Preferred Citation: Kertzer, David I., and Peter Laslett, editors Aging in the Past: Demography, Society, and Old Age. Berkeley:  University of California Press,  c1995 1995. http://ark.cdlib.org/ark:/13030/ft096n99tf/


cover

Aging in the Past

Demography, Society, and Old Age

Edited By
David I. Kertzer
Peter Laslett

UNIVERSITY OF CALIFORNIA PRESS
Berkeley · Los Angeles · Oxford
© 1995 The Regents of the University of California

To Matilda White Riley,
an inspiration to us all



Preferred Citation: Kertzer, David I., and Peter Laslett, editors Aging in the Past: Demography, Society, and Old Age. Berkeley:  University of California Press,  c1995 1995. http://ark.cdlib.org/ark:/13030/ft096n99tf/

To Matilda White Riley,
an inspiration to us all

PREFACE

We present this volume to our readers in the confidence that they will acknowledge the importance of its subject and that they will accept the description of its content in the title of the first chapter, "Necessary Knowledge," necessary knowledge from the past. Old age and old people are now a commonplace theme in all Western and industrialized countries, which is understandable as people realize that their populations are exceedingly old and getting older. They are well aware of the implications for transfers between age groups and for generational justice, especially in the United States where the crises in welfare and health costs are persistent themes in the media and among politicians and administrators. Every such discussion begins with a rigmarole about aging, the facts being variously selected and interpreted, seldom with much accuracy or understanding and frequently in a way that misleads. Perpetually and inevitably, there is a lack of historical depth, with consequent obfuscation.

The attempt here is to fill out the historical horizon and to make use of it to clear away some of the misapprehensions. This is done in the recognition that we cannot expect to understand ourselves as we now are in the industrialized countries—and what we shall become—unless we also understand what we have been. It is not only true, as a case in point, that Western and Japanese populations are very, very old. It is also true that they are the oldest human populations that have ever existed and that they never will be young again. Indeed, all the other populations in the world will join them in their elderly condition and are beginning to do so already. Our hope is that this book may help those in the nonindustrial world take cognizance of what they are and of what they may become.

Although we shall claim that this is the first work of its kind to be published, it is far from the first book to be published on the history of aging.


x

In recent years such efforts have challenged many of the established assumptions about aging and enriched our understanding of how the nature of old age has changed through the centuries. In spite of these advances, our ignorance is still formidable. Meanwhile, the strictly demographic study of aging has been making progress, if not enough to satisfy us, and the historical side of that study has been going forward.

But historical demography, like demography itself, has been preoccupied for the most part with fertility and mortality. Lacking so far is a conjunction of the history of age and the old, that is to say, how we have become old, with its historical demography, why we have become old. Such is the need that is addressed in this volume, with the addition of some elements—rather few, because this is the most difficult and least cultivated part of the terrain—of comparative analysis. The significant contrast is between the condition of developed societies when they were younger and the condition of the developing nations of today.

In the discussions that have gone on between us as coeditors, we have been impressed with the intellectual vistas that open up once we adopt a viewpoint from the perspective of later life, rather than of youth, which is seemingly the natural and inevitable vantage point of our civilization. Like speleologists breaking for the first time into the ample space of a cave, we are overwhelmed by entirely unexplored avenues leading off in various directions in the crepuscular light.

But familiar territory also begins to look different. Such is the area of the history and the development, in relation to aging, of the co-resident domestic group. In his final chapter, David Kertzer is able to limn the contours of a new definition of arguably the most important, because the most widespread, form of that group, the nuclear family household. The informed reader may recognize that this new interpretation is scarcely consonant with the published views of Peter Laslett on the nuclear family household.

We hope that this difference of interpretation will add to the intellectual interest of our contributions to the volume. How long will it be before the study of aging loses that air of grayness and tedium that has always hung about everything to do with becoming old and is instead recognized as an arena of brilliant intellectual opportunity?

References are made in the introductory chapter to the collected volume published two decades ago, Household and family in past time (1972). There it was confessed that deliberate aim was being taken at the opening up of a new field of inquiry, and so it proved in the sequel. It has also transpired that the rigorous standards of data collection and analysis that were laid out at length in the introduction to that volume have been progressively neglected. So much has this been so that the controversies that have arisen about family and household at various times and at various places in the world seem to have become interchanges about approximations and im-


xi

pressions rather than about rigorously examined entities. The findings of family history are to that extent less significant or even trustworthy. It is fervently to be hoped that this will not come about in the field of the history and historical demography of aging.

This book is the product of our efforts to stimulate interest in these pursuits and to suggest the kinds of insights that can be obtained by paying more attention to demographic issues in conducting historical research on the lives of older people. Toward this end, a conference on the historical demography of aging was held at Bowdoin College Breckinridge Public Affairs Center in York, Maine, in spring 1990. Among the participants were historians, anthropologists, sociologists, economists, and demographers from several countries.

Our interest was in the interaction between demographic events and processes, on the one hand, and both societal age norms and individual-level behavior involving older people, on the other. While the emphasis was on demographic forces, these were examined in interaction with other relevant influences: economic, political, social, and cultural. The focus was on the West, both Europe and North America.

Of the thirteen chapters that follow, ten are revisions of papers first given at that conference. The focus is on a number of Western societies, including Britain, Italy, Hungary, and the Baltic states, though the United States receives the most attention. In his contribution, Gene Hammel provides a demonstration of the role to be played by microsimulation techniques in advancing the historical demography of aging, as he takes on the classic question of the prevalence of the stem family in the past.

In addition to those whose work is found in this volume, a number of other scholars participated in the Breckinridge conference. Their contributions have enriched these pages. We should particularly like to thank Matilda White Riley, Arthur Imhof, Tom Ericsson, Maris Vinovskis, Pier Paolo Viazzo, Richard Suzman, Timothy Guinnane, Jack Riley, Lee Craig, and Robert Whaples.

The conference was made possible by a grant (1R13 AG08429-01) from the National Institute on Aging (NIA). We would like to acknowledge the important role of Richard Suzman of the NIA in encouraging and nurturing this project. Further support for manuscript preparation was provided by the NIA supplement to Brown University's Population Studies and Training Center grant (NICHD P30 HD28251). We would also like to thank Bowdoin College and the staff of the Breckinridge Public Affairs Center for providing us with an idyllic setting for the conference. A final note of thanks is owed to Stanley Holwitz, of the University of California Press, for his continued encouragement and faith in this project.


xiii

CONTRIBUTORS

George Alter is Professor of History at the University of Indiana, Bloomington.

Rudolf Andorka is Rector of the University of Economic Sciences, Budapest.

Allen C. Goodman is Professor of Economics at Wayne State University.

Myron P. Gutmann is Professor of History at the University of Texas, Austin.

Michael R. Haines is Banff Vintners Professor of Economics at Colgate University.

E. A. Hammel is Professor Emeritus of Anthropology and Demography at the University of California, Berkeley.

Tamara K. Hareven is Unidel Professor of Family Studies and History at the University of Delaware.

Nancy Karweit is Senior Research Associate, Center for the Study of the Social Organization of Schools, Johns Hopkins University.

David I. Kertzer is Paul Dupee University Professor of Social Science and Professor of Anthropology and History at Brown University.

Peter Laslett is a Fellow of Trinity College, Cambridge, and co-founder of the Cambridge Group for the History of Population and Social Structure.

Andrejs Plakans is Professor of History at Iowa State University.

Roger L. Ransom is Professor of History and Economics at the University of California, Riverside.


xiv

Daniel Scott Smith is Professor of History at the University of Illinois, Chicago.

Richard Sutch is Professor of Economics and History and Director of the Institute of Business and Economic Research at the University of California, Berkeley.

Peter Uhlenberg is Professor of Sociology at the University of North Carolina, Chapel Hill.

Richard Wall is the senior researcher at the Cambridge Group for the History of Population and Social Structure.

Charles Wetherell is Professor of History at the University of California, Riverside.


1

PART ONE
INTRODUCTION


3

One
Necessary Knowledge: Age and Aging in the Societies of the Past

Peter Laslett

For our purposes here, necessary knowledge, that is, necessary knowledge from the past, is a body of information that all persons must have to understand themselves as they are today. The historian, especially the historical sociologist, might claim that in theory at least the whole of history is relevant to present people and might demand a demonstration of how far age and aging should be accorded priority over other things that present people ought to be aware of. Such a demonstration is clearly required in the introduction to a book about aging in the past.

Why Do People Have to Know About Aging in the Past?

The populations of developed societies have grown old at an amazing pace. Within the last hundred years, and to a considerable degree within the last fifty years, the populations of Europe, North America, Australasia, and Japan have become far and away the oldest human populations of which we have knowledge. These populations are older—and still getting older—in two important senses: average individual lifetimes last for very much longer than they ever have before anywhere or at any time, and these populations have among them quite unprecedented numbers of elderly people. It is obvious that the situation could not always have been like this, and during the last century some European populations were getting younger. In England, for example, proportions of elderly people were decreasing slightly between 1800 and 1830. A knowledge of aging in the past, that is to say, the historical demography of aging, is evidently necessary in order to recognize how novel the situation now is in the advanced countries. There is no other way of grasping that fact.


4

I shall maintain that aging has been so sudden that there has not yet been time enough to take account of the transformation. Contemporary Europeans and citizens of societies of European origin must, along with the Japanese, be living in a state of cultural lag, even of false consciousness, at least to some degree. They are continuing to make assumptions about age and aging that, though they had always been true before the present century, have incontinently disappeared. They have to be brought up against the facts of aging as they always used to be and compare them with the very different facts as they now are.

The information produced by the historical demography of aging is therefore first and foremost knowledge with a view to ourselves. But the developed societies are not the only ones to be getting older in the last decade of the twentieth century. All contemporary populations are aging, including those classed as less developed or even "primitive." Young as they still are, sooner or later they will join the industrialized countries in their age-transformed condition. We shall illustrate this below for some Southeast Asian countries.

Nevertheless, the developed nations are much farther along the road. Except for Japan, they have spent a much longer time during the aging process than these other nations will, even though the change in the West has had to be described as so sudden and unexpected. It is the advanced countries that have to learn to adapt themselves, to modify their outlooks, to found the new institutions, to bring forth and develop the ideas and attitudes that are now for the first time required and that will be required into the indefinite future. At the present moment, they are in a situation not unlike that of Western Europe, especially of Britain, at the dawn of industrialization, the first to adventure into an entirely new world. The history of aging in the now developed countries has a significance for Brazil, shall we say, or India or China, analogous to that of the economic history of those developed countries, and the analogy is the more telling because of the fundamental similarity of demographic processes wherever and whenever they occur. The sooner the cultural lag or false consciousness, as I have called it, is made to disappear from the countries that already have age-transformed populations, the better for the world as a whole.[1] The first requirement is that the facts should be known, the historical facts in particular.

There is another reason why the historical demography of aging is highly significant. Together with gender, ethnicity, and class, aging is one of the four dimensions of individual and social experience, though it has hitherto been given much less attention than the other three. It has to be known how the age composition of societies has changed over time, along with the longevity and the life course of individuals, if a properly historical social analysis, a durational analysis, is to be undertaken. Until this is done, we shall not fully comprehend the societies in which our ancestors lived, nor


5

shall we be able to grasp the contrast between their situation and our own. This means that we shall not be able to understand ourselves, to understand ourselves in time, to place ourselves in the procession, as I shall finally put it.

Remarkable as the aging of the populations of advanced societies has been, the phenomenon is rather more complex than it may seem and must be viewed in proper proportion. It is now frequently said that in a country like Britain, longevity has doubled and relative numbers of the old have trebled within the hundred years preceding our own day and that this has mostly happened within the lifetimes of its older citizens. As the figures in table 1.3 will demonstrate, however, this is somewhat of an overstatement as to changes in the two variables, though the historical period during which the change has occurred is correct.

In England the gain in expectation of life at birth over that period has been more like two-thirds for men and just slightly more for women, with a multiplication of some two and a half times for men and somewhat under three times for women in proportions over 60. [2] When the more revealing measures that are suggested below in the section "Experiential Measures in the Historical Demography of Aging" are applied, the change has been somewhat less than this in England with respect to longevity. Length of life after the fifteenth birthday, the suggested statistic for general comparison over periods of time, has not doubled but increased by more like a half, and at age 50, more appropriate for comparisons of this kind for those in later life, the rise is about one-half for men and two-thirds for women in Britain since the 1880s. If the share of those over age 60 of all those over age 25 is reckoned, the realistic experiential measure suggested for the weight in the population of senior people, there has been a twofold rather than a threefold gain, though rather less for men.

We need not exaggerate these historic changes, therefore, and we should be prepared to face the complications that arise when deciding on the most informative ways of describing them. In making out a case for the great importance of the history of aging, moreover, we must take due account of other drastic and unprecedented demographic changes that have recently taken place in Western countries and in Japan as well, though not to the same extent. The extraordinary increases in divorce, in births outside marriage, in numbers of persons living alone, and in consensual unions have gone forward over the last twenty or twenty-five years alongside the process of aging and have had their own individual and collective effects on social life. The combination of these developments with aging has made the position of the Western countries in the 1990s singular indeed. But these other alterations in behavior, though frequently dramatic, have been more recent and rather less general than aging, showing abrupt vicissitudes and varying from country to country.[3] Their occurrence, in my view, does nothing to modify the claim that the historical demography of aging represents necessary knowledge.


6

This is the first volume, as far as we know, ever to be devoted to the historical demography of aging. The introductory chapter accordingly begins with a discussion of it as a subject itself and an appreciation of the extent and reliability of its results. We go on to describe and discuss the course of aging in developed societies, under the heading "The Secular Shift in Aging," a shift from a lower plateau to a higher plateau that has taken place, as has been seen, for the most part during the present century. The two temporal areas will alternatively be referred to as the Before and the After.[4] We will insist that longevity, that is, expectation of life at birth and at later ages, is as important a constituent of the history of aging as proportions of elderly, though not generally recognized as such by demographers. New indicators will be suggested to supplement those in use at present, which are ineffective for historical purposes and removed from experience, experience in our own day as well as in the past.

No apology will be offered for giving so much space to these indicators and to the discussion of definitions, conventions, and measures for this newly emerging study. When in the early 1970s a book of this type was issued, opening up the then-novel field of the history of household and family (Laslett and Wall, Household and family in past time , with a comprehensive introduction by Laslett), much space had to be used in this way.

The Historical Demography of Aging

Historical demography proper began its career fairly recently, in the 1960s. Statistics for age and aging have always been among its results, but it is only during the last year or two that much attention has been paid to them, and the information so far available is rather restricted. Scarcity of data imposes stringent limits on what can be accomplished by historical demography in any field. Fertility and mortality, both essential to the reconstruction of the history, of aging, can only be worked out for societies that have recorded births, marriages, and deaths (or in Christian areas, baptisms, marriages, and burials), societies whose records of this kind are still extant and available for study. Migration, often critical in the study of age composition, generally requires information additional to the registration of vital events. The fact that these are preconditions of satisfactory analysis means that we shall never have anything like precise numerical knowledge of aging in wholly illiterate societies, past or present, or in literate communities that have failed to carry out the necessary recordings and to leave them to us for study.

Hence accurate estimates of life expectation and age composition even in highly civilized earlier populations like those of Greece or Rome, or in any European population before the end of the Middle Ages, will always elude us. Along with them go the peoples of the whole of the rest of the world in the premodern past, though here and there a tiny pool of uncer-


7

tain light glimmers in the darkness, nearly always fitfully illuminating groups of elite individuals.

It has to be said that this enforced ignorance makes it difficult for us to observe at all closely any population that could be supposed to have the aging characteristics of a wholly traditional society of the preindustrial kind existing on the lower aging plateau in a wholly traditional world. It might be thought that the developing societies, especially the "primitive" ones, surviving in our own day would provide just what we would like to have, instances of something like "natural" aging. This is scarcely the case, however, because these contemporary societies, if they carry out the required registrations at all, do so at the behest of literate minorities whose very presence may alter their aging characteristics to some degree. Moreover, such "backward" communities exist in a world so dominated by the highly industrialized countries that their longevity and age structure are hardly likely to be at the traditional level.[5] We are in the same position as those who study birthrates, and our best chance of estimating a "natural" standard for aging lies in reconstructing the demographic history of the one or two countries that did maintain usable recordings for the whole of their populations during at least some stretches of time before the demographic transition and the secular shift. Since they have now become industrialized themselves, however, and since their material standards were probably already higher than those of the rest of humanity at the periods in the Before for which their situations with regard to aging can be recovered, the evidence of these countries is even less suited to show a "natural" aging condition than is the case with "natural" fertility.

Archaeodemography, the establishment of a general notion of vital rates through the examination of exhumed skeletal remains, has sometimes been used to prompt the record of the historical demography of aging (e.g., Laslett [1976] 1985: table 2a). Through this technique, indications of ages at death, of the life span, and of gender differentiation over the life course have been recovered; life tables have even been constructed. This evidence is all that is open to biologists of aging who have to interest themselves in the very distant past, distant enough to allow for natural selection. Historians concerned with much later periods have contrasted its outcomes with those derived from other sources with rather discouraging consequences. Although archaeodemography may serve for approximate aging estimates, the reckoning of limiting values for the most part, the sketchiness of its re-suits, and questions as to their accuracy and representativeness prevent it from being a source of much importance for the historical demography of aging as it will be expounded in this volume.[6]

Studies of what could be termed the history of aging as distinct from the historical demography of aging are subject to somewhat the same judgment. Where particular written records have survived from the prestatistical era


8

which can be manipulated for the purposes of demography, though not created to that end, they can yield some estimates of age. Conspicuous examples are Ulpian's life table (3d century A.D. ) or the inquisitiones post-mortem in medieval England (inquiries at a landholder's death about his properties and claimants to them). The first appears in many works on mortality and life expectancy not concerned with history and the second, along with the evidence of poll taxes, in a comprehensive historical study such as J. C. Russell's British medieval population (1948) and in many more limited analyses.[7]

When proper allowance is made for their limitations, biases, and inaccuracies, materials of this kind are of considerable use to the historical demography of aging. They could not be said, however, themselves to constitute that study over tracts of time in which nothing more systematic, interconnected, and informative is, or can be, forthcoming. Every detail pertaining to aging in past time is important to this book. But if nothing of greater value to our purposes could be recovered for former generations, our present project could never have been undertaken.

There are yet other studies of particular subjects in the history of aging that are of relevance to our purposes, and they will be touched on when we come to the aging of particular groups and restricted areas and aging at particular times. But we must glance here at the general, outline histories of the topic that its salience at the present moment has given rise to. They seem to be of much more limited value than the writings we have gone over. Composed, like some of the special studies, with the use of traditional historical sources, mainly literary evidence, and sometimes with scant respect for demographic analysis or even demographic reality, they seem to be preoccupied with past ideas about the division of the life course. They are to be classed as histories of aging attitudes and attitudes to the old rather than as histories of aging and will perhaps impede the progress of the study rather than forward it.[8]

Nevertheless, no great stress will be laid on the contrast between the historical demography of aging and the history of aging. As we approach demographic analysis proper, however, it has to be pronounced that by and large the historical demography of aging is confined to the now developed countries and, with salient exceptions in the history of those developed countries themselves, to the period that is called the statistical era. The statistical era is the time during which states have carried out exact recording of vital events, analyzed and preserved those records, and made them accessible. For most of the developed countries, the statistical era begins some one hundred fifty years ago, in the middle of the nineteenth century, though a century earlier in Scandinavia, notably Sweden. For these Western nations a great deal of what the historical demographer wishes to know about age and aging can now be recovered. Census and local census-type materials, where and for what years they are open to examination,[9] make it


9

possible also to investigate such things as the position in the household of persons of various ages, their kin relationships, and, by reference to their vital statistics, what has come to be called their life course transitions. Data of this kind have been extensively used in the chapters in this volume.

There are even some developed countries, notably those of Scandinavia, along with the Netherlands, Belgium, and Italy, where population registers were regularly maintained by local authorities in past periods and are now accessible. These are lists of all the inhabitants of a particular place distributed into households, giving ages and updated for every birth, marriage, death, entrance, or departure. In these very special cases, if the registers have survived, in principle virtually everything that the historical demographer wishes to know about aging is available.[10]

Since the 1950s and because of the activities of international statistical bodies such as the demographic division of the United Nations, data are plentiful for populations at all stages of development and multiply rapidly. Here the task of the historical demographer is to disentangle the figures that can be used to produce an intelligible account within that excessively narrow time period. As he or she works backward toward the beginnings of the statistical era, the evidence thins out rapidly, but in comparative terms the temporal depth remains woefully shallow. For by far the largest part of history in its conventional definition—past time during which written records were maintained and can be consulted—all there is to go on is demographic theory and analogy. This makes the historical demography of aging a very different pursuit from previously established forms of historical study. Not only is nearly all the available material confined to the last two or two and a half centuries, compressed so to speak in a dense layer on the surface of time, but almost all the notable action is in fact confined to that time space. The temptation that has to be resisted is to follow the unfortunate example of social scientists generally and to read history backward from the present moment. Our enterprise as to aging in the past is undeniably demanding and difficult, replete with somewhat hazardous inferential argument.

This comes out in the fact that the statements already made about necessary knowledge apply to all the world's populations over the whole of their history, while the relevant evidence is so recent and so concentrated. We have to do everything we can to make use of the records of a handful of national populations, as opposed to individual communities, for which reliable numerical knowledge of aging can be pushed back for a hundred years earlier than the statistical era, to the middle of the eighteenth century. France and the Scandinavian countries and parts of Italy are in this position. In one country, England, using the same means but in what might be called heroic form, the time horizon has been stretched backward another two hundred years to the mid-sixteenth century for the entire national population.[11] All these statistics from the time before the official statistical era


10

have been derived from registrations made by the Christian church. We shall see, however, that even the four and a half centuries of the English historical record on aging, from the 1540s to the 1990s, are rather short for our purposes.

We now turn to the actual process of aging in populations of some size, nearly always national populations, and begin with the demographic transition itself.

The Aging of National Populations

Since the demographic transition consists in a monotonic and irreversible fall in fertility and in mortality, from generally stable high levels to less stable low levels, it must always lengthen life and inevitably give rise, in due course, to an increase in the proportion of elderly people in a population. However, like fluctuations in total numbers, aging in this latter sense, that of our second aging variable, is a second-order demographic effect. The increase in proportions of elderly persons does not come about as a direct result of change in first-order variables such as fertility, mortality, and migration. It is an outcome of the interplay between those variables as they change over time, on the one hand, with the initial age composition of the population, on the other hand. It is the case, however, that in most situations, certainly those that characterize populations on the lower aging plateau, a fall in fertility, that is as large and continuous as it has to be during the demographic transition always initiates and maintains a proportionate increase in the numbers of older persons. The part played by falling mortality, which goes forward at the same time, is negligible at these earlier stages of the change in age composition, though it is of course wholly responsible for increase in longevity.

The process is not easy to explain in terms accessible to those unfamiliar with demographic analysis. It is important to recognize that because the rise in the numbers and proportions of the old is a delayed rather than an immediate outcome of the demographic transition, the momentum which causes that rise continues after the transition is over. Different countries therefore show different patterns over time in changes in their age composition, in accordance with the differing calendar years over which they experience the transition and with the varying course that the transition may take. They also age at different speeds. We shall only be able to refer here to the historically more important of these differences between the records of the countries we are considering.[12]

Countries have also varied in the timing and pace of aging in our first sense, in longevity, though we shall not be able to linger on this matter either. But the recognition of longevity as of equal significance to change in age composition in the aging process and its analysis is an extremely important matter.


11

Life expectation has been regarded by demographers hitherto as a function of the mortality prevalent in any given population and as nothing more. Accordingly, it is classified by social scientists and by administrators as a measure of general health and well-being, not in itself as a manifestation of aging, individual or social. Since their interest stretches over the whole of the life course, it is expectation of life at birth that is the commanding statistic, worked out and cited almost to the exclusion of longevity at later ages, except of course when life expectation at subsequent birthdays is the particular interest in mind. From our point of view, that of historical sociology, this is wholly too narrow a view of longevity and has to be expanded. Expectation of life at every age and rises or falls in its duration, the probability at one birthday of reaching a particular later birthday, are all crucial aging phenomena on their own account, both for the individual and for society at large, in our own day and assuredly in the past. What people have always wanted and needed to know is how long persons like themselves are likely to live, their friends and their relatives included, and social scientists need to know this too, if they are to grasp the durational reality.

