

\0

Tí- -Η Ο (Ν

m νο

I I

β •g a υ ο Ζ

ι I

Ή τ*

8 S

I I

ο\

ο

IN O

68

31 8

vi σ\ 00 CS Tf

σ\ O

ci 00 ο ο

α o

Household Patterns of the Elderly

i

8

I I

o o

0\ ©

•ο t r-'

o\

o o «S vi

I I

I I

ο σ>

I I

I I

Ό SÌ

ê

I I

ΓÖ

o\ Cl

I I

I I

I SR

8

4

68

Ò ο Ή

-t

* * *

en 00 no o

O env V 00") oo o en

m oo es m S S

2S o

es w-i vo

es oTI en

*

ω

o

* *

1—1 oOv 00 o On en 1 es NO es

+

**

en en N esO tes en

*

es —r-— en es es es es

ej\ S tv> es v>

+

vo en

8

es ttes

Living Arrangements of the Elderly in America: 1880-1980

* *

Ό Ι— VC Ο

ω c/j

269

*

ο es *η •t

ο δ ο

8 νο ο

ml

00 Ο Ο en τ-Η tι-Η es

*

oo rcs Μ

VO Tt Vj 00 •t t-

rTf σν Vi

\o en es o

00

en en

σ\ o* νο es

es

(N vo t-



o T-H

ON

ω CO

031

ω

COI ο

ι es , ζ ç f s ω > •ag g > « .ζΛ . a C S Ρ " E i 60 Γρ "5b ' 6 0 'oo CO 4H o « £ (4 t O y y y o >

O >

o H

3 o H

C4

1

o >

o >

•ì

s1 a1 a1 > '60

£ '60

S; '60

Ο y ο >

Ο >

ο >

3o 3 3 3 3 3 3 3 3 3 3 3 o o o o O ο ο ο ο ο ο H

H

H

H

H

H

Η

Η

Η

Η

Η

Η

D. Hogan, D. Eggebeen, and S. Snaith

340

** VO 00 >n es en

5 o



ΡS O3h

o VO en o Ö o

o

Ö

Ϊ I ^ CL,

Ses O o ö o

eS

o

~ β O tu ζ s α.

00 ηo

VO en VO o o ö

en

"o >

"3 >

1 H

1 [2

1 H

1 [2

3

3

S

S

'S 'S

S a»

s

I υ "

.-3 [Λ

«o

S o EC

eo ·α •ωs S

O1 ?

Ί25 1 O. 9 d t/5

> "βο

>O "O

S; 'so

' >δ

1 s1 'ο a

o> 3o Η

o>

o> 3 ¡2

ο> 3ο Η

i "o >

o>

o>

C/3 •a 1Η '8 co

1 H

1 H

3o H

•ao

¡§

Κ

w

•a υ £ I

g o

tí ·§

«

'S

H

•ao Η

1 s>1 'o 8

o> 3o Η

The Well-Being of Aging Americans With Very Old Parents

345

Ό •Π

O

a

•a

λg I»

§

73

•a •s Sf H (N O ¿í C > Ö o ^ o a '-a -3 Q 'Sa

I

α o

§

1o i

I •s ε §

o

o ιo V D ÍrNC 00 V£> O ο O ι 9 o1 Ö ö —' 9

I I I I I I I I I

•5 > — V νft*

-a •a

^ l a u

399

The Demography of Family Care for the Elderly

I ι

"g

31 o Ht

Tt NO »-H es Tt Ξ CS

o31

Q If] IO M t^ ^CS v>. Tt tfi Ol C W S ΙΛ Tt

V> σν 8 « VO

h h O ri h O t~v> o — (S Tf

ool om \

σ\ g £ oo O s o ^ 5

S?

o\ en Tt f- Tt ¡G « N vi r» 00 csO vi

ni v>

V> Tt V> t- Tt O Ό tTt CS en η

ONI ON

t n »iv>- _· m· M w.· O _· — Ν C 0S \ ^CS . Tt

CT>

e s 8

V3

~

.Ό O 1 a ü L •s g s g •S 1 ! σ\ CS

o v> tí" vd O vo

en τι·' o\

VO vd

O O C

"t CO

α o 3

en p-°

Μ "Λ (Ν O 8

"O od

•a

CS σ\ 00

•t -t VD O o

t^cJ

C/2 S è2

00 σ\

•ψ 00 ci vi o O

röt'-

σ>

5 •a H C/3 — ρ. δ

T-H •ri

t- es τ-Η en O o

en vi \o

α o 3

CS"

en 00 TÍ O o

vi CS

•a

o

on i (2 •a •a S M Ρ ò2

o

VO vi 0\

T) σ. ri o o

ov vi VO

CS \o o\

(Ν (S o O

en vd en o

en en 00

vd es

00 vo vi ro vo

σ > 00

VO o es VO vd o vo

o es' v>

en •t' cn

00 t^ 00

00 V> V> o o

«S r-

V) 00 «S Ov vd o en

rcs

σ\ m ot·»' S

es es

es vo vi en es vi

Ο od vo

v> VO v> Tt t- •Or O r» σ\

«S

es ren — r»

00 vd 00

en od en

cs rV> es" σν o•«t t-'

o σ!

O o CS öVO

vo oò

en T-H en »ri VI

o es es

00 es 00 00° σ\ CS Ti- VO

ο C ì/> C

t

> é

9 c/5

g H

Ο Q

i l « Si i l Ϊ ΌωΌ" •δ «S δω £ .δ ·δ ω C " do uo ,ο C C α "9 o. a ο ε ε '6 Ιh O

00

>£>

σ\

C *o ase 3* C/J J

o IE



•o υ O g; ' α ö (j> o n G 1 oS* UÌ §* CÄ Ό co υ < 4>υ • o •7 Q

o •o

s

α 0

•o U Iai O •o o> Ό

'tí

1»H OH

res Γ»



I e I

a

o\

en es

α o Ζ

ε o 3 S S 3

te a

Sí "r! •δ .δ 'li 'q e e S g· g-s C/3 Vi

aa

α 0 1&

Types of Support for the Aged and Their Providers in Taiwan

•3 O

σ\ es OD

\o vi i-Η o

00 Γ-'

C/5 i—.

oo t O en

o

CO Β

•s f o »«>Ν > .s» a» u S u a s S ì α e e O S 8.8,

J

— ο o —•

Tf en 00 σ> en vi ϊ-Η r-' o o

00

CS o Ό vi en r-' Tf m 8 O

x> en r» vo o o o vi ri vo >T1¡ l'- es o

o o

o

888

vo 00 00 τί 00 vi en 5 es

Vi

en o oo 0\ r»' en en

S—

en es

Vi

T-H en 00 00 00 Ò r-' rn rt

1-

00

rt

es

·

TÍ es es "O en es

s

00

en en O o >o en en en en 8 O

oN
ll;