The historical aging process in national societies will therefore be discussed here in terms of two variables, not one, two variables treated as if they were discrete. The first is change in longevity, or life expectation, that is to say, not only average expectation at birth in a population but increasingly as we proceed expectations at later ages. The second is change in the proportions of the population of those in the higher age groups, generally those over 60 but increasingly the share that these older persons made up of all those in mid- and later life, that is, all those over age 25. Tables 1.1 and 1.2 set out the statistics of these two variables in their initial form, expectation of life at birth and proportions over 60, for the populations of a number of now-developed countries between the 1880s and the 1980s. Numbers in the final columns in both tables will give some idea of the extent of the rise in values between the last quarter of the nineteenth century and the present day.

It is evident from the figures that in the 1880s there were considerable differences between the selected countries with respect to the two chosen aging variables. It is also evident that low mortality (high expectation of life) was not always associated with large proportions of elderly people. This comes out clearly in the contrast in the tables between France, on the one hand, and England and Wales, on the other. Longevity was lower in France in the 1880s, yet France had a two-thirds larger proportion of persons over 60. This contrast is even more marked when it is recalled that in Britain there was in progress during the 1880s a considerable out-migration that had continued for a long time. This made the British population older than it would otherwise have been. What made Britain older in this sense, of course, made countries of immigration like the United States, Canada, and Australia younger, because many of the migrants were younger Britons. The


12

TABLE 1.1
Change in Life Expectancy (at Birth) during the Secular Shift, 1880s to 1980s

 

1

2

3

4

5

 

1880s

1920s

1950s

1980s

Risea

 

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Australia

47.2

50.8

59.0

61.4

66.8

70.5

73.0

79.4

156

156

Belgium

43.8

46.9

56.2

59.8

62.9

68.3

70.4

76.9

161

164

Bulgaria

45.9

46.6

58.0

62.0

68.4

73.6

149

158

Canada

43.5

46.0

55.6

58.4

66.4

70.8

73.0

79.8

168

169

England and Wales

44.2

47.5

55.9

59.9

69.8

68.2

71.0

77.6

161

163

France

40.8

43.4

52.3

56.7

63.6

69.4

72.0

80.3

176

185

Germany

35.6

38.5

56.0

58.8

64.6

68.5

71.8

76.4

202

198

Greece

36.0b

37.5b

48.4

50.0

63.4

66.7

72.2

76.4

200b

204b

Italy

33.3

33.9

48.8

50.4

63.9

67.9

72.0

78.6

216

232

Japan

42.4

43.7

44.8

49.6

59.4

62.7

75.5

81.3

178

186

Spain

33.8

35.7

40.3

51.6

59.8

64.3

72.5

78.6

214

220

Sweden

48.5

54.1

57.5

60.2

69.1

72.3

74.2

77.0

153

142

Switzerland

43.3

45.7

54.5

57.5

66.3

70.5

73.9

80.7

171

176

U.S.A

42.5c

4.5c

55.0d

57.1d

66.3

72.0

71.3

78.2

168c

173c

SOURCES : Dublin, Lotka, and Spiegelman 1949; National Statistical Yearbooks; Preston, Keyfitz, and Schoen 1971; and various.

a For Bulgaria, values from column 2 = 100; for all other countries, values from column 1 = 100.

b 1879.

c 1890s, Massachusetts whites only.

d Registration states (17% of U.S. population 1880s, 82% 1920s, whites only); see U.S. Bureau of the Census 1937.


13

TABLE 1.2
Change in Proportions over Age 60 during the Secular Shift, 1880s to 1990s

 

1

2

3

4

5

 

1880s

1920s

1950s

1980s-1990s

Changea

 

Male

Female

Both

Male

Female

Both

Male

Female

Both

Both

 

Australia

7.7

7.3

7.5

11.5

13.4

12.4

11.9

Belgium

9.4

10.2

9.8

9.4

10.9

10.1

14.4

16.6

15.5

15.8

161

Bulgaria

7.0

6.4

6.7

8.6

8.3

8.5

9.1

10.1

9.6

Canada

5.2

6.0

5.4

6.8

7.8

7.5

11.5

11.1

11.2

11.3

209

England and Wales

6.9

7.8

7.4

8.8

10.9

9.4

13.9

17.8

15.9

17.0

230

France

12.0

12.6

12.3

13.0

14.7

13.9

13.4

19.0

16.3

16.6

135

Germany

7.5

8.2

7.9

8.6

9.6

9.1

13.0

14.4

13.8

16.7

211

Greece

5.0

4.1

4.6

8.5

8.9

8.7

9.1

10.4

9.8

15.2

330

Italy

9.0

8.9

9.0

10.8

10.9

10.8

11.3

13.1

12.2

15.5

172

Japan

7.0

8.4

7.7

12.9

Spain

7.8

7.9

7.8

10.0

10.0

10.0

14.1

180

Sweden

8.4

10.2

9.4

11.1

13.3

12.2

14.2

15.8

15.0

17.7

188

Switzerland

8.6

9.1

8.8

8.3

10.1

9.2

12.7

15.4

14.1

16.3

185

U.S.A.

[5.4]b

6.7b

6.8b

6.7b

11.8

12.5

12.2

15.9

[294]

SOURCES : National Statistical Yearbooks; and various.

a Values from column 1 = 100.

b Registration states (17% of U.S. population 1880s, 82% 1920s, whites only); see U.S. Bureau of the Census 1937.


14

variation between these populations and other similarly placed populations as they were in the 1880s and the rates of fall in fertility and mortality. that ensued during the transition ensured that they would age at different paces. But the effect to which we have already referred, monotonic fall in fertility initiating and sustaining change in age composition, was a constant, and it could be shown that the accompanying fall in mortality had very little to do with it until the last decades of the secular shift.

We have now reached the point that has to be called the pons asinorum of demographic aging, a bridge that has to have been crossed by everyone undertaking historical analysis of the subject. It is universally assumed by those who have never had occasion to examine the issue that populations grow old invariably because deaths go down and life expectation goes up. This mistake is the more understandable in that the demographic transition itself starts in most cases with a fall in mortality. But it is a serious mistake, nevertheless, particularly often made by historians without demographic knowledge. They think it natural to expect that if there is evidence of rising longevity at a particular period, proportions of older persons must have been growing, and vice versa. It usually does not occur to them that a fall in births will work to expand the proportion of the elderly in the population, that this effect could be reinforced by the very young living longer, least of all that a fall in deaths might contract that proportion because, especially in the Before, such a high proportion of deaths happened to babies.

The Secular Shift in Aging

The secular shift in aging in England is set out schematically in figure 1.1 in ideal terms and on a scale that goes back for some eight centuries into the pretransitional past, in the Before, and is projected forward in the After some six centuries from the 1990s.

The general shape of the curves conveys the idea of the secular shift as an upward thrust from a lower aging plateau stretching backward in time from the last decade of the nineteenth century up to a much higher aging plateau sloping upwards and stretching forward from the first decades of the twenty-first century. But in this hypothetical model, the higher plateau is a particularly speculative construct, and its course over time after the first few decades of the twenty-first century is very uncertain. Projections made by demographers specify a range of alternative futures, some of which are quite extreme. These uncertainties make it necessary to represent the future courses of both longevity and proportions in later life as areas of possibility rather than as determinate lines. The earlier plateau and the rise from the lower level after the 1890s are, however, quite well documented for a number of populations, as will be shown. In England, its steepest climb in both graphs evidently took place between the 1920s and the 1950s. These


15

figure

Fig. 1.1.
Secular shift in aging: England, 1000s-2500s.


16

figure

Fig. 1.2.
Expectation of life at birth and proportion of 
population over age 60: England, 1540s-1990s.

effects are very clear in the detailed representation of data for England set out in figure 1.2 covering the four hundred fifty years over which the requisite data are available. It is not possible to divide the genders until the 1870s, but the widening gap between male and female as we approach the 1990s is unmistakable.[13] In figure 1.3, the English curves are set out again, along with directly comparable curves from France and Sweden, two other countries that also have evidence for both longevity and age composition stretching some way back from the beginnings of the statistical era, distinguishing male from female at the outset. Figure 1.4 compares Tuscany, that area of Italy in which Florence is situated, with England in expectation of life at birth for three hundred years after the early seventeenth century. The stretch of years between the 1640s and the 1890s in this figure represents the longest period during the Before over which life expectation at birth can at present be compared for two substantial European areas.

Apart from these pieces of relevant evidence, the English record is all we have to go on before about the year 1740 to assess the experience of aging


17

figure

Fig. 1.3.
Expectation of life at birth and proportion of population 
over age 60: England, France, and Sweden, 1740s-1990s.

in Western countries—and in large part, the experience of the world as a whole—in properly numerical terms. This makes the statistics from England (and in more recent years, from England and Wales) strategically so important that all the relevant numbers for that country have been set out as table 1.3.

With this spread of information about England before, during, and after the secular shift in aging before us, there are two issues that have to be confronted if we are to appreciate critically its significance for the course of aging over the centuries and for the general subject of the historical demography of aging. These two issues are whether the English curves exaggerate the steepness of the rise in the two chosen variables and so cannot be called typical of the movement in all the countries concerned and whether the Before and the After are in fact to be thought of as plateaus. The upper plateau can scarcely be our present concern. Only the analogy of the level trajectory of the lower plateau, which we are about to discuss, justifies the suggested


18

figure

Fig. 1.4.
Expectation of life at birth: England, 1540s-1990s, and 
Tuscany, 1630s-1930s. Data for Tuscany from Breschi 1990.

stretching forward of the higher plateau into the indefinite future, though the narrowing of gaps between survival curves in figure 1.5 hints at such a future development.[14] Its possible or probable existence, however, gives a striking appearance to the secular shift in relation to what came before. It is to the lower aging plateau that we have to direct our attention.

Relative Historical Constancy in Aging Before the Secular Shift

Since England is unique in the way we have discussed above and since it was from the English evidence[15] that the model of the secular shift in aging was originally constructed (see Laslett 1984), the question of whether England has been typical is not without its importance. There would seem to be little doubt that this was so with respect to the increase in expectation of life at birth, the first of our aging variables. France and Sweden, whose curves


19

TABLE 1.3
Expectation of Life at Birth and Proportion over Age 60, England, 1541-1991



Year

Expectation of Life at Birtha

Proportion over Age 60b



Year


Expectation of Life at Birth


Proportion over Age 60

1541

33.75

8.67

1746

35.34

7.99

1546

32.50

8.47

1751

35.57

8.22

1551

37.99

8.35

1756

37.29

8.37

1556

30.73

8.16

1761

34.23

8.60

1561

27.77

7.29

1766

35.04

8.76

1566

37.97

7.21

1771

35.60

8.50

1571

38.22

7.32

1776

38.17

8.36

1576

40.26

7.49

1781

34.72

8.24

1581

41.68

7.59

1786

35.93

7.97

1586

38.31

7.80

1791

37.33

7.41

1591

35.51

7.93

1796

36.76

7.37

1596

37.65

8.08

1801

35.89

7.26

1601

38.12

8.27

1806

38.70

6.99

1606

40.82

8.62

1811

37.59

6.89

1611

37.27

9.08

1816

37.86

6.86

1616

36.79

8.90

1821

39.24

6.68

1621

39.95

8.02

1826

39.92

6.54

1626

33.96

7.90

1831

40.80

6.56

1631

38.71

8.03

1836

40.15

6.58

1636

36.14

8.12

1841

40.28

6.58

1641

33.70

8.27

1846

39.56

6.57

1646

38.47

8.59

1851

39.56

6.56

1651

37.82

8.92

1856

40.39

6.80

1656

34.11

9.28

1861

41.19

6.87

1661

35.71

9.40

1866

40.32

6.93

1666

31.79

9.73

1871

41.31

7.05

1671

33.18

9.89

     

1676

36.37

9.95

 

Male

Female

Male        Female

1681

28.47

9.71

       

1686

31.77

9.10

1881

44.2

47.5

6.9                7.8

1691

34.87

9.06

1891

41.9

45.7

6.8                7.9

1696

34.13

9.18

1901

48.0

51.6

6.8                8.0

1701

37.11

9.38

1911

49.4

53.4

7.3                8.6

1706

36.44

9.81

1921

55.9

59.9

8.7              10.0

1711

35.93

9.97

1931

58.4

62.4

10.7            12.3

1716

37.10

10.08

1941

59.4

63.9

—                 —

1721

32.51

9.46

1951

66.2

71.2

14.6             17.7

1726

32.41

9.11

1961

67.9

73.8

15.3             17.9

1731

27.88

8.41

1971

68.8

75.0

15.9             21.9

1736

35.64

8.35

1981

69.8

76.2

16.2              22.7

1741

3170

8.11

1991

70.1

78.3

16.5              23.1

SOURCES : British official statistics; Wrigley and Schofield [1981] 1989.

a 1541-1545 to 1871-1875: Mean, 36.45; standard deviation, 3.25; minimum, 27.77; maximum, 41.68.

b 1541-1545 to 1871-1875: Mean, 8.31; standard deviation, 1.48; minimum, 6.54; maximum, 10.08.


20

for the course of this statistic over time are also depicted in figure 1.3, show a steep or very steep rise similar to that for England, a rise that began at some point in the last decade of the nineteenth century or the first decade of the twentieth. In the case of Tuscany in figure 1.4, the resemblance to the curve for England is quite striking, rather surprising in view of very different conditions of health and welfare. There seems to have been a slow upward incline of expectation of life at birth in all Western countries during the nineteenth century before the secular shift, and the Scandinavian countries were well above England by the time the really steep ascent began. Such a circumstance might be taken as putting England into the median position. Longevity has been excellently assessed in all now-developed countries for this period, and it is possible to be confident that England can stand for the rest. And if this was so in the run-up to the secular shift and for the shift itself, why should it not have been so in earlier decades and centuries?

The same cannot be so easily pronounced for our second aging variable, traced by the lower graph in figure 1.3, increase in proportions over the age of 60. Here the English figures show a tendency to fall in the earlier nineteenth century, as is evident in table 1.3, putting the English graph for proportions of elderly well below those for France and Sweden, and this can be confirmed in other countries. The subsequent precipitate ascent in England may well exaggerate the abruptness of the aging transformation at the secular shift. Even in other English-speaking populations, the share of the old rose more slowly than in the "mother country," though their character as immigrant receivers may have something to do with this.

The contrast between English experience of growth in the weight of older people in the population at large is particularly marked in relation to France. But it is well known that the demographic transition itself was more diffuse in France than elsewhere, and it has to be expected that in this respect the secular shift would also be more diffuse in that country. The general allure , as the French would themselves say, of the French curves is quite similar to that of the English and the Scandinavian. With such a small sample and in so novel and uncertain an area of investigation, the correspondence between these graphs seems acceptable. I am prepared to regard the shape of the secular shift in England as an ideal type of that development, ideal type in the sense used by Max Weber. We shall return to this point when we come to discuss more revealing aging measures than those we have used so far.

As for the levelness of the lower plateau, which is a way of expressing the long-term constancy of these two aging variables in historic populations, the course taken by the graphs in figure 1.3 certainly suggests that in the long term relative stability can be assumed as well. It is here that the length of the record over time becomes highly significant, so that even the near half-millennium covered by the English data may appear somewhat inadequate.


21

But the line representing the English statistics of proportions over 60 is conspicuously flat for the complete run, and those for the other countries being compared are very much the same for their interludes of record. A survey of the quinquennial figures for England contained in table 1.3 shows that proportions varied between extremes of 10.08 percent (maximum) in the years 1716 to 1720 and 6.54 percent (minimum) in the 1820s and 1830s but that three-fourths of the values fell between 6.80 percent and 8.50 percent. In spite of differences in the later eighteenth century and earlier nineteenth century, the impression of constancy remains with respect to the weight of older persons in the population at large. But we shall find ourselves wondering whether this was quite so evident for the proportion of elderly persons in the adult as opposed to the whole population and whether the rise was as considerable at the secular shift in that respect.

The course of expectation of life at birth before the secular shift has a much less even appearance than the course of proportions of elderly: ups and downs succeed each other in the English figures in a way that recalls the heraldic description dancetty . Nevertheless, the claim for an underlying constancy seems strong in this matter too. Variation between extremes of 41.68 years (maximum) in the 1580s and 27.77 years (minimum) in the 1560s—the propinquity in date of the maximal peak and minimal trough should be noted—is accompanied by a concentration of two-thirds of the values between 35 and 40 years. When the graphs for expectation of life at birth are smoothed by the use of moving averages,[16] they also have a decidedly even appearance. Work in progress at the Cambridge length of life project suggests a surprising stability over time in the longevity figures for elite groups, back to the later Middle Ages for British peers and members of Parliament, back to the beginning of the Christian era for Chinese mandarins. It seems safe to assume that the secular shift is properly represented by a sudden, precipitous rise from a lower plateau to a higher level in both dimensions.

The implications of these considerations for the historical demography of aging must be quite evident. Although they were subject to quite sharp fluctuations in life expectation at birth-and here the effects of epidemics, wars, and food shortages spring to mind—our ancestors never seem to have been subject to aging changes on anything like the scale that has been experienced by the populations going through the secular shift. Recovery in duration of life was rapid after episodes of disaster, though it must be remembered that this recovery consisted largely of better prospects for newly arrived infants and children. Proportions of elderly persons in the population at large remained fairly constant, showing the same tendency to revert to the average after rises and falls, for centuries on end as far as we can tell.

This conclusion is reinforced by demographic knowledge and demographic theory. Although we have had to recognize that no example of a population in what might be called a "natural" condition with respect to ag-


22

ing is or is likely to be available, demographers have had extremely wide experience of populations that had not yet entered the demographic transition, or are in its early stages and hence so far not much affected by the secular shift. Their aging characteristics are indeed plateaulike. Episodes of low fertility along with relatively low mortality cannot have lasted long enough among these populations for them to grow old in the dual fashion that has been described for the secular shift. This has to be the case, since pretransition populations had and still have high birth- and death rates by definition.

Duration of Life at Every Age and Age Composition Historically Considered: The Rectangular Survival Curve

The theme of the historical demography of aging has been treated so far as if it were concerned only with those who reach, who have reached, or who are likely to reach late life. This may seem appropriate because gerontology, the scientific study of aging, is itself pursued in this way and because the chapters that follow here deal with such topics, many of them belonging interestingly enough to the period of the secular shift when it occurred to the population of the United States.

But the elderly have only ever been the topmost of the age levels of any society, and aging at all the other, lower levels are alike of importance to the social scientist, the historical sociologist especially. The system of distributing individuals into age groups—childhood, youth, middle age, and so on—found, but in differing forms, in every, society and at all times together with the interrelationships between those groups and their members are all of significance. Generational relationships are rightly recognized as a subject for anthropological and sociological as well as for historical and gerontological research. Taken as a whole these studies go by the name of age structuring, until recently pursued mainly by anthropologists, though with some reference to historical instances.[17] Analyses of how age structuring differs from society to society and how far it has changed historically certainly belong with the historical demography of aging, since the relative size of age groups will change with demographic change. What is more, all members of all age groups above infancy bear the marks of their experiences as members of younger age groups. This highly significant fact is of greatest importance for those who have gone through the largest number of age divisions, or life course stages, that is to say, those who are old at any one time.[18]

Both expectation of life and the relative sizes of the membership of age groups are relevant here once again. For it must never be forgotten that everyone is always getting older at all points in the life course and that prospective length of life to spend in each life course division, or in divisions


23

to come, makes a considerable difference to individual attitudes and other features of age structuring. The question that presents itself is how far expectation of life at all ages, not simply at birth, is likely to change over time (or to differ between countries and cultures) and how far relative age group sizes do the same.

An interesting conspectus of capacity to go on living and its fluctuations over time is provided in figure 1.5, which traces not so much values of years still to live as they have changed over the centuries but survival itself. Survival curves plot for successive dates the numbers of an original 1,000 newborn babies still alive after 1 year, 5 years, 10 years, 20 years, and so on, up to 85, or to 100 for the most recent case. The record is again the English one, the first country for which such estimates have become available so far into the past. Up to that for 1691-1695, each curve is based on evidence from one five-year period; after that the curves relate to briefer periods or a single year.

figure

Fig. 1.5.
Survival curves for cohorts of one thousand newborns, by age group:
 England, 1541-1991. Data from Cambridge Group back projection files
 and English Life tables up to no. 15. Work of James Oeppen.


24

The lower aging plateau is evident in the curves up to that for 1841-1845; here lines are separated by intervals of half a century each. These lines are not only very close together but they change positions with each other in an order that is certainly not chronological. The one for the five-year period 1691-1695, for example, traces a course only just above that for 1541-1545 and well below that for 1591-1595, which in its turn is closest to that for 1841-1845, nearly two hundred fifty years later. The run-up to the secular shift is visible in the wide gap between 1841-1845 and 1891, when mortality was evidently falling but at some ages may still not have been below the levels previously reached for an individual year or so on the lower aging plateau. The subsequent course of the secular shift itself shows up vividly in the more and more conspicuous spaces between successive curves, this time temporally successive curves and separated by ten-year and not by fifty-year intervals as is the case for the earlier lines. The approach to the higher aging plateau in our own day is strongly suggested by the marked narrowing of the spaces after 1951.

Students of aging will notice that the figure as a whole is reminiscent of the theory of what is called "The rectangularization of the survival curve," which is discussed in chapter 6 of A fresh map of life (Laslett 1989; see especially the revised version in press). This predictive theory, propounded by the American researcher James Fries and based on the very much more accurate survival curves now available, might provide a seductive concluding act to the historical drama we are commenting upon here. The finale would be a dramatic compression of mortality, showing itself in the survival curve pattern as a conspicuous narrowing of the gaps between the lines in figure 1.5, tending toward their elimination, particularly at the extreme right-hand side. Simultaneously these overlapping curves would change their shape even further in the direction shown forth in all their predecessors, tending more and more closely toward the rectangular. These shifts would indicate that survival will get better and better until finally virtually all persons born would be alive in their later 80s or their 90s. At some point during those years, it is further claimed, the end will come for a larger and larger number as time goes by, and after briefer and briefer final illness a "natural" death will supervene, a "natural" death at the completion of the "natural" span of human life.[19]

In Fries's model, then, morbidity, the tendency to fall ill, is to be compressed along with mortality. In favor of this dual hypothesis is that at all periods before the secular shift, on the lower aging plateau, that is to say, sickness and death were distributed over the whole life course: during and since the secular shift there has been an increasing tendency toward postponement to the later years. When the higher plateau succeeds after the secular shift is over, Fries's theory predicts that both sickness and death will be almost entirely concentrated within a very short period, most likely during the middle or late 80s of the life course. If the model were to be validated, its ef-


25

fect would indeed be that the survival curve will eventually become rectangular, fulfilling the tendency so conspicuous in our diagram of English development over half a millennium.

It is becoming evident, however, that survival curves in our own day are resistant to such an interpretation. Some of the number of competing theories of aging at the present time imply that they will finally approximate the rectangular shape. Otherwise the life span, known to be lengthening, will do so indefinitely. But the attractive thesis of a prospective compression of the incidence of final illness and death has to be regarded as not proven, a fact with evident implications for the uncertainties over the higher aging plateau. The issues it raises, however, decidedly intensify the interest of this topic in the historical demography of aging. For our present purposes, the significance of four and a half centuries of English survival curves, which as we have seen do not necessarily represent in detail what has gone on in other comparable populations, lies elsewhere. It is in the fact that their shape was indeed remarkably constant over time until the secular shift. Habits, attitudes, and institutional arrangements that were based on the assumption that most people would be alive in the same proportion of numbers at the same age as time went by and have roughly the same numbers of years to expect to live were apparently well founded.

It can be quite simply asserted, then, that over the three and a half centuries of English history before the secular shift for which the required knowledge has been established, age structuring was subject to very little change with respect to longevity. As has been hinted, the most interesting of the effects that finally supervened at the height of the shift itself is the emergence of the Third Age, which will be initially defined demographically in terms of survival by the establishment of a novel relationship between the number of those alive at age 25 and the number alive at age 70.

Exactly the same conclusion as to relative constancy during the Before can be drawn from figure 1.6, which traces the fluctuating boundaries between age groups composing the entire English population over the same long period. (Note that estimates before 1870 cover England and Wales; after 1870, the two entities can be distinguished in official figures.) All the age groups show continuity over time in their relative sizes until the secular shift sets in. It is true that the line bounding the elderly (those over 60, the lowest proportion of course until the end of the nineteenth century) is noticeably smoother than the lines bounding other age groups (0-4, 5-14, 15-24, 25-59). But all five boundaries are quite regular before the shift, regular, that is, within short-term fluctuation up and down. The only comparable reconstruction, that for Tuscany from 1810 to 1940, has a very similar pattern (Breschi 1990: 165). It seems to have been exceptional fertility falls in England during the 1550s and the 1640s to 1670s that led to the frontiers between shares of children (5-14) and adolescents (15-24) crossing each other for a little while. But otherwise age composition displays overall con-


26

figure

Fig. 1.6.
Age composition of the population: England, 1540s-1990s.

stancy, and the statements made above about survival and life expectation apply here as well.

Stability in age composition did not survive the onset of the secular shift, and the area near the right margin of figure 1.6 demonstrates as much. Proportions over 60 rose abruptly until by the 1950s they became the second-largest age group and went on growing without interruption until the end. It is well known that this relative expansion was, is, and decidedly will be accompanied by a parallel change in composition within the open-ended group of those over 60. The very old—the 80s to 84s, 85s to 89s, the over 90s—will increase rapidly, along with their longevousness. This is a radical change from traditional age structure and one that creates an entirely novel and very formidable prospect for the present generation. Although we do not have the information to confirm the generalization in detail, it is very probable that at all earlier times the composition of the group of those very late in life tended to be constant in all larger populations.

Our ancestors, that is to say, those who were citizens of our countries before the secular shift in aging, could rely on a relatively stable age structure as well as a relatively static life expectation at every point in the age range.


27

All over their populations and at every period, the proportions of infants (0-4), children (5-14), adolescents (15-24), and those at working age stayed about the same, as did proportions of the elderly. The same statements could probably be made with some modification about the populations still on the lower aging plateau at the present time. The secular shift upsets this perennial pattern quite radically, and all age boundaries and traditional assumptions as to length of future life tend to lose their stability as the shift proceeds. The whole of the evidence we have so far sun, eyed suggests that this represents a unique occurrence in human history.

Aging by Locality and by Social Group

Up to this point, we have talked about duration of life and about age composition over long historical periods and always in terms of averages over whole populations. If we are to reconstruct the aging landscape and habits, attitudes, and institutional arrangements of the past, and if we are to get some insight into the results of the secular shift in aging, this procedure is satisfactory enough. Societywide beliefs and norms were based on perduring and very general experience, nationwide experience for the most part, and still are in whole tribes or whole societies, whatever is the proper unit to take. But it is of great importance to recognize that there could be very different situations with regard to aging in different localities and in different social groups before the shift, in spite of the continuities and the constraints on extreme variations that we have examined. Relative constancy did not necessarily obtain at the local level or among select samples of people during the Before.

Individual towns or villages, even fairly large areas with particular ecologies, could show noticeably irregular aging characteristics. At a place like Hartland in the English county of Devon, quite a sizable but isolated settlement at the end of a road leading up to higher land, expectation of life at birth could be maintained at something like fifty years for the three centuries from the mid-1500s to the mid-1800s. This was about ten years above the estimated peak level reached for the whole population of the country over that period and one not generally attained in England until after 1910. The operative cause was undoubtedly markedly lower infant and child mortality than elsewhere, though we shall see that this did not necessarily imply better survival in later life. Similar circumstances can be found in Oxford-shire villages at the same time.[20]

As for differences by social class and social group, it is natural to expect that there would have existed during the Before some form of hierarchy in longevity, descending from top to bottom of the society, a hierarchy that would be concealed by the averages we have cited. The landowning nobility and gentry could be supposed to have lived longer than the substantial farm-


28

ers, the wealthy traders, the small landholders, and so on down through the modest tradesmen, the handicraft people, the laborers, and finally to the cottagers and paupers. Although this may have been the general tendency, and it can be confirmed for infants and children, there were surprising irregularities. In England during some of the earlier part of the period we can examine, the aristocracy actually died somewhat earlier on average than the rest of the population. Since their longevity was not higher at later ages, this may not have been due, as might be expected, so much to their having put their babies out to wet nurses rather than breast-feeding them themselves, like their social inferiors, and to the use of nourishment other than human milk (see Fildes 1986, 1988). There were evidently other causes such as a predilection of the nobility to lose their lives in warfare.[21] Of one thing we can be confident, however. Everywhere in the traditional world death came sooner among the inhabitants, even the richer inhabitants, of cities and towns than among the rural majority. The explanation is the life-endangering conditions of urban existence, of which putting children out to nurse was one.

A study of a particularly interesting example of a privileged group, the advocates of Edinburgh, has recently been published showing that these prosperous lawyers could expect to live after age 25 to 30—the start of their professional lives—for longer than the peers of the realm, and by the later eighteenth century, for as long as any of the select European groups that have been investigated.[22] Such bands of well-placed individuals could anticipate a considerably larger number of additional years after midlife than those in lowlier positions. But difference by locality was not confined to length of life: it existed in age composition. Wide variation is found from place to place in our second variable, proportions of elderly. In Arezzo, Italy, in 1427, 15.9 percent of the citizens were recorded as over 60 years old, half as many again as the maximum on the lower aging plateau for England and a figure well up to late-twentieth-century proportions in developed countries. The national census of Iceland in 1729 gives a figure of 16.7 percent.

Arezzo and Iceland were evidently exceptional cases, and we should be very cautious in trusting their accuracy. The other communities in the list of thirty or so from which these figures for proportions of elderly have been recovered for earlier times fall well within the limits given above for England before the shift, and some are as low as those for contemporary developing societies.[23] These exceptional figures, nearly all of them for relatively small bands of persons, may perhaps require us to modify a little the account that has been given of the contrast between the Before and the After. Their greatest significance for the historical sociology of aging, however, lies in two other circumstances of prime importance.

One is the caution that has always to be exercised in generalizing from the statistics of a group or a locality, either as to length of life or proportions elderly, to larger populations and to general conditions. Apart from the vari-


29

ability always encountered when numbers are restricted, there is the question of migration, migration that was never absent from preindustrial European society and the extent of which consistently surprises those who first encounter it. The immigration of poor elderly widows to places where public and charitable support might be forthcoming or the opposite, the emigration of younger people in search of employment opportunities, could easily make the share of the elderly in the population of a small community much larger, and the contrary effect might reduce that share accordingly. The chapter by Andrejs Plakans and Charles Wetherell in this volume discusses some of these issues in an eastern European context in the eighteenth century. Likewise in the matter of life expectation. It is transparently clear that one should not argue from the characteristics of a select group of the better off, or of a particularly favorably situated community, to a whole region or country. Great ingenuity is necessary if any guidance as to general conditions as to the life duration is to be gained from such evidence.[24]

The further point about arguing from figures from localities and from exceptional groups of people needs particular emphasis. In spite of the low or very low general expectation of life in the Before, particular individuals could live very long lives, even in places where mortality was high. Our early research into maximal length of life makes this evident. A survey at Cambridge of the information generously supplied by genealogists and local researchers, mostly studying their own forebears using parish registers and the official registration system, yielded the following results. Some 3,500 individuals were fairly reliably reported to have survived to age 80 or beyond, and a fair proportion of these died in their 90s. There were 89 (2.5 percent), 52 females and 37 males, who attained 100 years or more, 2 dying in the seventeenth century, 19 in the eighteenth century, 21 in the nineteenth century, and 47 in the twentieth century.[25] This may seem an unimpressive outcome, especially in view of the very numerous references to the extremely old, particularly to centenarians, in the literature of former times and up to our own day. But it could be said to indicate that survival to the tenth decade of life or, very, very occasionally, even to the eleventh decade (centenarians) was just a possibility on the lower plateau.

This exercise also demonstrates a salient fact about the very old, which no one interested in such matters can fail to observe. Both in the past and the present, in the developed countries, and those yet to undergo the secular shift, exaggeration of the later ages of individuals was and is lamentably widespread, especially when it comes to reaching 100. It is so for the alleged centenarians themselves, for their relatives and even for officials, local and national. It is blatantly evident on tombstone inscriptions and other family records. The worldwide interest in this topic continues, and even grows, not unexpectedly in view of the known tendency for the survival of ancients to increase. It may be intensified by the distinguished recognition given to cen-


30

tenarians, as for example by the telegrams sent out by the Queen of England and by professional gerontologists who regularly include sessions on centenarians in their conferences, national and international, and who show little initial insight into the registration data. This encourages exaggeration and the disposition to point to numbers of centenarians as a mark of prestige, national and local. The historians of aging cannot be too heavily warned against these distortive influences.

A particularly exasperating and highly misleading fable about survival to a great age is the claim so often made for very special areas such as the Caucasus. Usually remote and mountainous, these are alleged not only to have exceptionally high levels of survival but to produce centenarians at an unwonted rate. Such localities were extremely unlikely to have maintained the highly reliable registration practices which alone would make such claims acceptable, and the stories about them have been described as being "as mythical as Methusaleh" (Laslett 1989: 109). It may be a long time before the wheat is sorted from the chaff in these matters.

Experiential Measures in the Historical Demography of Aging

We have yet to consider the actual impact of the secular shift in aging on the societies in which it has so far occurred. Before we go on to a preliminary assessment of these effects, however, we have to acquaint ourselves with the experiential measures that have been alluded to already. These are numerical measures that it is hoped come closer to the personal and social realities of aging change over time and of aging differences between societies. We shall then find it useful in relation to the secular shift to discuss an illustrative example of the application of the principles of the historical demography of aging and of the measures that will be suggested.

We begin with longevity and meet at the outset an obstacle to understanding that recalls the pons asinorum described earlier. It is natural for someone unacquainted with what the demographers call the life table to suppose that expectation of life—or average expectation of years still to come, to give it its full title—is always highest at the very beginning, at age 0. But this is not so, except under the very special demographic circumstances of the advanced countries at the moment and over the last few decades. Even in their case longevity is at its peak not at birth itself but a few weeks later. In all other populations average expectation of years still to come is higher at exact age 1 than at birth, higher again at age 2, and usually higher still at age 5, the expected peak value in all but contemporary developed countries. It can often be higher at age 15 or 20, or even sometimes 25 or 30. Let us take an example.


31

In the life table calculated for England and Wales as it was during the years 1989 to 1991, expectation of life for women was 78.67 years at age 0, 78.21 at 1, 75.29 at 5, and 64.41 at 15, all the later values being lower than the value at the beginning. For Canada in 1831, in contrast, it has been reckoned that expectation of life for women was 39.84 years at birth, 46.49 at age 1, 49.48 at 5, and 42.61 at 15, still over three years longer than at birth. The outstanding difference between the two populations, and such is always the case when comparing longevity between the Before and the After, was the very much higher infant and child mortality in Canada during the 1830s. The death rate of females in the first year of life in Canada in 1831 was 162.36 per thousand, whereas in England and Wales in 1989 to 1991, the corresponding figure was 6.78. In the second year in Canada, it was 135.76 and in England and Wales, 5.59.

If the rise in expectation of life in the Canadian figures at successive birthdays after the first seems paradoxical, because it appears to indicate that an individual person had longer to live at 2, 5, or 15 or later than at 0, this is due to the fact that the character of life expectation has not been recognized as an average, an average of the years still to be lived by a specific group of persons of a given age. Since such an average is the result of dividing the pooled total of years of life remaining to all members of the group by the number of members of that group, it must be possible for that average to go up rather than down after an interval during which an appreciable proportion have died.[26] This follows from the fact that such a loss could well reduce the numbers of persons surviving by relatively more, even very much more than it reduced pooled years still to live. Such would be the case in spite of the fact that the comparatively very large number of pooled years still to come would have been lessened by the years lived by those surviving through the interval, or through any part of it. A suitable metaphor might be that those still present at the end of the interval say to each other with some satisfaction, "Now that those unfortunates who were due to die during the interval are out of the way, we can all go on to live longer." Here the word "all" must be taken to indicate the new average of years remaining to each of them, that is, their expectation of life.

There are several implications of this explanation for the comparative study of longevity over time and between populations and for the selection of more realistic longevity measures. It demonstrates that expectation of life at birth has serious disadvantages as a measure, only appropriate for comparison of length of life between the populations of advanced societies in very recent times indeed. In the 1950s in the United Kingdom, longevity at age 1 was still nearly a year greater for females than it was at age 0 and over a year greater for males. The efficiency of expectation of life at age 0 as an indicator of the experience of longevity in other populations is obviously


32

woefully impaired because it is so heavily affected by high death rates in the earliest years.

If we were to contrast the life expectation of Canadian women in 1831 at its maximum, that is, at age 5, with that of English women in 1989-1991 at its maximum, that is, at birth (49.48 years vs. 78.67 years), this would seem to be a more realistic comparison than that produced by contrasting expectation of life at birth on both sides. The outcome would be that there was a difference of 29.19 years in favor of the contemporary English women rather than 38.83, a substantial reduction of well over a quarter in the superiority of the present day. If both the populations had been on the lower aging plateau, of course, comparisons of maximum longevity would have to be for values at age 5 on both sides. Comparisons of this kind would have to be made, however, in the knowledge that babies and young children who would die before the age of 5 had been left out of account.

Nevertheless, the aging experiences of such young and very young persons might not necessarily be supposed to be of much significance to the subject we are pursuing, though reflecting on their position brings home the important fact that aging is different at different parts of the life course. The experience of proceeding from the fifth to the tenth birthday is certainly not the same as proceeding from the tenth to the fifteenth, let alone from the sixtieth to the sixty-fifth, and these obvious differences are even greater when the historical periods and social circumstances of the populations being compared are allowed for. It has to be reckoned that there is a degree of indeterminacy in comparing the experience of longevity over time and between societies and that there seems to be no easy way of overcoming it. This may be one of the reasons why demographers have not included longevity in their analysis of aging.

The compromise solution recommended here is that life expectation at age 15, not life expectation at age 0, should be taken as the value for comparing longevity between societies at all historical epochs and at all stages of development. This simple expedient has the following advantages over using expectation of life at birth. Taking life expectation at age 15 as the measure of longevity makes unnecessary the reckoning of life expectation at different ages on the different sides of the comparison. Moreover, it is a much better guide to changes in longevity in later life.

Taking up the contrast between Canadian women in 1831 and English women in 1989-1991 once again, the percentage difference in life expectation at birth was 97 percent, that is, English women had 97 percent longer to live than Canadian women as reckoned in this way. But at age 15 the difference was 51 percent in favor of modern English women, while at ages 50, 60, and 70, it was 50 percent, 57 percent, and 63 percent respectively, each value quite evidently considerably closer to the differential figure for expectation of life at 15 than the differential figure for expectation of life at


33

birth. The value at age 5 would have similar advantages over expectation of life at birth as an indicator for this purpose, and it has the attraction that it is the highest longevity value for a very large proportion of all populations on the lower plateau. But expectation of life at 15 marks the point in the age range at which in many of them the average of years still to come ceases to be higher than expectation of life at birth. It could be said, moreover, that expectation of life at age 15 represents a closer approach to the median experience of longevity in the middle range of the population. But once again it has to be borne in mind that the aging experience of all immature persons is being omitted, and in populations on the lower aging plateau this means a quarter or even a third of the total.

If expectation of life at 15 is adopted as that measure of longevity that is least likely to misrepresent the aging experience of populations that are being compared, the question arises, what change might this imply for the secular shift in aging? Values for expectation of life at age 15 (e15 ) are accordingly graphed in figure 1.7 for England, France, and Sweden, the countries shown in figure 1.3. Values for Canada, the United States, and Japan are also represented for as far back as records go.

Comparison of the curves in figure 1.7 for e15 with those in figure 1.3 for expectation of life at birth (e0 ) makes it immediately apparent that the suddenness and steepness of the secular shift is much more pronounced with respect to longevity when expectation of life is reckoned for age 15. It is also evident that the experience of all the Western countries that appear in the figure has been very close, that England could indeed be taken for each of them without misrepresentation. Interesting, too, is the fact that the two North American, European-descended populations should follow the domestic European populations so faithfully but that their statistics are somewhat higher during the nineteenth century. The graph for Japan, however, though it follows a parallel course from a markedly inferior initial level, climbs from the lower to the higher plateau with an amazing steepness after the 1950s, almost vertically in fact, and rises above all the others. The aging trajectory for that country could be said to be something of a caricature of the secular shift as it has been experienced by Western countries, the extraordinary drop in the 1940s being an outcome of the Second World War. We shall return to these differences later on.

The generalizations that have been made about the secular shift, then, are emphatically confirmed for longevity by the use of the more realistic measure of life expectation at age 15. It might perhaps be asked why this indicator was not adopted at the outset. The answer has to be that it has a grave disadvantage in practice. Because the only statistic for longevity usually published for any population is expectation of life at birth, any other measure but this is very seldom likely to be available. For further discussion of this and other circumstances in the measurement of longevity in experiential


34

figure

Fig. 1.7.
Expectation of life at age 15: Canada, England, 
France, Japan, Sweden, and the United States.

terms, see the appendix to this chapter together with table 1.A1 and figure 1.A1. The outcome of our consideration of longevity to this point might have to be that we must continue to accept expectation of life at 0 as the standard measure but that this measure must be used bearing in mind all the drawbacks and obliquities that have been set out.

We are not quite at the end of the complications about the reckoning of longevity for comparative historical purposes. There is an important difference between what is called period or generation life expectation, which has been used so far, and cohort life expectation, which has not been mentioned.

Period life expectation is nearly always the one that is used. But it is a synthetic construct. In reckoning the total years still to live, which have to be averaged out among every member of the population in question, it assumes that the death rates being experienced at all ages during the year at issue re-


35

main the same over the whole time during which the people concerned will go on living. It is almost as if everyone had his or her whole life experience in the current year. In reality, of course, these mortality rates will undoubtedly change as the years go by, perhaps not by very much but by enough to make this synthetic calculation a hypothetical estimate. The enormous advantage of the period calculation, however, is that it can be done for any chosen year or, more usually, for any two or three years, provided only that mortality by age is completely known for the year or years concerned.

Cohort life expectation makes use of the actual number of deaths experienced by a cohort (persons born in the same year) from the time of its appearance until the last member of that cohort has died, that is, something like one hundred years later, or even more. Cohort life expectation, then, is about as realistic as such a statistic could be. But it has the enormous disadvantage that it can only be calculated after every member of the collection of people in which we are interested has ceased to exist. It is for this reason entirely impractical to recommend cohort life expectation as being more realistic and true to experience than period life expectation. Only for historians studying populations existing at least a century ago would cohort life expectation be of use. Though such a statistic could in theory be used for comparisons between past populations, all of which existed a century or more ago, it cannot be used for comparisons involving any more recently existing population. What is more, the necessary data for such calculations so far into the past only survive for a small number of historic populations. Nevertheless, cohort life expectation can usefully be calculated for some purposes on occasion, such as reckoning of the Third Age Indicator, which will concern us later on. It is available for English cohorts born between 1541 and 1781 and again for those born between 1841 and 1876.[27]

This brings us almost to the end of the considerations necessary in measuring expectation of life for historical purposes, though in the appendix to this chapter we shall look at yet a further possible indicator. This is the age at which there are a given number of years still to live (say, two years more or five years more), an age that differs in a most interesting way between the Before and the After. It goes down rather than up during and after the secular shift (compare especially Bourdelais 1993). As with all measures, that which is the most satisfactory, or perhaps in this case the least unsatisfactory, for general purposes may not be of much use for particular purposes. In describing the emergence of the Third Age as an outcome of the secular shift, a somewhat different statistic from that of e15 or of expected years still to live will be suggested in relation to longevity. This will be the chances of reaching age 70 from age 25, the Third Age Indicator, or 3AI. It would be possible to argue that from the point of view of the interests of most students of aging—and certainly of most of the contributors to this volume—the 3AI, reckoned either in the period or the cohort mode, might


36

be even more realistic as a general measure than expectation of life at age 15, and closer to the experience of past people. I shall suggest later on that a low level of the 3AI might have disposed those people to write off the possibility of ever becoming old. These points are illustrated and enlarged in the appendix to this chapter.

Interesting as this possibility, might be thought to be, the major concern of those who now study aging is in the numbers, experience, condition, and prospects of those already in late life and their relations with their juniors and their juniors with them. It is this concern that informs the suggestion of a revised measure of proportions in late life, our second aging variable, a measure more realistic than a simple fraction of older persons in the whole population. The realistic, experiential measure suggested here is the share of those over 60 of all adults, of all those over 25, in symbols 60+/25+. A more detailed impression of the significance of older persons with respect to their numerical size in relation to that of other age bands, together with changes over time in such relationships, can be gained of course from figure 1.6.

This realistic age-proportional measure, relative weight of the elderly and old among all adults, passes over an even greater number of the nonadult, the immature, than the realistic longevity measure just discussed, life expectation at age 15. The justification for doing so is just as strong in my view, or perhaps even stronger. When attempting to sense the presence of those in late life in a society, it is not easy to see why every individual member of that society should weigh as much as every other. The young and very young are undeniably of significance in the social structure. We in our own day are not the first body of people to be conscious of the immature and maturing members of our society and of the crucial character of our relationships with them and theirs with us.[28] The young and the very young are also of evident importance because they are largely dependent and their numbers and proportions must be known to study support relationships and transfers between age groups. In this respect, the young are in a position very similar to that of the dependent old. These are highly significant examples of generational interchanges and of age structuring. Where changes in the proportions of age over time are accompanied by changes in the flow of support, with the result that some cohorts are privileged over other cohorts, a fascinating set of issues to do with intergenerational justice comes into view.[29]

These issues can certainly be classed as consequentially related to the historical demography of aging, but relationships as to aging experience between the various age groups are scarcely affected. Figure 1.8 traces the course of the ratio recommended, proportion of all adults who are over 60, for the four hundred fifty years of known English aging experience and the two hundred fifty years of French and Swedish experience, along with the briefer periods observable in Canada, the United States, and Japan. In my


37

figure

Fig. 1.8.
Proportion of adult population (i.e., all those over age 25) over 
age 60: England, France, Japan, Sweden, and the United States.

view, it presents the clearest and most revealing comparative numerical account of the weight and importance of the elderly over the stretches of time which it is possible to observe and complements for aging experience of this kind the results we have just surveyed for the experience of longevity. The figure includes, it will be noticed, a line for the threshold of the Third Age, which, along with the 3AI itself, will be discussed below.

The features that stand out in figure 1.8 are once more the similarity with regard to the abruptness and the shape of the secular shift to what is to be seen for all the populations included in figure 1.3 and the somewhat greater unevenness of the lines, none of which shows quite the same smooth progression that marks the long English graph for proportions over 60 in the whole population portrayed in figure 1.2. The pronounced peak in the English profile in figure 1.8 during the early eighteenth century is somewhat disconcerting, since it reaches a height not seen again until the secular shift was well under way in the 1930s. At 22.4 percent, the proportion that those over 60 made of all English adults in the five-year period 1706 to 1710 was over two-thirds of its level now (1991 = 31.2 percent). It might be thought that the share of the elderly in the population was not as constant as has been made out, or that the experience of the English was less than representative.

Closer examination shows, however, that more modest peaks of this kind are present in the graphs for the other populations represented and that the general resemblance to the lines in figure 1.2 is quite pronounced. Once again the relative shortness of the lines in the figures, the woeful lack of temporal depth in our data, makes judgment difficult. Inspection of figure 1.6 and of the quinquennial figures in table 1.3 reveals that, during the


38

early decades of the eighteenth century, proportions over 60 were at their highest in the whole English series before the 1930s and that the age band 25-69 fell into a trough at the same time. Though neither movement was particularly sharp, their coincidence seems to have produced the effect we are examining.

The adoption of a more realistic, experiential measure for proportions of elderly scarcely confirms the arguments that have been presented using the whole population as a divisor in quite the decisive way in which the substitution of expectation of life at 15 does for longevity. The rise during the secular shift in the proportion of adults who were over 60 was only 57 percent in England. But the dissonance cannot be said to affect in any great degree the theory of relative constancy on the lower plateau, the abruptness of the secular shift, or of the general typicality of the English data. The reexamination of the aging processes and particularly of the secular shift by the use of experiential measures is highly illuminating, nevertheless. We shall find this point confirmed as we take up an illustrative example in the aging history of two neighboring European countries whose comparative development has preoccupied us so much.

Aging in France and England in the Eighteenth and Nineteenth Centuries: an Illustrative Example

The idiosyncratic demographic behavior of France has come up several times, just as it tends to do in the discussion of the demographic transition itself. It is of significance for understanding the secular shift in aging and the emergence of the Third Age for the following reasons. If it could be shown that French longevity, during the period when it is accurately known for the later eighteenth century and nineteenth century, along with proportions of elderly rose to levels that have been claimed never to have been reached before the secular shift was well under way in the twentieth century, then the assertion that there was a sudden upthrust from plateau to plateau would be less persuasive. So also would be the case for assuming an overall constancy in aging in the Before and, by analogy, in the After.

Perhaps more important would be the effect on the claims made here about the uniqueness of our own contemporary, developed societies. Experience of aging in a very important European nation during the eighteenth century might have anticipated to a significant extent the experience that has been described as confined to the developed countries at the present day, though eventually to be shared by all societies. Since what happened in France might well have happened elsewhere in the illimitable time space of the Before, then it would be likely that all social structures along with their constituted mentalities would have a resilience in the face of the effects of pronounced aging built into them. The secular shift, though the very ex-


39

treme of change, could perhaps have been accommodated by the traditional social structure and so not have led to that incoherence which, as has been claimed, marks the very old, developed societies of today.

Alternatively, if it could be assumed that France in the eighteenth century was the first country ever to have a foretaste of what was to come two hundred years later, then the French attitudes and practices at that time would have a peculiar significance in the history of aging. They would have taken into account the aging experience not simply of a restricted locality, or that of a select group, but of a whole national society, at that period a national society preeminent in the life of the Western world. During its Grand Siècle, France was intellectually dominant in Europe, the center of the Enlightenment yet the seed bed of that revolutionism that was to shatter the political, cultural, and religious framework of the traditional European world.

Though no commentator on age and aging in France in the eighteenth century has made statements as positive as these, a recent authoritative work by a historian of aging, David Troyansky, Old age in the Old Régime: Image and experience in 18th-century France (1989), certainly hints at them. He insists, quoting Pierre Chaunu, that the lengthening of human life was "the only great event of the 18th century" (14) and sees this demonstrated in the literary and cultural life of France at the time. They reveal in themselves, he seems to be maintaining, that France was then already exceptionally old. The scene changes for the nineteenth century, when French supremacy in these directions had disappeared and British industrialism had supplanted the Enlightenment as a hallmark of national success. Then it was the French statesmen and wiseacres of the time rather than subsequent historians who interpreted what they saw in a very different sense. They tended to blame French decline after the revolutionary and Napoleonic interlude, the military reverses and economic sluggishness, on the drag of excessive numbers of the old in the population.[30] The object of this discussion is to show how easily interpretations of this kind could have arisen when the historical demography of aging was of an entirely unsystematic and impressionistic character and the comparative historical evidence for the secular shift in aging had still to be worked out.

Now that we have something like principled, if often inferential, knowledge of aging in Western countries over the two centuries with which we are here concerned, it is not difficult to see how ill-founded such judgments must be for France during the century of the Enlightenment. The simple juxtaposition in figure 1.3 of the French trajectories with those of England and Sweden shows such judgments to be improbable, and the graphs in figures 1.7 and 1.8 should dispose of them altogether. When the records of that country began in the 1740s, France was catching up to England and Sweden in longevity (expectation of life at age 15) but never clearly surpassed them, and the rise was not exceptional but of the same order as the fluctuations to be seen in longevity over the whole length of English record-


40

ings. Meanwhile, our more realistic indicator for the proportion of elderly among French adults behaved in rather the same way, except that it did go above the English level between 1800 and 1850 and was no doubt very high by European standards throughout the nineteenth century. But Swedish proportions were rising even faster and were well above the French by 1900, as were those of other Scandinavian countries. These circumstances for the French add particular significance to the possibility, suggested by recent work at Cambridge, that it was in the early eighteenth century that the longevity of adults amongst the elite began to differ from that of the masses in the Western populations.

If we were looking for an exceptional situation to justify an exceptional attitude arising from longevity and the relative weight of the elderly, it would not be in France over this period that we should be seeking it but rather in England. The population of that country was near or at the top of the scale of longevity in Europe from the 1740s to the 1940s and after the 1800s had an exceptionally low proportion of adults who were elderly. But this situation was certainly not to be classed as an anticipation of the present day, since while it continued the two aging indicators had opposite tendencies, high longevity and low proportions of elderly.

The imperfections of this illustrative example for the history of aging and its historical demography have to do with the methods and sources of historical sociology as well as the objective reality of the secular shift. Important and illuminating as may be the literature current in the past on the subject of aging and the elderly in the history of society and of culture, to seek in such sources indications of longevity in earlier times or of the numerical weight of the elderly in the adult population is a misdirected enterprise. It is particularly unfortunate, even potentially disastrous, when evidence of this kind alone survives to provide any indication of longevity or of proportions of elderly persons. The facts are, as we have seen, that such was the position as to knowledge of aging at almost all locations and points in time before the last two or three centuries of Western history.

With the rather uncertain exceptions cited earlier, such writings as the Old Testament, the poems of Homer, or the philosophical essays of medieval canonists and Renaissance philosophers contain practically all the evidence available. Under such circumstances, the proper policy for the historian is surely to make no attempt to use literary allusion as positive evidence for age and aging. The assumption has to be that the society in question must have been located at some level on the lower aging plateau, with the characteristics set out here for the Before and derived to a considerable extent from the theory and practice of demography in relation to primitive and developing societies. Textual allusions in such sources to growing old and to old people have to be taken as illustrative or complementary: their relationship to the demographic situation as it actually occurs can never be a reflection of reality.[31]


41

The title with which this discussion began, Troyansky's Old age in the Old Régime , is an attempt not so much to recover from cultural sources the facts about aging as to confirm from all the indications the author can lay his hands on that old people actually existed in some numbers in eighteenth-century France, so as to drive home his claim for the importance of aging under the Enlightenment and the Ancien Régime generally. It might be supposed this is something that any knowledge of the historical demography of aging would make unnecessary. Troyansky goes to considerable lengths in this enterprise, citing a rise in the proportion dying after the age of 60 or 70 in nine villages whose families have been reconstituted by French demographers; an increase in the age at death of French bishops; an inaccuracy in the low figures worked out for adults at Caen in 1988; a growth in the economy and its general capacity for supporting the dependent elderly; and many other circumstances. He even cites figures for proportions of adults who were over 60, a ratio similar to that used here, but using age 20 rather than age 25 in the denominator. It has not been possible to cover all the bewildering array of arguments he presents, arguments that sometimes suggest he has not yet got across the pons asinorum of aging. It should be clear from figures 1.7 and 1.8 that France only became exceptional with respect to the relative weight of elderly persons during the nineteenth century and that it was never exceptional in longevity before the mid-twentieth century. The disposition of historians to advance a confusing medley of numerical and seminumerical arguments of this kind has to be described as an obstacle in the way of the understanding of age and aging in the past.[32]

The Secular Shift in Aging: Its General Historical Position and its Outcomes

Industrialization and Modernization

The chapters in this volume can be read as a commentary on the results of the secular shift in aging, results that are shown to have been brought about during its historical course at particular places and for particular areas of activity. This is evident in Richard Wall's essay on the elderly in England and those with whom they resided, comparing the preindustrial past with the industrial present, the lower with the higher aging plateau in the language we have been using. Wall makes repeated references to the aging process, nearly all of them to rises in longevity, as does George Alter in analyzing railway pensions, though he refers to declining longevity as well. The other contributors make similar allusions, and it is evident that the secular shift is seldom far away from the changing situations that are being described, even when the subject is American home ownership or the elderly on the Texas frontier. The descriptions of facets of life as it was experienced by older persons at sites in traditional Europe—Hungarian experience in the case of An-


42

dorka, Latvian in the case of Plakans and Wetherell—seem deliberately aimed at conveying an impression of the position as it was in the Before. Some notion of what changed and what remained the same, a critical part of the task of historical sociology, can be gathered in this way.

Understandably, perhaps, considering the shortcomings of the available data that have had to concern us so much and the difficulties of the analysis, almost nothing is said in these chapters about the pace or extent of population aging change and their consequences. This introduction was not in the hands of the writers, and no compendium of the relevant demographic facts was available. None of our authors goes any way toward an estimation of the overall results of aging in Western societies, or of the historical situation and significance of the secular shift. Insofar as that estimation is yet feasible, it is very briefly attempted here, particular reference being made to such general outcomes as the transformation of the life course, the redirection of social transfers, and the swaying of the balance between the sexes leading to the feminization of the elderly population. Finally comes what in my view has to be called the most important of the outcomes for the present and future of developed societies and the future of societies yet to develop, that is, the emergence of the Third Age.

It cannot be expected that so profound a change in social relations and in the life of individuals should be adequately comprehended so soon after its crystallization as the present day, if indeed this change can yet be regarded as complete. This is particularly the case if, as has been claimed, the transformation itself has yet to be recognized and accepted and if even those who run the political and administrative systems of the advanced societies and determine their attitudes persist in a state of lagged awareness about the aging transformation. In the late 1980s, British Parliamentary candidates actually supposed that the way to get in touch with older voters was to go to the institutions where the infirm and dependent elderly lived "in spite of the fact that 96 percent of older people reside in ordinary housing."[33] No better case could be made for the historical demography of aging being necessary knowledge.

Singling out the effects of the secular shift is made peculiarly difficult because of the other major movements taking place during the century or so when the shift occurred. In most but certainly not all the countries, this was also the time of industrialization, of social and political mobilization, and finally of procreative transformation, the time in fact of all those movements that usually go under the unsatisfactory rubric "modernization." The particular results of radical aging have furthermore to be distinguished from those of other demographic changes to which it is causally connected, continuing shrinking fertility, perpetually lowering mortality, and radical change in marriage, all supposedly embraced within the overall "modernization" model. This is not the occasion for dwelling on the difficulties and


43

the distortions, especially the historical distortions, that any such overarching concept as the modernization model is bound to bring about.[34] But certain distinctions can be made here.

The secular shift in aging was by no means necessarily associated with the other changes being mentioned and is not so today. In some important countries, in Germany and Italy, for example, or the Scandinavian countries and Spain, the phase of intense industrialization did coincide with the early part of the secular shift. But these instances can be misleading, especially when they are seen as historically parallel with what is happening in the now-developing countries.

Industrialization in Britain occurred wholly on the lower aging plateau and in that country, modernization, insofar as the concept is applicable at all, was complete before industrialization began. In France, as we have seen, the secular shift was considerably more diffuse over time than elsewhere. In China as well as in other East Asian countries, as we shall see, the shift is already in progress. The risk in China and especially in the really laggardly countries is that the shift will proceed so quickly that economic development will be behindhand in providing means of support for the hugely expanding elderly population.

It may turn out indeed to be a fortunate historical chance for the Western countries that they have not experienced the secular shift under the conditions and at the pace that seem likely to characterize its trajectories in other areas of the world. In short, an analysis of the historical position of the secular shift leads to the conclusion that it has been independent of these other changes, though causally connected with some of those that are clearly demographic. The interconnections supposed by the modernization model and its theory of convergence seem not to be evident, in spite of the difficulty of disentangling aging from those other threads in the skein.

Age Structuring and the Life Course

Some of the outcomes are fairly straightforward nevertheless and have been hinted at as the shift has been described. The transition from a position of perduring constancy in life expectation and in age structuring to one in which family members live in company with each other for longer and longer and age groupings undergo radical disturbance has upset the relationships between age groups. It has changed the life course of individuals along with the developmental cycle of the family and of the domestic group as a whole. The immemorial association in family time between the exit of the parental generation and the beginnings of the independent family life of the child generation has gone forever. The consequent proliferation and prolongation of the situation known as the empty nest stage, in which the parental couple stays together in the home after the departure of their children, a situation so familiar in the contemporary West, are almost exclu-


44

sively the result of aging. The accompanying phenomenon, the rapid rise of the primary, individual, that is, single persons living alone, has causes other than aging, though this must be mainly responsible. Both existed in the Before, but it is probably the secular shift that has made them major characteristics of late-twentieth-century Western social structures.

As part of this development, age at inheritance has risen inordinately.

Long-expected one-and-twenty
Ling'ring year, at length is flown:
Pride and pleasure, pomp and plenty,
Great *** ****, are now your own.

Samuel Johnson's overfortunate youth, deliberately unnamed, be it noted, whose father had evidently died long before the great day, would be more like 58 or 60 years old at inheritance in contemporary Britain or America.

The accompanying elongation of vertical kin links, great-grandfather, grandfather, father, son, grandson, great-grandson, or even great-great-grandson, is entirely due to the secular shift. It reminds us that the populations of contemporary advanced societies, unlike any of their predecessors, cannot be envisaged in pyramidal form with respect to the distribution of age groups. They are more like rectangles posed on their shorter sides. The shift in flows of support between generations, particularly the appearance and spread of multiple dependency, in which a woman has at the same time to attend to the needs of her children or grandchildren and to those of her parents or even grandparents, is due to the same cause. This whole galaxy of consequences of aging has been brilliantly described by Michael Anderson but in a strictly preliminary fashion, nevertheless, for it is scarcely possible yet to know how many changes of this character have been, are being, or will be brought about.[35]

The least specific but still extremely important possible outcome of the secular shift has still to be touched on. This is the possibility that the growth in the relative weight of the elderly in the population—the ever-increasing number of people with halting gaits, bent shoulders, lined faces, and shapeless figures—brought into being a hostility and even a hatred that was absent or less prominent in the Before, became strident during the shift itself, and persists in the After. This is a topic of considerable controversy, even of contradiction, and the interplay between demographic causes and cultural influence is especially hard to unravel. For it has to be insisted, as Daniel Scott Smith has repeatedly pointed out, that there was cultural change with respect to the elderly during the secular shift which went on separately from aging itself: mere demographic determinism is quite unacceptable in trying to understand what happened. There is a good case for supposing that there was a change for the worse for elderly British people in the later nineteenth and twentieth century, and historians of aging in the


45

United States are at odds as to when a similar but earlier change occurred in that country, although they seem convinced that it was an objective fact.[36]

Uncertainties such as these can only be registered here and remain for historians of aging to investigate further. We glance now at two sets of changes much more securely attached to lengthening longevity and population aging: feminization and retirement.

Effects on the Sex Ratio and on Retirement.

As is evident from figure 1.9, comparing the growth of proportions of elderly in England and France during the secular shift in aging, the process altered numerical relationships between the sexes in later life in a conspicuous way. The widening of the gap between female and male proportions of elderly persons is evident in France from the 1820s but came rather later in England. This effect intensified as the secular shift proceeded and has continued until the present day. There is some indication in table 1.2 that in the southern and eastern European countries, the sex ratio of the elderly had been more equal on the lower aging plateau just as it often is in developing societies in the contemporary world. In 1880, there were more old men than old women in Italy,

figure

Fig. 1.9.
Proportion of population over age 60: England
 and France, 1770s-1970s. Adapted from Laslett 1977.


46

Greece, and Bulgaria, as was the case in the eighteenth-century Hungarian village reported on by Rudolf Andorka in his contribution to this volume. The discrepancy in proportions lasted in Bulgaria into the 1920s, but at none of the dates in table 1.1 was expectation of life at birth higher for men than for women, close as the figures sometimes get. By the 1950s, the whole of Europe seems to have been fairly uniform as to the excess of women in the higher and highest age groups and as to their superior longevity.

We should not infer that this growing imbalance was caused by aging in a billiard ball sense; rather, the two accompanied each other, and living longer or being allowed to live longer gave women an opportunity, so to speak, to demonstrate more and more their superior capacity to persist in the face of natural hazards. However this may be, we have to recognize that in coming to terms with our age-transformed condition, we shall have to provide for a more and more feminine society of those in late life, with no tradition in our culture or in any culture to give us precedents or inspiration. We know little about the sex balance among senior people in the West before the secular shift, though in the small and unreliable samples analyzed for England (see Laslett 1977: 201-208), women already predominated in the seventeenth and eighteenth centuries, if not to the same degree that they do today. It would seem that neither in the Before nor in the After have social arrangements in this respect taken the facts of demography properly into account, though we are obviously less realistic than our forebears. The same indifference to, even defiance of, demographic "reality" is to be seen in the equally important matter of retirement.

Nothing whatever in the changes in age structure over the secular shift as portrayed in the figures could be said to justify the rapid and continuous withdrawal from the workforce at earlier and earlier ages that has marked recent Western social history. We have seen that the proportion in the working ages (15-59) was quite stable over the secular shift in England, though the mean age of the workforce has risen and will rise. As that change nears its end in the West, there is no surplus of older workers in the demographic sense in the way that there is a surplus of elderly women. Although account has to be taken of actions of government and employers in getting senior persons off the rolls of the employed, it is tempting to interpret retirement as a move by individual people in spite of demography, the assertion of a wish and a right to take advantage of lengthened life by changing the use of time. Above all this, of course, is the immense growth of wealth that was proceeding in all Western countries contemporaneously with the secular shift and that is doing so on an even greater scale in present-day Japan.[37]

The growth of retirement in spite of the lengthening of later life, like the growth we shall observe in solitary living in spite of greater opportunities for elder people to reside with their married children, illustrates the fact that demography does not always have the expected effect on life-course transi-


47

tions, that fascinating topic handled in several of the chapters in this volume, particularly by Tamara Hareven and Peter Uhlenberg. We can go no farther in this direction here, but we must consider for a little while the familial effects of the shift, since the familial position of the elderly and old has been such a preoccupation of those who have pioneered the history of aging and who are well represented in our collection.

Familial Effects.

It might perhaps be expected that larger proportions of older people and the better survival of parents and grandparents would have led to an increase in multigenerational households all over Europe despite "modernization" as the secular shift proceeded, certainly in its early stages. It would seem, however, that this expectation is in general unjustifiable and that the effects were complex and multidirectional. Research has established (Wall 1989) that European countries vary very widely with respect to the living arrangements of their elderly citizens at the present day. There is no reason to suppose that this variation was less in the 1880s. It has been shown for England that while that society was on the lower aging plateau, independent living was quite a normal thing, though not very widespread for elderly people, even if they had married children with whom it was possible to co-reside. The tendency to live on their own is scarcely surprising in the European areas known to historical sociologists as maintaining neolocal household formation rules, that is, mostly the countries in the north and the west of Europe. In a nuclear family formed on such rules elderly persons have no right to join the families of their married children, any more than those married children have the right to rejoin the families of their parents. But it was always open to children to invite their parents into their households out of affection or because of failing health, or simply from a sense of duty. This evidently frequently happened. In southern and eastern Europe, where household formation rules were not the same, the secular shift may have progressively increased the proportion of multigenerational households, at least initially, but of this we have as yet little knowledge.[38]

In all developed Western countries, however, there has been the movement we have mentioned several times, one which is distinct from the secular shift and opposite in its impact on family forms, a tidal drift toward solitary living at every adult age, especially at late ages. In Asia, where about three-fourths of elderly persons in the 1990s live with their married offspring, the position has been very different.[39] It can be seen how intricate are the questions that have to be faced if the Western countries are to meet the challenge of providing precedents of use to those who will follow them on to the higher aging plateau.

There can be no doubt that the increase in the numbers of elderly, potentially dependent parents in European societies which came about as a re-


48

suit of the shift has led to an intensification of kin interchanges, particularly between mature children and their still-surviving parents. It is fascinating to see in a recent study of an industrializing town in the United Provinces where the population registers were kept how these interrelations grew more frequent and important in the later nineteenth and early twentieth century and how the domestic group itself frequently changed its structure in reaction to these developments (Janssens 1993). It can in general be said, however, that the final stage, the era of the older married couple living alone after the departure of their children and of the elderly solitary widow, widower, divorcée, divorcé, spinster, or bachelor had already begun by 1950.

Yet it is notorious that the image of the family group in the eyes of Western journalists, politicians, advertisers, and citizens themselves obstinately remains that of the married couple with two or more young children living with them. The great and growing numbers of elderly people living alone and the virtual absence of three-generational households have still to penetrate the consciousness of Western populations and to be recognized as a reality by Western politicians and intellectuals.

Effects on Social Transfers.

The secular shift, wherever and whenever it occurred, must have moved to some extent the burden of support for those at the working ages from maintaining large numbers of children to maintaining increasing numbers of the elderly. It seems unlikely that these changes would perforce have increased the total burden of dependency appreciably, since in most societies at most times dependent elderly persons are no more costly than dependent infants and children. We have just seen that support to the elderly in need was not necessarily given in the way of maintenance in the home, and it has now to be added that an appreciable proportion of that support, in England and perhaps other countries, took the form of transfers through the collectivity.

By collectivity is meant not simply the state but all agencies of social transfers that cannot be described as familial. The introduction of pensions in Bismarck's Germany and in Britain in 1908 marked an entirely new era in the extent and regularity of central state support for persons quitting the workforce. But the collectivity in European countries, evidently to varying degrees between country and country, had always taken rather more of a responsibility for demographic casualties such as failing elderly people than was realized until results of research on the point began to become available.[40]

The traditional English Poor Law has been shown to have been comparatively generous to the elderly, providing something like pensions to widows as early as the seventeenth century, even continuing that support under the much more parsimonious New Poor Law of early Victorian times. In addition to these sources, transfers to the dependent elderly from towns and charitable institutions of all kinds, together with a system of elderly privileges, such as licenses to beg, to glean after the harvest, to ransack the hedgerows


49

for firewood, or even to pilfer, had been present in the traditional European social structure from time immemorial. There were multifarious arrangements for dealing with the incapacitated elderly, often agreed between the family and the collectivity, circulating them between households—of children, of kin, and of neighbors—and putting them into institutions where such existed. Nevertheless, transfers of this kind were unstable over time, plentiful at one period and scarce at others. There was evidently much ambivalence. But the truly decrepit, those entirely unable to care for themselves, could count more reliably on collective assistance. In England and probably elsewhere, it is an interesting fact that about the same proportions of old persons lived in institutions before the secular shift as do so today.

It seems clear, however, that the effect of the introduction of state pensions was to reduce the proportion of transfers reaching the elderly from these miscellaneous sources. After the middle of the twentieth century, very high proportions of the support of working-class elderly people began to come from the sole source of a state pension, a situation that seems to have become widespread elsewhere in Europe. It can certainly be observed among the workers of London in the 1930s.[41] The point of importance is that the increase in the longevity and numbers of older people at the secular shift elicited collective support on a larger scale but certainly did not bring it into being.

Although usually assigned to the invention and expansion of the welfare state, more recent scholarly opinion is that universal pensioning at a steadily increasing level of those past work in twentieth-century Europe is perhaps better regarded as a continuation of traditional policies into a time when the enormous increase in disposable wealth made it possible to contemplate transfers on such an unprecedented scale, even to a rapidly expanding and increasingly older population. It would seem that given in this way the means to live independently, older people readily did so. There were other effects as well. The state pension has been shown to have led, for example, to the decline of the responsibility felt by factory and other employers for members of their workforce who had grown old in their employment, providing not simply occasional pensions but also a scale of diminishing demands in the way of effort, by assigning workers near the end of their time to such tasks as being night watchmen.[42]

By the middle of the twentieth century, therefore, the secular shift had brought those retired from the workforce into direct contact with the nation state, in societies whose ever-increasing riches made it possible to support persons in later life at something like the level of workers in their active earning years. Except among the proletariat, savings over the life course, however, were surprisingly common even before the secular shift began, particularly in the United States, and were much enhanced in the decades of economic growth in the earlier twentieth century. Along with state support these resources gave to those in late life a prospect of per-


50

sonal independence on a mass scale quite unprecedented in the history of the old. Such were the financial prerequisites for the emergence of the Third Age in the fullest sense of that term, and they stand in exquisite contrast to the position as it was in all countries when they were on the lower aging plateau.

Nevertheless, dependent persons are always relatively poor, and dependent elderly proletarians in traditional societies like those of the preindustrial era in Europe are the poorest of all. The impression must not be given that the modestly placed 70-, 80-, and 90-year-olds of the past were ever anything but badly off in comparison to their juniors, and the historian is often puzzled that they managed to survive in the numbers that they did. In rich industrialized societies, as everyone should know, the working-class elderly are still wretchedly poor in comparison to the rest of the population. They are relatively poorer probably than the old persons of preindustrial times and getting relatively even poorer, at least in Britain. The outcome of this composite situation has been to create for contemporary citizens of developed nations a conviction that the now enormously expanded and expanding society of retired persons is an economic problem for themselves and a burden for posterity. They find it difficult to recognize that in the emergence of the Third Age their own personal future is being provided for in an entirely novel society.

The Emergence of the Third Age

The definition of the stage of life which in European countries has come to be called the Third Age is still uncertain. The term belongs to a division of the life course into four ages, put forward in Britain during the 1980s, which is in increasing use there and in Australia but not so much in the United States and Canada. Many circumstances have to be taken into account in determining the character of the Third Age and in maintaining its independence of birthday age. The time of retirement from a job marks its beginning for most people, and for them the Third Age will last until death, or the onset of the Fourth Age.[43] Unless that interlude is substantial, the Third Age cannot come into being. The existence and development of retirement and the extent to which transfers and savings are adequate to make a life of self-realization possible for the retired are crucial circumstances. These, in turn, depend on the wealth of the society to which individuals of later life belong and on such things as the level of instruction, the numbers in education, and the availability of education over the whole life course. In this way, therefore, the emergence of the Third Age is inseparably bound up with national development and with political and social policy.

More fundamental perhaps than retirement from a job, especially for women, is the change in personal life that comes with the final departure


51

of children from the parental family and the arrival of what has already been termed the empty nest stage of the development cycle of the family. There has to be a sense of release from the responsibilities and trammels of the Second Age and a recognition of the opportunity to live a life of self-fulfillment thereafter. It goes without saying that the physical condition of the retired person has to be good enough for her or him to seize that opportunity. Judging this circumstance depends on the reconstruction of the history of health over the life course. The assertions on this topic made by several of our contributors, especially by George Alter, hint at one of the most important of the tasks of the historical sociologist of aging. Though more usually justified by the intimidating growth in the cost of invalidism in late life, a theme not pursued here but evident in practically all of the discussions of the subject, its relevance to the emergence of the Third Age of independence and creativity is clear.[44]

The general theory of the emergence of the Third Age and of its relations with the Second and Fourth ages goes as follows. Though elements of such a life stage can be described during the Before, it could not properly exist then because its essential preconditions were absent, especially the demographic. The mass of individuals went without much in the way of an interlude between active life and old age, and old age was defined in terms corresponding largely to those we have used for the Fourth Age.[45] In the absence of widespread and compulsory state-guaranteed retirement, the transition was not dictated by calendar age and would vary with family and generational position, with social class and occupation, with the economic situation, but above all with physical and mental condition. There is ample evidence of this in the essays published here in Part IV, Retirement and Mortality.

Prior to all these necessary features of the Third Age, therefore, are the demographic conditions that permit a large enough proportion of the population to live for long enough after the Second Age is over. It is the theory of the early discussions of the Third Age and its emergence (Laslett 1984, 1987a , 1989) that all the circumstances named, along with others that have to be passed over for lack of space, began to exist for the first time in most of the countries of Europe and in countries with populations of European origin by the middle decades (1940-1960) of the twentieth century, though the countries of eastern Europe may have to be excepted. The demographic qualifications have had to be decided in an entirely arbitrary fashion, and the suggestion is that the Third Age can only be present when at least half of a country's male population can expect to survive from age 25 to age 70 and when at least a quarter of adults, those over 25, are beyond the age of 60 years.

As for the survival condition, in life table terms, numbers alive at age 70 divided by numbers alive at age 25 has to be equal to one-half or more (170 /125> .5 for males). This is called the Third Age indicator, or 3AI.[46] In table 1.4, values of the 3AI are given for England at fifty-year intervals from


52

TABLE 1.4
Demographic Qualifications for the Third Age, England and Britain, 1540s-1990s

 

figure

 

Number over Age 60b

Year

Number over Age 25

1541-1545

.292

19.19

1591-1595

.348

18.43

1641-1645

.325

18.67

1691-1695

.310

20.53

1741-1745

.297

17.87

1791-1795

.338

16.90

1841-1845

.391

15.25

1881-1885

.374

16.47

1891

.301

16.21

1901

.375

15.42

1911

.416

15.66

1921

.492

17.23

1931

.497

19.66

1941

.471

21.82

1951

.532

24.50

1961

.581

26.77

1971

.625

30.08

1981

.634

31.73

1991

.690

31.21

SOURCE : Calculated from data in the files of the Cambridge Group for the History of Population and Social Structure.

a Males only.

b Females and males.

the 1540s to the 1880s, along with proportions of adults over the age of 60, and at ten-year intervals thereafter until 1991. The fundamental differences between aging on the lower and higher plateaus are visible once again in the table, as is the complex relationship between the demographic transition and the secular shift. It will be noticed that the 3AI varied narrowly above and below .33 until 1901, a one-third chance for a male at age 25 ever attaining age 70. By 1951, however, the 3AI for men exceeded .5 for the first time, and proportions of all adults above the age of 60 exceeded a quarter for the first time by 1960. England (and Britain) had qualified demographically for the Third Age, on the criteria we are using, almost exactly at the midpoint of the twentieth century.

Climacteric in the Middle of the Twentieth Century

It would not be expected that all the developed countries should comport themselves in the same orderly fashion. It can be stated, nevertheless, that all the populations of the countries appearing in tables 1.1 and 1.2 had a


53

3AI for males of at least .5 by 1960 and some of them several decades earlier, but that none of them had a quarter of all adults over the age of 60 before the 1940s. Reference to figure 1.8, however, shows that within that decade or its successor, every population represented except Japan had risen decisively above the line denoted there as the qualifying level. In contrast, figure 1.10, with the further detail given in figure 1.11, shows that the dates by which countries qualified with respect to survival were spread over a much longer period. Sweden and New Zealand had a 3AI of .5 just after 1900; Denmark, by the 1920s; and the Netherlands and Italy (not in the figure), by the 1930s, still twenty years before the accession of the United Kingdom and thirty years before Japan.

It is only, therefore, when the course over time of our realistic, experiential indicator of the weight of older persons in a society is taken alongside the course over time of the 3AI that England, or the United Kingdom, can

figure

Fig. 1.10.
3AIs (males only): Canada, Denmark, England, 
France, Japan, New Zealand, Sweden, and the 
United States. Data from National Life Tables. 
3AI = 170 /125 .


54

figure

Fig. 1.11.
Detail of figure 1.10.

still be said to have had the standard, the paradigmatic, experience of the secular shift in aging. The accession of that country to the Third Age in the 1950s can, under these highly specific conditions, be said to have marked the movement of the whole body of the rich, highly developed countries of western and northwestern Europe, with their companions overseas, into the new demographic era where the Third Age can exist. The most significant of all the outcomes of the secular shift in aging had unmistakably manifested itself in the West.

Yet another tilt of circumstance occurred in England in the 1950s. It must have been in this decade that the almost vertical upthrust of the secular shift began to bend over toward the higher aging plateau, problematic as that plateau may turn out to be in the next century. The demographic history of the other Western countries appearing in the figures we have been at such pains to compare with the English would most probably show this same slight change of direction detectable in the drawing together of the decadal survival curves in figure 1.5 at corresponding points in their aging trajecto-


55

ries. Inconspicuous as this reorientation may appear to be, this presaging of the disappearance of the secular shift and the succession of the higher plateau were pregnant with the future of aging in the Western societies we have lingered over. It is surely correct to think of the 1950s as the time of the "Great Climacteric" for the West in the history of age and aging. It remains to be seen what will correspond to it in the history of populations that have yet to be caught up in the secular shift.

Age Trajectories Over the Secular Shift in the West and in Some Countries of East[*]

We have made use of the demographic statistics of Japan in the course of the argument, as a supplement and a contrast to those of the Western developed countries. We have also repeatedly asserted that the historical de-mography of aging in every country, and this means, of course, a collection consisting as to a vast majority of less-developed countries, will sooner or later follow the same course. The higher will succeed to the lower plateau in all of them, and the secular shift in aging will do much to transform their social landscapes.

Here we will make use of the Japanese data as well as data from one or two other East Asian countries to give substance to these propositions and to justify the statement made at the outset that the Western aging experience has a transcendent importance because it has happened first. With these additional facts in front of us, we shall proceed to a conclusion.

Because our sources for this section are published demographic statements, it will not be possible in this very brief exposition of such facts as have become known to us to keep to the experiential aging measures that have here been declared necessary for the comparative pursuit of the historical demography of aging. We shall have to content ourselves with the cruder measures we began with and which have been subjected to criticism. We shall proceed, moreover, in a summary fashion, by way of commentary on the six figures that follow.

In figure 1.12, two features have to be noticed. The first is that the trajectories of the growth in proportions of elderly (here those over 65 rather than those over 60) are measured on a scale of years from starting points placed thirty years apart, that of Japan from 1950 and that of China from 1980. This circumstance makes the second feature even more remarkable, that the two trajectories should follow each other so closely. This is in spite of the gap between their starting points, which implies that the shifts were initiated in very different world situations and very different local, North-

[*] Written with Zhongwei Zhao using the data and figures that he prepared. For an extended version, see Zhao in press.


56

figure

Fig. 1.12.
Change in proportion of population age 65 and older: 
China, from 1950, and Japan, from 1980. Data from
 Keyfitz and Flieger 1986; Wu and Du 1992.

east Asian, North Pacific situations. The enormous dissimilarities between the two countries—in size, in culture, in social structure, in history (especially recent history), in wealth, in degree of development—seem not to have mattered.

This parallel effect is certainly surprising and suggests a definiteness and autonomy in the secular shift that could scarcely have been inferred from the evidence of the other countries we have surveyed. It is true that the comparison is largely based on projected trends of population aging (from the fourth decade in the case of Japan and the second in the case of China). It is also true that such projected results are essentially determined by the demographic parameters that have been used, and they could vary considerably according to the assumptions being made. But many other projected results indicate that in China, the aging trajectory is likely to be very similar to that which has been and will continue to be observed in Japan unless unexpected changes in fertility or mortality occur to alter them.[47] Short of an entirely improbable set of circumstances, or of borrowings from one set of projections to form the other, the trajectories and their close rapprochement must represent a genuine phenomenon.

In figure 1.13, four trajectories have been added to those represented in figure 1.12. The South Korean case is based mainly on forecasts, while the European cases rely to a far smaller extent on such projections. It is difficult to decide precisely the time when each population embarks on the secular shift, but for the purpose of comparing its speed in various countries, a starting point is arbitrarily chosen from the trajectory of the growth in propor-


57

figure

Fig. 1.13.
Change in proportion of population age 65 and
 older: China, England and Wales, France, Japan, 
South Korea, and Sweden.

tions elderly. Here the point selected is that at which the proportion age 65 and over is close to 5 percent of the total population and thereafter rises rapidly and monotonically. Figure 1.13 shows that a rise in the proportions of those over 65 from a level of 5 percent to a level of 15 percent occupied about 150 years in France, about 115 years in Sweden, and about 90 years in England and Wales but about 60 years in China and Japan. In South Korea, if the projected trend continues after the year 2020, the increase in the proportion of the elderly may be even more rapid.

This conspicuous foreshortening of the time taken for drastic aging to occur has been noted before. In 1988, Naohiro Ogawa published figures showing that the years required for the proportion of the population over 65 to rise from 10 to 20 percent were 24 in Japan, 48 in Finland, the fastest European case, and ranged from 54 to 85 (Sweden) in the eight other European countries he selected (Ogawa 1988: table 8). All this implies that the secular shift has been very much briefer in Japan—about half of that in most northwestern European countries—and will no doubt be somewhat the same in China. The speeding up of the secular shift and the indication that it may intensify as country succeeds to country undergoing the shift in East Asia must be classed as singular phenomena.

We have dwelt a great deal upon the suddenness of the aging process in the West, on the extent to which it has been overlooked, and on consequent "false consciousness" about it. It seems that East Asia and perhaps other countries embarking or about to embark on the secular shift will undergo it with even greater rapidity. The Western precedent seems to be less help-


58

ful here. No one can yet tell what may happen in the way of social discontinuity and disorientation with regard to aging and age relationships after a shift of such rapidity.

A very similar reflection is suggested when figure 1.14 is added to the series, delineating comparisons in longevity. Here the time scale is a real one and the statistics observational. Once more the Southeast Asian populations have increased and are increasing their longevity far more quickly than ever the Europeans did, and they will outstrip them in short order if Japan is to be taken as precedent.

Figure 1.14 suggests something further (it suggests only, because the temporal depth is so shallow for the Asian populations). It may be—and the little historical work that has been done and that is known to us does nothing to contradict this—that expectation of life on the lower plateau in Asia, or in parts of Asia, was below what it was in Europe. There is a hint of this in the trajectory for Japan in figure 1.7. Early and near universal marriage for women, which we suppose to have been a widespread characteristic of

figure

Fig. 1.14.
Expectation of life at birth (females only): China, 
England and Wales, France, Japan, South
 Korea, and Sweden.


59

these countries, would certainly have made it possible for their populations to survive in the long term with lower expectation of life than has been found in the West in the past. We must be a little wary in comparison here, of course, because expectation of life at birth, which is all we have, may interfere a little with an objective contrast. But our knowledge of the plateau-like character of longevity over long periods certainly seems to imply that the rises in Japan, China, and South Korea are very unlikely to be a short-term variation in an otherwise generally even trajectory.

If this conjecture is correct, the lower longevity plateau in these countries may have been somewhat beneath that for Western populations but probably within its range of variation. We have as yet nothing to indicate the extent to which it was horizontal and flat. A further inference, therefore, that can be made from the figures is that the secular shift may start from a more modest level than that which obtained on the lower plateau in the West and so may have a longer slope upward. It could therefore take less time, travel farther, and be markedly steeper in these East Asian countries than it has been in Europe, America, and Australia. This would apply to both of our aging parameters, expectation of life and proportion in later life, and it has to be said that the first East Asian country to proceed through the shift, Japan, has displayed all these characteristics. The still more remote possibility that Western countries were perpetually somewhat longer lived than others on the lower plateau has no support known to us other than the indications of these figures. It is an intriguing possibility nevertheless.

Figures 1.15, 1.16, and 1.17 modulate the inferences we have made but certainly do not overset them. They are all in real time, over the period spec-

figure

Fig. 1.15.
Expectation of life at birth (females only): Asia,
 China, Europe, Hong Kong, Japan, Singapore, 
South Korea, and Taiwan, 1950-2020.


60

figure

Fig. 1.16.
Proportion of population age 65 and older: China, France, 
Japan, South Korea, Sweden, and the United Kingdom,
 1950-2050. Data from Keyfitz and Flieger 1986; Wu and Du 1992.

figure

Fig. 1.17.
Proportion of population age 65 and older: China, Hong 
Kong, Japan, Malaysia, Singapore, South Korea, Taiwan,
 and Thailand, 1950-2050.

ified. The trajectory shown in figure 1.15 for expectation of life in Asia as a whole during this interlude (taken from Keyfitz and Flieger 1990) does not suggest that the entire continent will have a longer-lived population than the West. The pattern of the lines in that figure points to East Asian exceptions, however, and hints once again that longevity, could have been less before the secular shift began than in England (see, e.g., the figures in table 1.3, above). It is important to note that all the individual countries and areas shown in the figure have by now stepped over the threshold of the 3AI.


61

Figure 1.16 shows Japan overtaking European countries in proportions of elderly persons and rapidly acquiring the oldest age composition yet contemplated. The conceivable long-term difference between the Asian countries and the West with respect to both parameters on the lower plateau— Asia being the younger—is difficult to appreciate in view of the difference in the measures in figures 1.16 and 1.17 for proportions elderly from those used earlier, where proportions over 60 were used rather than proportions over 65.

But if it is remembered that proportions over 60 are likely to be about a third as high again as proportions over 65, a small margin in favor of the hypothesis is apparent. Finally, figure 1.17 makes it clear that in spite of the probability that all the Southeast Asian populations represented will show a similar pattern of change during the secular shift, the present temporal relationship between their trajectories is somewhat complex. The closeness of the observed and projected trajectories for Thailand and Malaysia, South Korea and China, Taiwan, Hong Kong, and Singapore brings us back to what was said about figure 1.12 as to the near-identity in the course of aging in Japan and China.

This exercise illustrates many of the issues about the Before and the After in aging, the two aging plateaus, and the secular shift in rather unexpected ways. It puts the historical aging processes that the Western developed countries have experienced into a comparative context. And it adds emphasis to the contention that the historical demography of aging has to be seen in the widest possible context over time—past, present, and future—and over all the world's political and cultural areas.

Conclusion: Processional Knowledge

A lengthy exposition has been required to demonstrate that the historical demography of aging is indeed necessary knowledge from the past. Some of this knowledge has had to be technical, even more of it numerical, not a little of it to consist in invoking visual judgment as to whether one line of a figure does or does not resemble other lines. These are not the usual qualities of historical discourse. And in the end, the knowledge and the inferences expounded turned out not to be from the past alone but from the past and the projected future, with the historically insubstantial present taking up a great deal of space.

For the title of a recent analysis of the grave issues of justice over time, the phrase "processional justice" had to be adopted.[48] It was found, and this will not surprise a demographer, that the realities to be understood escaped the analyst if they were confined to a time point. A cross-sectional view was only useful so far as it could indicate not only what is or was at a point in time but also what had been and would be. There is a similar logic in the use made of projections in the section above. We do not understand our-


62

selves in time, and this understanding is the peculiar duty of the historical sociologist, unless we see ourselves as placed in a procession. Hence the description of necessary knowledge from the past as processional.

There are two further examples from the historical demography of aging that help us to grasp what we should be after. One was touched on in the discussion of the Third Age Indicator, the 3AI. If at the beginning of your productive life the expectation of ever growing old is so low that the prospect can be sensibly disregarded, then your attitude toward your late and very late life is likely to be quite different from that of a person who can confidently expect to be old. We do not know how far this did and does mark the attitude of people placed on the lower aging plateau. There, as table 1.4 and figure 1.10 show, the value of the 3AI was often below .33 and was likely to have been lower for a fair proportion of those "decreasing the wealth of the kingdom." We do know one relevant thing about their behavior as to age and aging, however. As late as the 1930s in England, working-class people failed to save for late life, but they did save penny by penny for their funerals (Johnson 1985), which could of course come at any age for them as it could for all their predecessors.

On the lower plateau, between 15 and 25 percent of an original male cohort were alive at the age of 70, whereas 65 to 75 percent are alive at that age today. If the proletarians did write off their old age in the Before, they were acting rationally, especially if public money was there to support them in old age as it certainly was in London during the 1930s. But they were behaving as if prompted by processional knowledge, knowledge of a kind that must inform the writing of the history of aging and especially its demography.

Our further example is the same set of circumstances looked at from another point of view. In discussing transfers of resources, social as well as familial, to those in late life in the West, it is properly supposed that the recipients receive these resources as of right, as a part of the normative structure of the society they live in. These rights have been of particular importance in the West because Western rules of household formation could be said to impose hardship on older people, especially widowed and decrepit older people. It does so by removing their children from their families at the marriage of those children, "nuclear family hardship" as it has come to be called.[49] On the lower aging plateau, these rights, although they can be supposed to have attached to every individual, were only ever exercised by very few. This is a point on which the strongest stress has to be placed. The others died too soon. However important their right to support at the end of their lives may have been to them when living, seen processionally, those rights were a nullity.

One way of looking at the much discussed crisis of support for the old in Western countries today is that the secular shift in aging makes it more and more difficult, perhaps in some estimates finally impossible, to sustain the


63

inherited, perduring Western social structure in this regard. To recognize this possibility is also to practice processional thinking. Such thinking will be even more urgent in developing countries, as we have seen. It should be evident that responding to the acquisition of necessary knowledge from the past about aging, even in the preliminary fashion in which it has been done here, is to widen the intellectual horizon of historians, sociologists, and social scientists at large, as well as to explore a hitherto almost unknown area of the human reality. It does much to consolidate the concept of historical sociology and to demonstrate its overarching importance.

Appendix: Indicators for Comparison of Longevity

Measures of longevity are extensively discussed in the text above. The section, Experiential Measures in the Historical Demography of Aging, proposes that values for expectation of life at age 15 (e15 ) be substituted for expectation of life at birth (e0 ) as the most useful and revealing measure for comparing longevity between populations, particularly between those on the lower and the higher aging plateaus. It was pointed out, however, that this measure would only very seldom be available because expectation of life at birth alone is usually published in the relevant documents.

One of the objects of this appendix is to make it possible to proceed in a rough and ready fashion from e0 to e15 . In columns 1 and 2 of table 1.A1, values for twenty-five "levels" of life expectation are set out for e0 and e15 , and the trajectories of both are depicted in figure 1.A1, along with four other sets of statistics having to do with longevity. The source of all the information is the regional model life tables and stable populations published by Ansley Coale and Paul Demeny (2d ed., 1983, with the collaboration of Barbara Vaughan). All the statistics, except those for the 3AI (column 3), are for females.

The sets of tables ranged under the title "North" in that volume have been chosen for citation because the figures there seemed to correspond more closely than those in "East," "West," or "South" to the experience of England, at least up to the early nineteenth century, that is, on the lower aging plateau (see Wrigley and Schofield 1989: 110, 198, and passim). If, as is claimed in the text, English experience was in general typical of other Western countries, then the values in table 1.A1 can be regarded as indicative of Western demographic development as a whole. After the early nineteenth century, however, the English figures (and so perhaps those of contiguous European societies and their descendant societies overseas) resemble Coale and Demeny's "West" rather than "North," and this tendency must be allowed for when using the table.[50]

What follows is a discussion of the columns of the table and the corresponding lines in the figure, together with the character and values of the longevity indicators that the numbers in the columns represent.


64

TABLE 1.A1
Indicators Facilitating Comparisons in Longevity over Time and between Cultures

 

1

2

3

4

5

6

7

8

 

e0

e15

3AI (Males)

Mean Age at Death

5 Years Come

10 Years to Come

12.5 Years to Come

Half of Cohort Dead

Level 1

20.0

32.8

.111

28.9

c.73

c.58

c.52

4

Level 2

22.5

34.3

.135

28.9

c.74

c.59

c.54

5

Level 3

25.0

35.8

.160

28.9

c.74

c.59

c.53

9

Level 4

27.5

37.2

.185

29.0

c.75

c.59

c.54

15

Level 5

30.0

38.6

.211

29.1

c.76

c.58

c.57

19

Level 6

32.5

40.0

.238

29.3

c.77

c.64

c.59

27

Level 7

35.0

41.3

.265

29.6

c.78

c.65

c.60

34

Level 8

37.5

42.6

.293

30.0

c.78

c.66

c.62

40

Level 9

40.0

44.0

.321

30.5

c.78

c.63

c.57

44

Level 10

42.5

45.3

.349

31.1

c.79

c.67

c.63

50

Level 11

45.0

46.6

.377

31.7

c.79

c.69

c.63

54

Level 12

47.5

47.9

.406

32.5

c.80

c.69

c.64

57

Level 13

50.0

49.0

.431

33.4

c.81

c.69

c.65

61

Level 14

52.5

50.2

.454

34.4

c.82

c.70

c.65

63

Level 15

55.0

51.4

.479

35.7

c.83

c.70

c.64

66

Level 16

57.5

52.6

.505

37.2

c.83

c.70

c.64

68

Level 17

60.0

53.8

.531

38.9

c.83

c.71

c.64

70

Level 18

62.5

55.1

.559

40.9

c.84

c.72

c.66

73

Level 19

65.0

56.4

.588

43.3

c.84

c.73

c.67

74

Level 20

67.5

57.7

.618

46.2

c.84

c.73

c.68

76

Level 21

70.0

59.0

.648

49.8

c.84

c.74

c.67

76

Level 22

72.5

60.3

.675

54.3

c.84

c.74

c.70

77

Level 23

75.0

62.0

.713

59.0

c.85

c.75

c.71

78

Level 24

77.5

63.8

.756

64.2

c.85

c.76

c.71

79

Level 25

80.0

65.8

.803

69.7

c.86

c.77

c.73

81

NOTE : Females only, except for column 3.

The numbers in column 1 (e0 ) necessarily draw a straight line in the figure, since Coale and Demeny's tables were set up so as to make an e0 of 20 for females the basis of level 1 and an e0 of 22.5 as the basis for level 2, and so on, increasing by increments of 2.25 years up to an e0 of 80. The relationship between columns 1 and 2 (e0 and e15 ) illustrates the points made in the text. Expectation of life at age 15 is as much as 12.8 years higher than that at e0 when e0 is 20 (level 1) years (e0 /e15 = 1.64), and this superiority diminishes gradually to 8.6 years when e0 = 30 (level 5, multiplier 1.28) and to 2.8 (level 10, 1.06) when e0 = 42.5. Longevity at age 15 does not fall be-


65

low that at age 0 until a point between level 13 and level 14. Extrapolation will have to be made by users of the table: the figure for e0 in question here is presumably about 51.5 years. Thereafter the value of e15 steadily falls below that for e0 , until at the highest level the difference is 14.2 years, proportionally about as much above e0 as it is below it at the age of 20. This interesting relationship comes out clearly in figure 1.A1.

The course of the 3AI, column 3, the only set of figures for males in the table, was surveyed in the section, "The Emergence of the Third Age," above. This indicator shows a very considerable range of values and must be regarded as a highly sensitive and realistic measure of longevity experience. It can be seen that its course is parallel over most of its range to that of e0 itself. It might be permissible to infer from these circumstances that when at level 16 female expectation of life at birth reaches 57.5 (male 53.8), the 3AI (male) reaches 0.5, the qualifying level in respect to longevity, it will be remembered, for the emergence of the Third Age.[51] It is interesting that at the highest levels, 22-25, the trajectory of the male 3AI ceases to be roughly linear and curves upward.

Mean age at death, column 4 in the table,[52] is sometimes used as an approximate indicator of longevity, which is the more understandable since details of the age of death of individuals on the lower plateau is the relevant evidence most likely to be available. Inspection of the table and the figure shows, however, that such a practice is almost entirely misleading, since the course of mean age at death is curvilinear and cannot be used as a guide to e0 or e15 unless complicated calculations are made. It so happens that mean age at death is within a year or two of e0 between levels 4 and 6 (e0 27.5 to e0 32.5), but the two curves diverge rapidly both above and below that narrow band. The distance is greatest at levels 17 to 19 (e0 60 to e0 65), when mean age at death is 20 years less than e0 and 13 to 15 years less than e15 . The differences certainly narrow toward the highest levels. But it should be borne in mind that age at death is a function of population growth as well as of longevity, which makes matters even more intricate.

The full description of columns 5, 6, and 7 is mean age at which there are 5, 10, or 12.5 years still to live. Statistics of this kind have sometimes been recommended as measures of longevity and have an obvious usefulness from the point of view of individual experience.[53] It is instructive for an individual to know when she (or he) is on average within 5, 10, or 12.5 years of the end of life and particularly instructive when Third Age living has become possible from the point of view of longevity, that is (as we have seen), at levels above number 16 in our table.

What may seem astonishing is that the graphs representing values in columns 6 and 7, particularly even in their courses or consistently parallel to each other, rise so very gradually throughout the twenty-five levels of e0 . These two sets of values increase by only 35 percent and 40 percent while


66

figure

Fig. 1.A1.
Longevity indicators. Females only, except males only for 3AI. Scale for 
3AI is .01 that for age in years. Numbers in square brackets after the
 titles of the variables on the lines correspond to the columns of table 1.A1.


67

the 3AI rises by 800 percent, and in columns 4 and 8 the values are 266 percent and 2,025 percent. These circumstances make transparently clear once again a point already heavily stressed in the text: in spite of the fact that very few could expect to live to the later decades on the lower plateau, those women who did might continue for an appreciable number of additional years. If the general expectation of life was as low as the early 20s (and a dozen years higher at 15), a woman could expect to go on for another 12.5 years if she was one of those lucky enough to reach age 60. As for column 5, it makes very clear how much later in life that limit to expectation comes, even in situations where longevity is low.

The remaining column, 8, headed "half of cohort dead," should explain itself as the age by which one-half of a set of persons born in the same year have disappeared. It is even more sensitive as an indicator than the 3AI and so of considerable potential usefulness in longevity studies. This seems to be recognized, however, more by demographers studying animals than those studying people. Unfortunately, this final indicator, like those in all the columns except 5 (whose disadvantages have been set out), requires life tables if it is to be calculated. For the really detailed historical study of longevity, therefore, a series of time-separated life tables is necessary.

It can be tentatively suggested from this discussion and from an examination of the table and the figure that during the Before in the West, that is, on the lower aging plateau, for women in the West expectation of life at birth was on average at a value corresponding to a point between level 7 and level 8, that is, in the middle and late 30s, and expectation of life at age 15, in the early 40s. Something over a quarter of their male counterparts could expect to live until age 70 if they reached age 25, only about one-half of the proportion required to satisfy the longevity criterion for the Third Age. Average age at death for women was late 20s or early 30s, but women who reached the early 60s could expect to go on for 12.5 years more, for 10 years more if they attained the mid-60s, and for 5 years more in their early 80s. One-half of those born in the same year would, on average, be dead by their early 40s.

During the After, on the higher aging plateau as it is beginning to unfold, expectation of life for women at birth is likely to be in the 80s, probably inching upward: at age 15, in the mid-60s. Mean age at death is likely to be in the late 60s or early 70s, and the prospect of men surviving from 25 to 70—from the beginning of the Second Age to the middle of the Third Age—is over 80 percent. From the point of view of length of life, therefore, the Third Age will be absolutely securely established. In her early 70s on this higher plateau, a woman could probably expect to live for 12.5 years more, for 10 years more in her later 70s, and for 5 years more once over age 85. At an average age of over 80, probably slowly rising, one-half of all women born at the same date will still be alive.


68

It can be concluded with confidence that we inhabit an entirely different world with respect to length of life and experience of living it than our ancestors did, both as individuals and as a collective society. Necessary knowledge, indeed.

References

Acsádi, G., and J. Nemeskéri. 1970. History of human life span and mortality . Budapest: Akademiai Kaido.

Anderson, Michael. 1985. "The emergence of the modern life cycle in Britain." Social History 10(3): 69-87.

Applebaum, Andrew. 1978. Old age in the new land . Baltimore: Johns Hopkins University Press.

Bonfield, Lloyd. 1989. "Was there a 'third age' in the preindustrial English past?" In An aging world , ed. John M. Eekelar and D. Pearl, 37-53. Oxford: Oxford University Press.

Bongaarts, John, T. K. Burch, and Kenneth W. Wachter, eds. 1987. Family demography: Methods and their application . Oxford: Oxford University Press.

Bourbeau, R., and J. Legaré. 1982. Evolution de la mortalité au Canada et au Québec, 1831-1931 [Evolution of mortality in Canada and Québec, 1831-1931]. Montréal: Presses de l'Université de Montréal.


74

Bourdelais, Patrice. 1993. L'âge de la vieillesse [The age of aging]. Paris: Odile Jacob.

Breschi, Marco. 1990. La popolazione della Toscana dal 1640 al 1940 [The population of Tuscany from 1640 to 1940]. Florence, Italy: University of Florence, Department of Statistics.

Bulder, Elles. 1993. The social economics of old age: Strategies to maintain income in the Netherlands 1880-1940 . Tinbergen Institute Research Series, no. 50 (P.O. Box 14791, 101LG, Amsterdam).

Chesnais, J.-C. 1990. "Demographic transition patterns and their impact on the age structure." Population and Development Review 16: 327-336.

———. [1986] 1992. The demographic transition: Stages, patterns and economic implications . Oxford: Oxford University Press. (Translation by E. and D. Kreager of La transition démographique: Etapes, formes, implications économiques . Paris: INED.)

Coale, Ansley J., Paul Demeny, and Barbara Vaughan. 1983. Regional model life tables and stable populations . 2d ed. New York and London: Academic Press.

Conrad, Christoph. 1991. "The emergence of modern retirement: Germany in international comparison." Population (Paris: INED), English selection no. 3.

Conrad, Christoph, and H.-J. Kondratowicz, eds. 1993. Toward a cultural history of aging . Berlin: Deutsche Centrum für Altersfrage.

Dublin, Louis I., A. J. Lotka, and M. Spiegelman. [1936] 1949. Length of life: A study of the life table . New York and London: Academic Press.

Dyer, Colin. 1977. Population and society in twentieth-century France . Sevenoaks: Hodder and Stoughton.

Elder, Glen. 1974. Children of the Great Depression . Chicago: University of Chicago Press.

Fildes, Valerie A. 1986. Breasts, bottles and babies . Edinburgh: Edinburgh University Press.

———. 1988. Wet nursing: A history from antiquity to the present . Oxford: Blackwell.

Fischer, David H. 1978. Growing old in America . New York: Oxford University Press.

Fries, James F. 1980. "Aging, natural death and the compression of morbidity." New England Journal of Medicine 303: 130-135.

Fries, James F., and L. M. Crapo. 1981. Vitality and aging: Implications of the rectangular survival curve . San Francisco: W. H. Freeman.

Gordon, Chris. 1988. "Familial support for the elderly in the past: The case of London's working class in the early 1930s." Ageing and Society 8(3): 287-320.

Hareven, Tamara. 1978. Transitions: The family and the life course in historical perspective . New York and London: Academic Press.

Hollingsworth, Thomas H. 1964. The demography of the British peerage . Supplement to Population Studies 18(2).

———. 1977. "Mortality in British peerage families since 1600." Population (Paris: INED), numéro spéciale.

Houston, Rab. 1992. "Mortality in early modern Scotland: The life expectancy of advocates." Continuity and Change 7(1): 47-69.

Howell, Nancy. 1979. Demography of the Dobe !Kung . New York and London: Academic Press.

Janssens, Angélique. 1993. Family and social change: The household as a process in an industrializing community . Cambridge: Cambridge University Press.

Johnson, Paul. 1985. Saving and spending: The working class economy in Britain 1870-1939 . Oxford: Oxford University Press.


75

Kannisto, Väinö. 1994. Development of oldest old mortality: Evidence from twenty-eight developed countries . Monographs on Population Aging. Odense, Denmark: Odense University Press.

Kertzer, David I., and D. P. Hogan. 1989. Family, political economy, and demographic change: The transformation of life in Casalecchio, Italy, 1861-1921 . Madison: University of Wisconsin Press.

Kertzer, David I., and K. W. Schaie, eds. 1989. Age structuring in comparative perspective . Hillsdale, N.J.: Lawrence Erlbaum.

Keyfitz, Nathan, and W. Flieger. 1990. World population: An analysis of vital data . Chicago: University of Chicago Press.

Kobayashi, Kazumasa, and Zenji Nanjo. 1988. Generational life tables for Japan, based on period life tables covering the years 1891-1986 . Tokyo: Nihon University Press.

Kobrin, Frances E. 1976. "The primary individual and the family: Changes in living arrangements in the United States since 1940." Journal of Marriage and the Family 38(2).

Laslett, Peter, 1966. "The numerical study of English society." In An introduction to English historical demography , ed. E. A. Wrigley, 1-13. London: Weidenfeld.

———. 1972. "The history of the family." Introduction to Household and family in past time , ed. Peter Laslett, with Richard Wall, 1-89. Cambridge: Cambridge University Press.

———. 1976. "The wrong way through the telescope: A note on literary evidence in sociology and historical sociology." British Journal of Sociology 26: 319-342.

———. 1977. "The history of aging and the aged." In his Family life and illicit love in earlier generations , 174-213. Cambridge: Cambridge University Press.

———. [1965] 1983. The world we have lost: English society before and after the coming of industry . New York: Macmillan; London: Methuen.

———. 1984. "The significance of the past in the study of aging." Introduction to the special issue "History and Ageing." Ageing and Society 4(4): 379-389.

———. [1976] 1985. "Societal development and aging." In Handbook of aging and the social sciences , ed. R. H. Binstock and E. Shanas, 87-116. New York: Van Nostrand.

———. 1987a . "The emergence of the Third Age." Ageing and Society 7(2): 133-159.

———. 1987b . "The character of family history, its limitations and the conditions for its proper pursuit." Journal of Family History 12: 263-284.

———. 1988a . "Family, kinship and collectivity as systems of support in preindustrial Europe: A consideration of 'The nuclear hardship hypothesis.'" Continuity and Change 3(2): 153-175.

———. 1988b . "La parenté en chiffres" [Kinship in figures]. Annales ESC: Economies, Societés, Civilizations (Paris): 1-23.

———. 1989. A fresh map of life: The emergence of the Third Age . London: Weidenfeld; U.S. paperback ed., Cambridge: Harvard University Press, 1991; 2d rev. ed., London: Macmillan, 1995.

———. 1990. Review of G. Minois, History of old age. Journal of the Social History of Medicine 3(3): 461.

———. 1991. "Child, family and society." Address to the European Community, Luxembourg, May.


76

———. 1992. "The family in the industrializing East and the industrial West." Address to the International Conference on Family Formation and Dissolution, Perspectives from East and West, Taipei, Taiwan.

———. 1993. "What is old age?" Paper for the IUSSP Conference on Elderly Populations, Sendai, Japan.

———. 1994. Review article forthcoming in Ageing and Society on the final report and research papers of the Carnegie Enquiry into the Third Age.

———. In press. "The aging of European populations and its social effects." Part 1 of chap. 7 of vol. 3 of L'histoire de la population Européenne , Paris.

Laslett, Peter, and James S. Fishkin, eds. 1992. Justice between age groups and generations . New Haven: Yale University Press.

Lesthaeghe, Ronald. 1991-1992. "The second demographic transition in Western countries: An interpretation." IPD working paper, Free University, Brussels.

Minois, Georges. 1989. History of old age . Oxford: Polity Press, Blackwell.

Mitterauer, Michael. [1986] 1992. A history of youth . Trans. Graeme Dunphy. Oxford: Blackwell.

Ogawa, Naohiro. 1988. "Aging in China—Demographic alternatives." Asia-Pacific Population Journal 3(3): 21-64.

Pelling, Margaret, and Richard M. Smith, eds. 1991. Life, death and the elderly: Historical perspectives . London: Routledge.

Petersen, William. 1975. "A demographer's view of prehistoric demography." Current Anthropology 15(2): 134-156.

Plakans, Andrejs. 1989. "Stepping down in former times: A comparative assessment of 'retirement' in traditional Europe." In Age structuring in comparative perspective , ed. David I. Kertzer and D. W. Schaie, 175-195. Hillsdale, N.J.: Lawrence Erlbaum.

Riddle, S. M. 1992. Doctoral thesis on the elderly worker in England during the later 19th century and earlier 20th century, Cambridge Group.

Riley, James C. 1989. Sickness, recovery and death . London: Macmillan.

Riley, Matilda W. 1987. "On the significance of age in sociology." American Sociological Review 52: 1-14.

Ruggles, Steven. 1994. "The transformation of American family structure." American Historical Review 99(1): 103-128.

Russell, Josiah C. 1948. British medieval population . Albuquerque: University of New Mexico Press.

Schofield, Roger, David Reher, and Alain Bideau, eds. 1991. The decline of mortality in Europe . Oxford: Oxford University Press.

Smith, Daniel S. 1981. "Historical change in the household structure of the elderly in economically developed societies." In Aging, stability and change in the family , ed. R. W. Fogel, E. Hatfield, S. B. Kiesler, and E. Shanas, 91-114. New York and London: Academic Press.

Smith, James E. 1987. "The microsimulation of kin sets and kin counts." In Family demography: Methods and their applications , ed. John T. Bongaarts, T. K. Burch, and Kenneth W. Wachter, 249-260. Oxford: Oxford University Press.

———. 1991. "Aging together, aging alone." In Life span extension: Consequences and open questions , ed. Frederic C. Ludwig, 81-92. New York: Springer.

Szreter, Simon. 1993. "The idea of demographic transition and the study of fertility change." Population and Development Review 19(4): 659-701.


77

Thane, Pat. 1993. "Old age in English history." In Toward a cultural history of aging , ed. Christoph Conrad and H.-J. Kondratowicz, 17-37. Berlin: Deutsche Centrum für Altersfrage.

Thomas, Keith. 1976. "Age and authority in early modern England." Proceedings of the British Academy 62: 205-248.

Troyansky, David. 1989. Old age in the Old Régime: Image and experience in 18th-century France . Ithaca: Cornell University Press.

U.S. Bureau of the Census. 1937. Vital statistics of the United States of America, Part 1 . Washington, D.C.: Government Printing Office.

Vallin, Jacques. 1991. "Mortality in Europe from 1720 to 1914: Long-term trends and changes in pattern by sex." In The decline of mortality in Europe , ed. Roger Schofield, D. Reher, and A. Bideau, 38-67. Oxford: Oxford University Press.

van de Kaa, D. J. 1987. "Europe's second demographic transition." Population Bulletin 42(1): 3-5.

Vinovskis, Maris A. 1989. "Stepping down in former times: The view from colonial and 19th-century America." In Age structuring in comparative perspective , ed. David I. Kertzer and K. W. Schaie, 215-225. Hillsdale, N.J.: Lawrence Erlbaum.

Wall, Richard. 1989. "The living arrangements of the elderly in Europe in the 1980s." In Becoming and being old , ed. Bill Bytheway, T. Keil, et al., 121-142. London: Sage.

Wrigley, E. A., and R. S. Schofield. [1981] 1989. The population history of England 1541-1871: A reconstruction . Cambridge: Cambridge University Press.

Wu, Cangping, and Peng Du. 1992. "Rethinking population aging in China." Population Science of China 3: 1-5. (In Chinese.)

Young, Michael, and Tom Schuller. 1991. Life after work: The arrival of the ageless society . London: Harper Collins.

Zhao, Zhongwei. In press. "Rapid demographic transition and its influence on kinship networks in East and South Asia, with particular reference to China." Population Studies .


79

PART TWO
LIVING ARRANGEMENTS


81

Two
Elderly Persons and Members of Their Households in England and Wales from Preindustrial Times to the Present

Richard Wall

There is, as Peter Laslett said, a deeply entrenched belief—or rather, misbelief—that in the past all older people in England (those over the age of 65) lived in families, either their own or those of their relatives, particularly those of their married children (Laslett 1989). A considerable amount of research has now been completed which shows that this is not the case, at least in preindustrial times, even though death rates were not so high as to prevent individuals from surviving long enough to see their children marry (Laslett 1977: 184; Laslett, Wachter, and Laslett 1978; Wall 1984, 1992). Disagreement continues, however, as to whether elderly parents were disinclined to co-reside with their married children because they wanted to remain independent as long as possible or whether co-residence was ruled out by the children because of the burden that would have been imposed on their own growing families immediately or in the future. Economic hardship (potential or current) is favored by Michael Anderson to account for the reluctance of the inhabitants of rural preindustrial England as well as of some nineteenth-century towns to welcome elderly persons into their households (Anderson 1972: 229). Laslett, however, appears to believe in the persistence over centuries in England of a family system that encouraged residential independence on the part of elderly parents (Laslett 1989: 119, 121).

This chapter presents a thorough reassessment of the living arrangements of the elderly in preindustrial England. In addition, the extent to which the family and household patterns of the elderly had changed by the end of the nineteenth century also receives consideration. Anderson has argued that the urban industrial revolution of the nineteenth century was associated with a considerable increase in the frequency of co-residence between parents and married children but with evidence of variation from


82

place to place depending on the availability of employment for married women outside of the home, which encouraged co-residence as an elderly parent could provide child care (Anderson 1972: 223, 230). Also relevant, again according to Anderson, was the prosperity of the local community, with co-residence of parents and their married children only practical when the poverty was not too biting (Anderson 1972: 230). The expectation is, therefore, that the residence patterns of the elderly will have varied from one community to another. However, little hard evidence has yet been assembled on this point, either for nineteenth-century or preindustrial England. I hope to rectify this omission by noting the degree of local variation in the residence patterns of the elderly and relating this variation to particular features in the economy and social structure of the areas concerned, despite the fact that direct evidence on the living standards of specific populations in the past is particularly difficult to assemble.

Another factor that demands attention is the impact of the introduction of old age pensions in 1908 on the family and household patterns of the elderly. Prior to the Office of Population Censuses and Surveys making available to the ESRC Cambridge Group anonymized census data on thirteen communities from the period 1891-1921, it had been impossible to investigate this issue as public access to the records of the census (enumerators' books and householders' schedules) only becomes possible one hundred years after their compilation. Two competing hypotheses, however, already await examination (see Anderson 1972: 230-231). The first is that the award of a pension will have improved the ability of the elderly to maintain their residential independence; the second is that a co-resident elderly parent with a pension was a more attractive proposition to a married child as a potential co-resident as the income from the pension could help alleviate any temporary "life cycle" poverty in the child's family.

Finally, an attempt is made to chart the changes that have occurred in the residence patterns of the elderly since 1921, drawing on a number of local surveys and for recent times the Longitudinal Study of the Office of Population Censuses and Surveys. Explanations for such changes are then sought from a range of demographic, economic, and cultural factors. Demographic factors are important because of the potential impact on residence patterns of changes in the proportion of the population marrying, in fertility, and in the spacing of births, as well as an improvement in life expectancy, while economic factors may alter the ability and cultural factors the desire to maintain an independent household in old age.

The "Sample" Populations

It is somewhat ironic that households in preindustrial England have received more attention than have households in other parts of Europe, given


83

that the census material that survives for preindustrial England is so much more fragmentary and poorer in quality than that which is available for other parts of Europe. For the period before 1800 in England and Wales, there are only about five hundred censuses, each listing the inhabitants of a particular parish or township at one point in time and no more than eight recording the ages of the inhabitants while also providing adequate detail on relationships of household members to the head of the household. It is impossible, therefore, to produce a random sample of preindustrial populations or to determine how representative the communities for which information does survive might be of English society in general. All that can be claimed is that surviving censuses come from many different parts of the country and from a variety of time periods.

For the purposes of analysis, the populations have been divided into two categories: a group of smaller communities and two larger communities, Lichfield and Stoke, both located in the county of Stafford and enumerated in 1692 and 1701, respectively. Ideally, each of the smaller communities should have been analyzed separately as well, given that the family and household patterns of the elderly may well have varied across both time and space just as did other features of the social structure (see Wall 1987, and for the dates of the censuses and the number of households in each community, see the note to table 2.1, below). In practice there are so few elderly recorded in some of these censuses that any such specific influences on residence patterns would be difficult to distinguish from the effects of random variation. Consequently, the only practical option is to analyze these communities as a group. On the plus side, it is possible to examine the nature of the economy of these smaller populations in some detail as in almost all cases the occupations of the male household heads were specified. This information is set out in table 2.1 and shows that despite the small size of these communities, agriculture was by no means the dominant employment. Farmers large and small (yeoman and husbandmen) represented less than a fifth of all heads of households and even together with the laborers, many of whom would not have been exclusively or perhaps even primarily involved in farm work, constituted slightly under half of all household heads. The remainder were employed in a wide range of manufacturing and service jobs. The majority of these were intended to meet the needs of the local community (as in the case of carpenters, blacksmiths, and wheelwrights, for example), but some were producing goods for more distant markets, as in the case of the nailers of Chilvers Coton. In view of the small size of these populations, however, it is appropriate to consider them as predominantly rural in character despite the presence of some protoindustrial employment.

The two larger populations fall into a quite different category. Lichfield was a city, and the cluster of settlements that would eventually become Stoke-on-Trent already had a reputation for the production of pottery. Both


84

TABLE 2.1
Occupations of Heads of Household in Preindustrial Rural England by Employment Sector


Employment Sector


Occupation

Percentage of All
Employed Heads
a

Primary

Yeoman or farmer

11

 

Husbandman

7

 

Fisherman

2

Laborer

Laborer (unspecified)

30

 

Laborer (nonagricultural)

0

Manufacturing

Textiles

5

 

Clothing

3

 

Food

2

 

Wood

4

 

Leather

5

 

Metal

2

 

Tools and furniture

4

 

Other products

2

Service

Building

4

 

Mining and quarrying

5

 

Transport

2

 

Distribution and trade

4

 

Service

4

 

Clergy and professional

3

Other

Gentry

1

 

Military

0

 

Pauper

0

 

Other

0

 

Total

 

100

SOURCES : Listings of the population of Ealing, Middlesex, in 1599 (86 households); Chilvers Coton, Warwickshire, in 1684 (176 households); Wetherby, Yorkshire, in 1776 (214 households); Corfe Castle, Dorset, in 1790 (256 households); Ardleigh, Essex, in 1796 (201 households).

a N = 823, employed household heads; N = 110, household heads with no occupation; total N = 933.

the censuses unfortunately lack information on the occupations of the inhabitants, but the character of each community is relatively clear from the remarks of travelers and later historians, while references about the relative wealth of the two populations can be derived from an analysis of the hearth tax. Daniel Defoe and Celia Fiennes both commented on Lichfield. Defoe, whether from personal inspection or not, found it "a fine, neat, well-built, and indifferent large city . . . a place of good conversation and good company, above all the towns in this county or the next, I mean Warwickshire or


85

Derbyshire" (Defoe [1928] 1974: 80). It was the presence in the close of clergy with positions in the cathedral and other wealthy residents, not included in the census, who earned Lichfield this reputation, but there was no doubt some benefit to the rest of the city, which was by far the larger and more populous part. Fiennes was far less impressed, wondering why the bishop and other officials together with the gentry did not remove to Coventry given that Lichfield "stands so low and watrish" (Fiennes 1947: 114).

Accounts of Stoke concentrate on its role as a center of manufacture rather than on what it looked like. Nineteenth-century historians of the pottery industry tended to refer rather disparagingly to the quality and scale of seventeenth-century production. Simeon Shaw and John Ward, for example, imply that no more than between five and eight men may have been employed at a single works and that the distribution of the coarse ware that they produced may seldom have extended farther than the neighboring towns and villages (Shaw [1829] 1970: 65, 96; Ward 1843: 46). There can be no doubt, however, that there was a considerable amount of industrial activity in the area even in the late seventeenth century. Robert Plot's near-contemporary account refers to peacock coal being dug at Handley Green, which is located within the area covered by the census of Stoke, and of iron ore being worked at Longdon (also part of Stoke), while Burslem and Keele contained, respectively, the greatest pottery in the country and one of the only two centers in the country for the manufacture of frying pans (Plot 1686: 122, 126, 158, 335-336). There is ample justification, therefore, for insisting on the independent analysis of the family and household patterns of the elderly in Stoke even though some areas that were included in the census, such as Seabridge, were still predominantly agricultural even in the early nineteenth century (Shaw [1829] 1970: 60).

It is to be expected that the population of Stoke would be considerably poorer than that of Lichfield. Analysis of the hearth tax from the second half of the seventeenth century tends to confirm this. Just over a quarter of chargeable households in Lichfield were assessed on just one hearth compared to two-thirds in Stoke. Twenty-six percent of households were certified as not chargeable in Lichfield against 42 percent of the households in Stoke, although over a third of all the households in Lichfield were themselves deemed too poor to pay either the poor or church rate (William Salt Archaeological Society 1921: 153f.; 1936: 145f.). Unfortunately, how many of this last type of household there may have been in Stoke is unknown, and if they have been included in the hearth tax returns rather than simply omitted, then the differences between Stoke and Lichfield in terms of relative wealth would be considerably less than has been suggested although not totally eliminated. There is some justification, therefore, for considering Stoke the poorer of the two populations, and it will be necessary to take note of this when examining the residence patterns of its elderly inhabitants.


86

For the period 1891-1921, it has been possible to select the populations it is intended to study in detail rather than rely on the chance survival of a few local censuses, as was the case with the preindustrial period. The choice was dictated by a number of factors. Most of the populations were of interest because aspects of their social structure in other time periods had already received attention, for example, Bethnal Green in the 1950s from Michael Young and Peter Willmott (1957), Swansea in 1960 from Colin Rosser and Chris Harris (1965), Stoke in the mid-nineteenth century through the work of Marguerite Duprée (1989), and working-class York at its end from the celebrated study by B. Seebohm Rowntree (1901). At the same time, care was also taken to ensure that the thirteen populations, all that could be obtained given both the limited research budget and the time that would be required to clean and organize the data, represented a range of geographic and economic environments. This process eventually led to the selection of two "rural" populations, one in the northwest of the country (Morland) and one in the southeast (close to Saffron Walden). Another largely rural population was located in the southwest in the rural district of Axminster but including also the small seaside resort of Seaton. Two additional rural areas surrounded the market towns of Abergavenny in Monmouthshire and Banbury in northeast Oxfordshire, although Abergavenny also embraced the coal and iron working district of Blaenavon. Other areas were primarily or even exclusively industrial: Earsdon, Northumberland, had coal mining; Bolton, Lancashire, specialized in cotton; and Stoke, Staffordshire, as already mentioned, was famous for its pottery and by the late nineteenth century had also developed mining and metal interests. Swansea, too, had an important metal industry, although it was chosen principally for its role as a major port. York was selected as a representative of a large provincial town. Its adult male population was employed in a wide variety, of economic activities, with the food, transport, and construction sectors being the most important. Finally, three districts from within the London metropolitan area indicate the varied experiences of an inner-London population (Bethnal Green) and two suburban ones (Walthamstow and Pinner), the former very largely working class, the latter destined to be middle class but in 1891 only just beginning to experience the outward push of suburban London.

For the period after 1921, there are unfortunately only a few ad hoc surveys of a particular population at a particular date. These are also somewhat sparing in their details, indicating the numbers of elderly persons who lived alone, or alone but for the spouse, but rarely how many co-resided with their children (Wall 1992: 66-69). Such surveys suffice, however, to show the pace of change in both the household and the family patterns of the elderly before 1971 when the first of the random cross sections of the national population, taken every decade by the Office of Population Censuses and Surveys


87

in connection with the Longitudinal Study, can be exploited to yield a fuller breakdown of relationships within households. Even this classification, however, is less complete than those obtained from the more flexible data sets that have been produced for the historical populations. It also has to be borne in mind that the Longitudinal Study data sets are too small to permit the study of residence patterns of geographic areas smaller than that of the region, while most historical data are derived from an analysis of residence patterns in particular localities. At present there is no viable alternative strategy. No detailed data set on household patterns at the level of the individual locality is available for the present-day population, while the production of random samples of the national populations in the past are so time consuming and expensive that only one (for 1851) has so far been attempted, with that yielding little more than an overview of family and household patterns (see Anderson 1988, 1990).

Elderly Persons and Their Co-Residents in Preindustrial Times

There are any number of ways in which the household and family patterns of the elderly can be investigated. One of the more interesting approaches is to consider how many of the elderly were residentially isolated in that they lived either alone or with nonrelatives only. These two types of residence patterns are best considered in tandem as it is notoriously difficult to maintain a consistent definition of the household, and what one enumerator saw as a separate domestic group, another might subsume into a neighboring unit. The problem is serious enough in the nineteenth century when at least there was some attempt by a central authority, the Registrar General, to introduce and enforce a standard definition of the household (Wall 1982), but it is particularly acute with the earlier local censuses that were taken for a variety of purposes by different enumerators who, in compiling their lists of inhabitants, drew lines or left spaces between blocks of names but seldom recorded the procedures they had followed.

Another residence pattern that it is particularly important to distinguish involves those married couples who lived on their own, a very common pattern in contemporary Western societies and one that carries with it the threat of residential isolation should one of the partners die. Also of significance is the frequency of co-residence with a child, married or unmarried, with or without the presence of other people. There was, of course, no guarantee that child and parent would continue to co-reside until the latter's death, but at least there was a certain measure of security that care, economic support, and companionship would be provided if parents became dependent.

Data on these lines are set out in table 2.2 for preindustrial England, distinguishing the group of rural communities from the two urban areas of


88

Lichfield and Stoke. It is immediately apparent that five of the residence patterns that have been specified (living alone, with nonrelatives only, with a spouse alone, with a spouse and other persons, and with a child) capture the vast majority of the experience. For an elderly person to live with other relatives in the absence of a spouse or child was extremely uncommon in any of the communities. It is also apparent that relatively few elderly persons in preindustrial times lived totally alone: never more than 5 percent of elderly men and at most 16 percent of elderly women. As is argued above, however, those who lived only with nonrelatives should perhaps also be considered at risk of residential isolation. There were considerable numbers of such people, particularly women, in Lichfield and Stoke, reflecting perhaps the movement of women late in life into towns where there would be a broader range of cheaper accommodation. Taken together with the number of elderly who lived totally alone, it can be seen that about one in ten elderly men rising to a fifth in Stoke and around a third of elderly women rising to almost half in Lichfield were residentially isolated on the basis of the definitions set out above. These proportions are not substantially different from the proportions of persons who lived totally alone in England and Wales in 1971, as we shall see later. One important qualification has to be added. The censuses of both Lichfield and Stoke do not always specify the relationships of those persons who were not part of the nuclear family of the head of the

TABLE 2.2
Preindustrial England: Residence Patterns of Persons Aged 65+

 

Residence Patternsa

Rural English Communities (1599-1796)


Lichfield (1692)



Stoke (1701)

Males

Alone

2%

3%

5%

 

Nonrelatives only

11

8

15

 

Spouse only

19

24

10

 

Spouse and others (no child)

15

11

13

 

Child with or without other persons

49

54

54

 

Other relatives (no spouse or child)

4

0

3

   

N

104

37

39

Females

Alone

16%

15%

8%

 

Nonrelatives only

16

34

31

 

Spouse only

17

8

3

 

Spouse and others (no child)

9

7

7

 

Child with or without other persons

37

34

46

 

Other relatives (no spouse or child)

5

2

5

   

N

101

67

39

a The residents of institutions are excluded from this and all subsequent tables.


89

household, and the number of occasions on which elderly women, in particular, lived only with nonrelatives may, in consequence, be overstated. In contrast, relationships in the censuses of the rural communities are well specified, and there seems no reason to doubt that overall a third of elderly women had no relative of any sort present in their households.

However, in all the preindustrial communities, urban as well as rural, more common than either living alone or with nonrelatives was for an elderly person to co-reside with a child. Around half of elderly men shared a household with a child, married or unmarried, as did more than a third of elderly women. That fewer elderly women than elderly men lived with a child may seem surprising, as it might have been expected that when elderly women were widowed they would be taken into the household of a married child. In fact, as table 2.3 makes clear, higher proportions of elderly women than of elderly men were living with a married child, but there were many more elderly men with a co-resident never-married child. Finally, it should be noted that it was relatively rare in preindustrial times for an elderly couple to live on their own. In Stoke, in particular, there were very few such households, but even in the rural communities fewer than a fifth of elderly men and women lived alone with their spouses. This is in marked contrast to the situation of the elderly in Britain today, when more than six in every ten elderly men and a third of elderly women are living just with their spouses (Arber and Ginn 1992: 99).

A different perspective on the residence patterns of elderly persons is suggested by measuring the frequency with which they lived with a range of related and unrelated persons regardless of the number and type of other persons who might (or might not) be present. Table 2.3 contains some data

TABLE 2.3
Preindustrial England: Percentage of Persons Aged 65+ with Co-residing Spouse, Child, or Nonrelatives

 

Co-residents a

Rural English Communities (1599-1796)



Lichfield (1692)



Stoke (1701)

Males

Spouse

59

70

59

 

Never-married child

38

46

51

 

Ever-married child

12

8

3

 

Nonrelatives

44

21

28

Females

Spouse

41

21

26

 

Never-married child

21

25

31

 

Ever-married child

17

9

15

 

Nonrelatives

33

50

51

a Combinations of co-residents might be present in some cases, hence the percentages do not sum to 100.


90

along these lines for preindustrial England. For elderly men, the most likely co-resident was a spouse, followed in the case of Stoke and Lichfield by unmarried children. Co-residence with a married child was considerably rarer. Even in the rural communities, only just over 10 percent of elderly men lived with a married child. By contrast, the presence of a nonrelative, whether servant, boarder, or lodger, was somewhat more likely, and in the rural communities where 44 percent of elderly men had at least one nonrelative in their household, they were a more frequent co-resident than unmarried children.

The position of elderly women was somewhat different. In the first place, far fewer than was the case with elderly men lived with a spouse. This is only to be expected as in preindustrial times, like today, women generally outlived their spouses and were in any case on average a few years younger than their spouses on marriage or remarriage. Even so, approximately four in ten elderly women in the rural preindustrial communities were still married. That in Lichfield this proportion fell to about a fifth indicates there may well have been considerable variation from place to place in the pattern of living arrangements, probably reflecting on this occasion not only the higher mortality of urban areas but, as was suggested above, the movement of widows into localities offering a range of cheaper accommodation. Second, elderly women were less likely than were elderly men to live with an unmarried child. In this case it could be argued that the living arrangements of elderly men and women differ, because in a significant number of cases a radical reconstruction of the parental household occurred only after the death of the father. Such a reconstruction might involve both the departure and/or marriage of any children still resident and the movement of the newly widowed mother into the household of a married child. Only repetitive high-quality censuses or population registers, neither of which are available for preindustrial England, could enable the process to be traced in detail. The third point to make is that in both Lichfield and Stoke, half of all elderly women lived with a nonrelative, whereas fewer than a third of elderly men shared a household with a nonrelative. This reverses the situation in the rural communities, where elderly men were more likely to co-reside with nonrelatives than were elderly women. As noted above, the numbers of women in Lichfield and Stoke who co-resided with nonrelatives may be somewhat inflated because not all relationships between household members were identified in the censuses, but it is unlikely that so many relationships were unstated as to reduce the proportions of elderly women living with nonrelatives to the same level as those for elderly men.

Elderly Persons and Their Co-Residents Between 1891 and 1921

The above account of the residence pattern of elderly people in preindustrial England can now be directly compared with the situation in the late


91

nineteenth century and early twentieth century by drawing on the anonymized data that the Office of Population Censuses and Surveys has recently made available to the Cambridge Group. Table 2.4 parallels table 2.2 by indicating how many elderly men or women were residentially isolated, or lived with their spouses, either alone or with other people, or co-resided with a child at each of the censuses of 1891, 1901, 1911, and 1921. Few differences from preindustrial times are evident. For men the most dramatic change is the rise from a very low level in the proportions who lived with other relatives in the absence of their spouse or child. There may also have been a slight increase in the proportion of elderly men living entirely on their own. In the case of elderly women, it is clear that far fewer in the late nineteenth and early twentieth century than in preindustrial times lived with nonrelatives in the absence of relatives, while more co-resided either with a child or with other relatives in the absence of a spouse or child. Yet even after a century of great social and economic change it would be difficult to substantiate the claim that the households of either elderly men or elderly women in 1921 differed in fundamental respects from those formed by elderly people in preindustrial times. The other point worth making is how little alteration there was between 1891 and 1921 in the household and family patterns of the elderly. For elderly men there is evidence of a slight rise in the percentage living on their own, and there was also a steady increase in the proportions of both elderly men and elderly women resident with a child. These increases were balanced by a decrease in the proportions of elderly people who were still married but who did not co-reside with a

TABLE 2.4
Thirteen English and Welsh Communities, 1891-1921: Residence Patterns of Persons Aged 65+

 

Residence Patterns

1891

1901

1911

1921

Males

Alone

5%

4%

6%

6%

 

Nonrelatives only

13

11

12

11

 

Spouse only

16

18

15

18

 

Spouse and others (no child)

12

9

9

10

 

Child with or without other persons

48

51

52

52

 

Other relatives (no spouse or child)

6

6

5

6

   

N

1,696

1,764

2,174

2,597

Females

Alone

11%

11%

10%

11%

 

Nonrelatives only

13

10

12

10

 

Spouse only

10

11

9

12

 

Spouse and others (no child)

7

6

5

5

 

Child with or without other persons

47

50

52

52

 

Other relatives (no spouse or child)

12

12

12

10

   

N

2,112

2,244

2,788

3,289

SOURCE : Anonymized data from the censuses of 1891-1921.


92

child. In all other respects it is impossible to detect any consistent trends affecting the residence patterns of the elderly in this period.

An absence of change between 1891 and 1921 does not, of course, preclude significant developments between preindustrial times and the 1890s. Detailed data on family and household patterns in almost any community in the country can be abstracted from each of the decadal censuses beginning with 1851, but surprisingly few analyses have been completed given the wealth of information available, and little enough has emerged even from Anderson's 2 percent nationwide sample of enumerators' books from the censuses of 1851. Anderson, however, has now calculated the proportions of elderly men and women living alone, with nonrelatives only, with a child, or with other relatives (Anderson 1988: 436).

Of the elderly men in 1851, 5 percent lived totally alone, 16 percent with nonrelatives only, 5 percent with relatives other than a spouse and child, and 45 percent with a child. Fewer elderly men co-resided with a child in 1851 than in preindustrial times or between 1891 and 1921. However, the percentage who lived with nonrelatives in 1851 is considerably higher than it was to be in the late nineteenth or early twentieth century and higher also than what it had been in any of the preindustrial populations. The frequency with which elderly men lived with relatives other than a spouse or child is also in excess of that occurring within the preindustrial communities, although generally below the levels recorded in the later years of the nineteenth century and in the early part of the twentieth century. Finally, the proportion of elderly men living alone in 1851 appears to differ little from that between 1891 and 1921.

In a number of respects the situation with regard to the residence patterns of elderly women appears to be similar. For example, there is the same rise between 1851 and the early twentieth century in the proportions who shared a household with a child (46 percent in 1851 and between 47 percent and 52 percent during the period 1891-1921), and the proportion who lived with other relatives is, as is the case with elderly men, above that of preindustrial times but below that of the late nineteenth and early twentieth century. Elderly women, like elderly men, were also more likely in 1851 to be living only with nonrelatives than would elderly women in the late nineteenth and early twentieth century. The differences in their residence patterns as compared with those of elderly men lie in the fact that in 1851 they appear to have been less likely to have lived alone than was the case either in the late nineteenth century or in preindustrial times, while they were much less likely compared to women in the preindustrial urban populations to have lived with nonrelatives in the absence of any member of their own family.

The above account of family and household patterns of elderly people between 1891 and 1921 can also be supplemented by measuring the frequency with which the elderly co-resided with a spouse, with never-married


93

or ever-married children, or with nonrelatives whether or not other persons were present. Data on these lines are set out in table 2.5 and point to a marked fall, particularly between 1911 and 1921, in the frequency with which both elderly men and elderly women lived with nonrelatives. The same decade also witnessed a substantial rise in the proportions living with a married child, perhaps as a result of the housing crisis that followed the end of the First World War (for a further account, see Wall 1989). Between 1891 and 1911, a steadily increasing proportion of elderly persons lived with an unmarried child. No definite trend, however, is visible as regards the proportion of elderly living with a spouse.

These proportions can now be compared with those that pertained in preindustrial England (see table 2.3). Focusing first on the situation of elderly men, it is apparent that by the end of the nineteenth century, fewer elderly men were living with a spouse than had been the case in preindustrial times. In addition, between 1891 and 1921, many fewer elderly men lived with a never-married child than had done so in Lichfield and Stoke. However, between 1891 and 1921, many more elderly men lived with their married children than had been the case in any of the preindustrial populations. Trends in the frequency with which elderly men co-resided with nonrelatives are, however, more difficult to interpret, as in 1891 it was somewhat more likely that elderly men would share a household with a nonrelative than would elderly men from Lichfield and Stoke but somewhat less likely than would men from the rural preindustrial populations.

In addition, a number of the contrasts between the living arrangements of elderly men in preindustrial times and at the end of the nineteenth century do not apply in the case of elderly women. First, the proportions of el-

TABLE 2.5
Thirteen English and Welsh Communities, 1891-1921: Percentage of Elderly Persons with Co-residing Spouse, Child, or Nonrelatives

 

Co-residents a

1891

1901

1911

1921

Males

Spouse

57

56

55

57

 

Never-married child

36

38

40

38

 

Ever-married child

16

17

16

20

 

Nonrelatives

35

31

31

25

Females

Spouse

30

31

29

32

 

Never-married child

29

31

34

32

 

Ever-married child

21

22

23

25

 

Nonrelatives

34

29

29

24

SOURCE : Anonymized data from the censuses of 1891-1921.

a Combinations of co-residents might be present in some cases, hence the percentages do not sum to 100.


94

derly women still living with a spouse were some way below the proportion of elderly women in rural preindustrial England who co-resided with a spouse but were not as low as the proportions reported for the towns of Lichfield and Stoke. This contrasts with the situation of elderly men, who were less likely to be resident with a spouse than were men from any of the preindustrial populations. However, more elderly women were living with a never-married child in the nineteenth century than had lived with a never-married child in the rural communities in preindustrial times and by 1911 more even than had lived with a never-married child in Stoke in 1701. By contrast, the proportions of elderly men living with never-married children were much lower than they had been in preindustrial Stoke and Lichfield. The one major similarity in the trends over time in the residence patterns of elderly men and women was that co-residence with a married child had become much more probable by 1891.

A further perspective on the residence patterns of the elderly in the late nineteenth and early twentieth century requires that some account be taken of the degree of spatial variation. The data supplied by the Office of Population Censuses and Surveys are ideal for this purpose, because, as mentioned earlier, the thirteen populations were selected partly because of their geographic and economic diversity. To the extent, therefore, that household and family patterns are influenced by the nature of the local economy, the range of variation suggested is likely to embrace as much of the experience of the time as one could reasonably hope to cover, given that budgetary and time constraints precluded the taking of a proper national sample. Moreover, the focus on distinctive environments was enhanced due to the fact that from within each area a specific sample was drawn comprising a number of usually contiguous enumeration districts.

To illustrate the degree of variation from place to place in the residence patterns of the elderly, tables 2.6 and 2.7 set out the proportions of elderly in all thirteen communities in 1921 who were residentially isolated, or lived alone with their spouse or with their spouse and other persons but without a child, or lived with a child with or without other persons being present, or co-resided with more distant relatives. As measured by the coefficient of variation, there was least variation in the proportions of elderly men living with a child and most variation in the proportions living alone. Overall, about 6 percent of elderly men had no one else present in their household in 1921 (see table 2.4), whereas in Bethnal Green 14 percent lived alone and in Earsdon not one elderly man lived alone (table 2.6). By contrast, the percentage of elderly men co-residing with a child ranged from a high of 64 percent in Stoke to a low of 45 percent in Saffron Walden, Morland, and Bethnal Green. More variation is visible in the residence patterns of elderly women living with a child: the range extends from 63 percent in Stoke to 39 percent in Axminster.


95

TABLE 2.6
Thirteen English and Welsh Communities, 1921: Residence Patterns of Elderly Males

Residence Patterns

Abergavenny

Axminster

Banbury

Bethnal Green

Bolton

Earsdon

Alone

5%

4%

9%

14%

7%

.0%

Nonrelatives only

18

8

8

10

12

11

Spouse only

12

24

22

21

15

17

Spouse and others (no child)

6

11

9

4

7

8

Child with or without other persons

51

49

49

45

56

58

Other relatives (no spouse or child)

7

4

2

6

4

6

 

N

249

226

249

125

104

146

 

Morland

Pinner

Saffron Walden

Stoke

Swansea

Walthamstow

York

Alone

9%

3%

12%

2%

6%

6%

5%

Nonrelatives only

6

13

3

12

8

12

11

Spouse only

18

14

23

8

16

18

19

Spouse and others (no child)

9

10

10

8

9

7

10

Child with or without other persons

45

54

45

64

54

53

50

Other relatives (no spouse or child)

12

7

7

6

7

4

5

 

N

139

182

251

170

225

258

293

SOURCE : Anonymized data from the census of 1921.


96

TABLE 2.7
Thirteen English and Welsh Communities, 1921: Residence Patterns of Elderly Females

Residence Patterns

Abergavenny

Axminster

Banbury

Bethnal Green

Bolton

Earsdon

Alone

9%

9%

14%

16%

13%

5%

Nonrelatives only

9

16

11

10

12

5

Spouse only

10

16

16

12

4

12

Spouse and others (no child)

5

4

7

3

4

5

Child with or without other persons

58

39

43

49

59

62

Other relatives (no spouse or child)

9

15

9

8

8

9

 

N

223

298

328

188

143

129

 

Morland

Pinner

Saffron Walden

Stoke

Swansea

Walthamstow

York

Alone

17%

4%

16%

7%

6%

11%

13%

Nonrelatives only

6

13

5

11

8

10

12

Spouse only

12

9

20

5

10

10

10

Spouse and others (no child)

5

4

6

5

4

4

5

Child with or without other persons

48

58

43

63

62

58

48

Other relatives (no spouse or child)

13

11

9

10

9

6

12

 

N

168

270

276

245

278

353

390

SOURCE : Anonymized dam from the census of 1921.


97

It might be expected that many of these differences could be readily explained by relating them to specific features of the local economies, even if these are broadly defined as "agricultural," coal mining, and inner urban, but in practice such relationships are not easily detected. For example, although two of the industrial areas, Earsdon and Stoke, have a very low proportion of elderly men living on their own, other industrial areas such as Bolton and Swansea have only "average" proportions. Conversely, comparable (and low) proportions of elderly men living on their own occur in places that are known to be strikingly dissimilar (e.g., Stoke and Pinner). It may well be, of course, that a more detailed specification of the economic character of the areas in question, or possibly of subareas within them, would clarify the nature of the relationships between economic factors and household forms. This must await further analysis of the data. What can, however, be pointed out now is that it is in the rural communities around Saffron Walden that the family and household patterns of the elderly in 1921 took on their most "modern" look, with higher proportions of both elderly men and women living on their own, or with only their spouse, and lower proportions with nonrelatives only, or with a child, than was generally the case in 1921.

Residence Patterns of the Elderly After 1921

Keeping track of the residence patterns of the elderly after 1921 is not a particularly easy task because of the scarcity of detailed surveys and the degree of variability in the patterns recorded in those surveys that were taken. As early as 1929, the New Survey of London Life and Labour reported that 19 percent of elderly men and 37 percent of elderly women from the London working class were living alone, percentages that for women in particular were far in excess of those in Bethnal Green only eight years earlier (Gordon 1988: 26). Yet as the Second World War ended, in another working-class population, the mid-Rhondda, just 1 percent of men and 7 percent of women over the age of 65 lived alone (Nuffield Foundation 1947: 140-141). Such a low incidence of living alone is matched among the populations enumerated in 1921 only by Earsdon in the case of elderly men and by Pinner, Swansea, Stoke, and Earsdon again in the case of elderly women (see tables 2.6 and 2.7).

To measure the pace of change in the residence patterns of the elderly in the nation as a whole, it is necessary to continue to limit the comparison to the percentage of the population who lived alone. Comparisons can then be made with the results of a national survey taken in 1945 as well as with the far better-known investigation of Ethel Shanas, Peter Townsend, Dorothy Wedderburn, and others into the living arrangements of the elderly in three Western countries in the early 1960s (Shanas et al. 1968). In


98

TABLE 2.8
Britain, 1921-1981: Percentage of Persons Aged 65+ Who Lived Alone

 

Males

Females

1921

6

11

1945

6

16

1962

11

30

1971

13

36

1981

17

42

SOURCES : 1921 anonymized data on thirteen local populations from census of England and Wales 1921; 1945, calculated from national sample of the British population in Thomas 1947; 1962, national sample of the British population in Shanas et al. 1968: 186; 1971 and 1981, calculated from the national samples of the English and Welsh populations taken by the Office of Population Censuses and Surveys for the purpose of the Longitudinal Study.

table 2.8, the evidence of these two surveys as to the percentages of elderly men and women living alone in 1945 and 1962 is compared with the percentages in 1921 for the group of thirteen local populations (see table 2.4) and in 1971 and 1981 for England and Wales as a whole, derived from the samples taken for the purposes of the Longitudinal Study. Between 1921 and 1945, there was no change at all in the propensity of elderly men to live alone, while just 5 percent more women over the age of 65 lived alone in 1945 than had done so in 1921. Much more evidence of change is visible by 1962, although the degree of change in the residence patterns of both elderly men and elderly women that occurred between 1945 and 1962 was still somewhat less than that which was to take place between 1962 and 1981.

One additional perspective is possible, and that is a consideration of the residence patterns of the nonmarried elderly, as in table 2.9. The decision to study this particular group of elderly was largely dictated by the fact that it is the only group whose residence patterns can be examined through to 1981 using the population samples of the Longitudinal Study, as the way in which the data streams were defined in the study renders it impossible to determine whether married couples shared their household with other relatives or nonrelatives. Even so, to extend the perspective to 1981, a fairly basic classification of residence patterns is all that is possible, and table 2.9 is limited to showing how many elderly men and women lived with relatives, or only with nonrelatives, or alone. However, on the positive side, it should be emphasized that the nonmarried elderly are a very important group, not only because they constituted a large proportion of the total population over the age of 65 (e.g., more than four in ten of the men and two-thirds of the women aged 65+ in 1921) but because, lacking a spouse, they were particularly exposed to the risk of residential isolation.

In the late nineteenth and early twentieth century, as table 2.9 makes clear, approximately two-thirds of the elderly women and just under two-


99

TABLE 2.9
England and Wales, 1891-1981: Percentage of Nonmarried Persons Aged 65 + Who Lived Alone, with Nonrelatives Only, or with Relatives

   

1891

1901

1911

1921

1962

1971

1981

Males

Alone

12

10

13

15

37

49

63

 

Nonrelatives only

29

25

27

23

8

15

12

 

Relatives

59

65

59

62

55

36

26

   

N

721

767

970

1,115

303

6,681

8,209

Females

Alone

16

16

14

16

45

57

65

 

Nonrelatives only

19

14

16

15

5

8

5

 

Relatives

65

70

70

69

50

35

26

   

N

1,474

1,540

1,953

2,224

986

24,318

27,752

thirds of the elderly men who were not currently married nevertheless lived with at least one person to whom they were related. By 1962, this proportion had fallen, but not dramatically so. The real change had occurred with a rise in the percentage of the nonmarried elderly, both men and women, who lived alone and a marked decline in the proportion living only with non-relatives. Much the greater part of the rise between 1921 and 1962 in the proportion of elderly men living alone can therefore be explained by the decline in the percentage who were in some senses already economically independent in that they budgeted separately from the nonrelatives with whom they lived but lacked the wherewithal to establish their own households. About a third of the change in the residence patterns of elderly women can also be explained in this way. After 1962, the situation appears to change considerably with a very spectacular fall in the percentage of both elderly men and elderly women living with relatives, particularly between 1962 and 1971, matched by a continuing expansion in the proportion living on their own. The proportions living only with nonrelatives, in contrast, appear to have drifted upward between 1962 and 1971, only to fall back again by 1981.

Cultural, Economic, and Demographic Determinants of the Living Arrangements of the Elderly

At this point it is appropriate to return to the issue of whether the primary factors influencing the temporal and spatial variations in the residence patterns of the elderly were cultural, economic, or demographic. Both demographic and economic factors have clearly had a large part to play in increasing in the longer term the numbers of elderly who live on their own or


100

as couples on their own, by decreasing the frequency with which the elderly live with either nonrelatives or their children. The role of economic factors is evident in the long-term decline in the frequency of sharing a household with nonrelatives. Some of these nonrelatives encountered in the households of the elderly in preindustrial society would have been servants, but many others would have been lodgers or boarders or even unrelated people put together into one household by Poor Law authorities anxious to economize by arranging for the younger poor to care for the elderly poor (Erith 1978; Robin 1990: 208). It is often impossible to establish just how independent some of the lodgers and other unrelated people may have been of the household to which they were attached, as living spaces were so circumscribed. Over time, as the rise in living standards has allowed the standard of accommodation to be improved beyond measure, it has become much easier to see which individuals are living independently in the sense that they occupy separate accommodation. However, the lodgers and boarders of earlier times will have budgeted separately from the main household and on the definition of "independent living" ought perhaps to be judged as forming their own "households." From this perspective, therefore, some of the increase in the numbers of elderly living on their own is more apparent than real, and the role of economic forces in helping to bring about the new household forms rather less significant than might appear at first sight.

It is also interesting that the introduction of old age pensions in 1908 seemed to have little visible impact on the living arrangements of the elderly. The award of a pension, even if the initial payments were neither universal nor particularly generous (Thane 1990: 34-35), should have enabled a higher proportion of the elderly to maintain their own household. As mentioned above, this was Michael Anderson's expectation, and B. Seebohm Rowntree hinted at the existence of such a tendency in York when he compared the households of the poor in 1936 with those of the poor in 1899 (Anderson 1972: 230-231; Rowntree 1941: 114). There was no sign of this, however, when the censuses between 1891 and 1921 for the thirteen communities were analyzed (see table 2.4). A slight rise is visible from 1911 in the percentage of elderly men living alone but not in the percentage of elderly women, nor in the percentage of either elderly men or elderly women who lived "independently" in that they resided alone or only with nonrelatives. Nor is there any evidence to support Anderson's alternative suggestion that the introduction of the old age pension may have made it more feasible for sons or daughters with families of their own to offer to shelter their elderly parents in case of need (Anderson 1972: 231, quoting the statement of one pensioner circa 1912). Table 2.4 shows that although the percentage of elderly men and women living with a child did increase between 1901 and


101

1911, this merely continued a trend of the preceding decade that had therefore begun well in advance of the introduction of the old age pension.

Other changes in the living arrangements of the elderly are demographic in origin. Improvements in life expectancy over the past two centuries have obviously increased the proportion of those who survive through to old age, yet in the Britain of today the proportion of elderly women who live with a spouse is lower than the proportion who lived with a spouse in preindustrial rural England. The reason for this is that there has been a greater improvement in the life expectancy of females than of males. This has prolonged the period women spend as widows while ensuring that many more men than in the past have a spouse to provide care and companionship in old age. Another demographic change of possibly even greater significance for the living arrangements of the elderly is, paradoxically, the fall in fertility beginning in the later nineteenth century and the altered pattern of birth spacing within marriage. Prior to the demographic transition, women continued to bear children into their late 30s or early 40s. When parents reached the age of 65, it was entirely feasible, therefore, for them still to have unmarried children in their household without these children necessarily having to postpone unduly the date of their own marriage, or not marrying at all, unless of course one or both parents survived into extreme old age and still insisted on keeping the parental household intact and unchanged. In a present-day population, it is much less likely that the elderly will have a never-married child on whom they can rely, while the earlier cessation of childbearing results in children leaving the parental household much earlier in the life of the parents. The increase in the proportions who eventually marry (or cohabit with a partner) will simply intensify these trends.

It took some time, of course, before the fertility fall of the late nineteenth century and the early part of the twentieth century came to affect directly the lives of the elderly. That is why there is so little sign of change in family and household patterns of the elderly by 1921 (see table 2.5). No firm evidence can be produced to demonstrate how soon after 1921 the decline in the percentage of elderly living with an unmarried child may have established itself, but it was by no means over when Ethel Shanas and colleagues undertook their research in 1962 (Shanas et al. 1968: 186). According to their study, 22 percent of elderly men and the same percentage of elderly women in Britain in 1962 co-resided with a never-married child, percentages that were indeed lower but not substantially lower than had been the case for the elderly in 1921 when 38 percent of men and 32 percent of women over the age of 65 lived with a never-married child (see table 2.5).

Economic and demographic factors also undoubtedly help to account for much of the variation in residence patterns that emerges when the experience of one local community is compared with that of another. The pre-


102

cise impact of the factors, however, is difficult to measure. In part this is because the fertility and mortality rates to which the families of the inhabitants had been exposed are not easily calculated, as a large proportion of the adult population will not have been born in the area in which they were resident at the time of the census. In addition, however, there is the problem of identifying the likely intercorrelations between a range of economic and demographic factors, on the one hand, and the various components of the household, on the other. One example would be when a buoyant local market pushed some sections of the native population toward both earlier marriage and an earlier exit from the parental home while at the same time the improvement in the standard of living and consequent fall in mortality make it less likely that the parental home would be broken by out-migration of the children following an early parental death. In fact, what is particularly surprising is that the residence patterns of the elderly from the various populations are not more different given the differences in their economies and in the level of epidemic and endemic mortality to which they were exposed. It was established above that the population of Stoke in 1701 was in all probability considerably poorer than that of Lichfield in 1692. Their economies certainly differed as both did from those of the rural populations that were also undoubtedly healthier. E. A. Wrigley and R. S. Schofield, for example, suggest that mortality in the most sparsely inhabited rural parishes such as Hartland in Devon might be two or three times lower than that in a substantial town (Wrigley and Schofield 1983: 178-179), although the difference in the level of mortality between Lichfield and Stoke and the pooled data on the rural populations included above may be somewhat less than this as the latter did not include any community quite as remote or with such a scattered settlement pattern as Hartland. Nevertheless, in terms of the household patterns of the elderly, Lichfield, Stoke, and the rural populations appear remarkably similar. Strikingly, the 54 percent of men over the age of 65 in Lichfield who lived with a child is the same as in Stoke and just 5 percent more than did so in the rural populations. For elderly women, the range is from 37 percent in the rural populations to 46 percent in Stoke.

Much the same point can be made in connection with the thirteen populations enumerated in 1921. There is undoubtedly some variation in residence patterns between one population and another, usually somewhat greater in the case of elderly women than elderly men, but the variation is not that large given that the thirteen include some of the least healthy areas of the country, such as Stoke, as well as some low mortality populations such as Morland and Axminster (Annual reports of the Registrar General). There is no evidence either that any of the different local economies, whether mining, industrial, suburban, or agricultural, produced a unique family pattern. This might seem to suggest that there might be different "cultural" preferences in particular areas favoring the formation of one type of house-


103

hold rather than another. However, the evidence that would prove that such preferences existed and were acted upon has yet to be produced, and it seems more likely that such limited variation as there is in residence patterns does reflect the different demographic and economic circumstances of the various populations. Moreover, it is worth emphasizing that even a shared characteristic, for example, a high proportion of elderly people living alone, can occur in different demographic and economic contexts, as in the case of Saffron Walden and Bethnal Green.

Over the longer term, cultural influences may have exerted a greater impact on the sort of households that were formed. Again, however, the evidence is lacking which would establish definitely the existence of norms prescribing the residence rules for various sections of the population. By contrast with the frequent references that can be found to various forms of behavior that were deemed to be morally offensive or contrary to the natural order, such as conceiving a child out of wedlock, mésalliances, and scolding wives, little appears to have been said about when, or even whether, children should leave the parental home, or about the rights of an elderly parent to live with a child. Such norms, if they existed, therefore, must have been internalized rather than embodied in a legal code, perceived as a matter of choice by the parties immediately concerned even though their neighbors would have reacted in the same way if faced with the same situation. Laslett (1984: 364) has labeled norms of this type "noumenal normative rules," holding them applicable in particular to the process by which in England children on their marriage regularly established a household independent from that of their parents. A wider applicability, governing, for example, whether the elderly should maintain an independent household even if widowed, is also possible. Yet merely to posit the existence of such norms in one sense solves very little as Laslett himself realized. The origin of the rule system and the reason why different rules are applied in some other parts of Europe remain to be explained and could conceivably be derived from the nature of the relationship between the population and its resource base at some distant point in the past.

Also requiring explanation is why in the recent past, particularly from the 1960s, there has been the explosion in the proportion of elderly living on their own. It was suggested above that demographic change, particularly the fall in fertility earlier in the century, probably accounts for a good deal of the decline in the proportion of elderly persons residing with their children. However, it should be noted that some other investigations of the trends in family and household patterns over recent decades in a number of Western countries have concluded that not all of the trend can be explained by improvements in the standard of living and modifications to the age structure of the population (for some of the arguments, see Michael, Fuchs, and Scott 1980; Pampel 1983; Schwarz 1983). This apparent inconsistency may arise


104

from the fact that too much of the research has focused on possible period effects, whether the primary concern is with cultural, demographic, or economic determinants of residence patterns, when cohort effects may be of equal or even greater significance. The data that would allow such hypotheses to be investigated do not exist for England, but it seems plausible that each generation that reached old age after 1945 was wealthier than its predecessor, rendering the members of these generations increasingly reluctant to disband their households to move in with their children or other relatives even in extreme old age. In countries, such as the Netherlands, with population registers that can be linked to income tax and real property records, such hypotheses could be subjected to a thorough testing (see Bulder 1993 on the Netherlands before 1940).

References

Anderson, Michael. 1972. "Household structure and the industrial revolution: Mid-nineteenth-century Preston in comparative perspective." In Household and family in past time , ed. Peter Laslett and Richard Wall, 215-235. Cambridge: Cambridge University Press.

———. 1988. "Households, families and individuals: Some preliminary results from the national sample from the 1851 census of Great Britain." Continuity and Change 3(3): 421-438.

———. 1990. "The social implications of demographic change." In The Cambridge social history of Britain 1750-1950 . Vol. 2. People and their environment , ed. F. M. L. Thompson, 1-70. Cambridge: Cambridge University Press.

Arber, Sara, and Jay Ginn. 1992. "In sickness and in health: Care-giving, gender and the independence of elderly people." In Families and households: Divisions and change , ed. Catherine Marsh and Sara Arber, 86-105. London: Macmillan.

Bulder, E. A. M. 1993. "The social economics of old age: Strategies to maintain income in later life in the Netherlands 1880-1940." Ph.D. dissertation, Tinbergen Institute, Erasmus University, Rotterdam.

Defoe, Daniel [1928] 1974. A tour through the whole of the island of Great Britain , ed. G. D. F. Cole and D. C. Browning. London: J. M. Dent and Sons.

Duprée, Marguerite. 1989. "The community perspective in family history: The potteries during the nineteenth century." In The first modern society: Essays in English history in honour of Laurence Stone , ed. A. L. Beier, David Cannadine, and James M. Rosenheim, 549-573. Cambridge: Cambridge University Press.

Erith, F. H. 1978. Ardleigh in 1796: Its farms, families and local governments . East Bergholt: Hugh Tempest Radfort.

Fiennes, Celia. 1947. The journal of Celia Fiennes , ed. Christopher Morris. London: Cresset Press.

Gordon, Chris. 1988. "The myth of family care: The elderly in the early 1930s." Discussion Paper 29, Centre for Economic and Related Disciplines, London School of Economics.


105

Laslett, Peter. 1977. Family life and illicit love in earlier generations . Cambridge: Cambridge University Press.

———. 1984. "The family as a knot of individual interests." In Households: Comparative and historical studies of the domestic group , ed. Robert McC. Netting, Richard R. Wilk, and J. Eric Arnould, 353-379. Berkeley, Los Angeles, and London: University of California Press.

——— 1989. A fresh map of life: The emergence of the Third Age . London: Weidenfeld and Nicholson.

Laslett, Peter, Kenneth W. Wachter, and Robert Laslett. 1978. "The English evidence on household structure compared with the outcome of microsimulation." In Statistical studies of historical social structure , ed. Kenneth W. Wachter, Eugene A. Hammel, and Peter Laslett, 65-87. New York: Academic Press.

Michael, Robert T., Victor R. Fuchs, and Sharon R. Scott. 1980. "Changes in the propensity to live alone 1950-1976." Demography 17(1): 39-56.

Nuffield Foundation. 1947. Old people: Report of a survey committee on the problems of ageing and the case of old people . London: Nuffield Foundation.

Pampel, Fred C. 1983. "Changes in propensity to live alone: Evidence from consecutive cross-national surveys 1960-1976." Demography 20(4): 433-447.

Plot, Robert. 1686. A natural history of Staffordshire . Oxford: Timothy Halton.

Registrar General of England and Wales. 1837-. Annual reports . London: HMSO.

Robin, Jean. 1990. "The relief of poverty in mid-nineteenth-century Colyton." Rural History 1 (2): 193-218.

Rosser, Colin, and Chris Harris. 1965. The family and social change: A study of family and kinship in a South Wales town . London: Routledge and Kegan Paul.

Rowntree, B. Seebohm. 1901. Poverty: A study of town life . 2d ed. London: Thomas Nelson.

———. 1941. Poverty and progress: A second social survey of York . London: Longmans, Green.

Schwarz, Karl. 1983. "Die Alleinlebenden." Zeitschrift für Bevolkungswissenschaft 9(2): 241-257.

Shanas, Ethel, Peter Townsend, Dorothy Wedderburn, Henning Friss, Paul Mihøj, and Jan Stehouwer, eds. 1968. Old people in three industrial societies . London: Rout-ledge and Kegan Paul.

Shaw, Simeon. [1829] 1970. A history of the Staffordshire potteries . Hanley: privately printed; reprint, Newton Abbot: David and Charles.

Thane, Pat. 1990. "Government and society in England and Wales, 1750-1914." In The Cambridge social history of Britain, 1750-1950 . Vol. 3. Social agencies and institutions , ed. F. M. L. Thompson, 1-61. Cambridge: Cambridge University Press.

Thomas, Geoffrey. 1947. "The employment of older persons." An enquiry carried out in mid-1945 for the Industrial Health Research Board for the Medical Research Council. The Social Survey , new series, 60(2).

Wall, Richard. 1982. "Regional and temporal variations in the structure of the British household since 1851." In Population and society in Britain 1850-1980 , ed. T. Barker and Michael Drake, 62-99. London: Batsford.

———. 1984. "Residential isolation of the elderly: A comparison over time." Ageing and Society 4(4): 483-503.


106

——— 1987. "Leaving home and the process of household formation in preindustrial England." Continuity and Change 2(1): 77-101.

——— 1989. "English and German families and the First World War." In The upheaval of war: Family, work and welfare in Europe 1914-1918 , ed. Richard Wall and Jay Winter, 43-106. Cambridge: Cambridge University Press.

——— 1992. "Relationships between the generations in British families past and present." In Families and households: Division and Change , ed. Catherine Marsh and Sara Arber, 63-85. London: Macmillan.

Ward, John. 1843. The borough of Stoke upon Trent . London: J. W. and S. Shaw.

William Salt Archaeological Society. 1921 and 1936. Collections for a history of Staffordshire . 1921, London: Harrison and Sons; 1936, Stafford: J and C Mort.

Wrigley, E. A., and R. S. Schofield. 1983. "English population history from family reconstitution: Summary results 1600-1799." Population Studies 37(2): 157-184.

Young, Michael, and Peter Willmott. 1957. Family and kinship in East London . London: Routledge and Kegan Paul.


107

Three
The Elderly in the Bosom of the Family: La Famille Souche and Hardship Reincorporation

E. A. Hammel

Perhaps the most outstanding demographic characteristic of the human species, one that may have appeared as long ago as one hundred millennia or more, is the extraordinary prolongation of life beyond the age of reproduction. There is no immediately discernible reason why animals should continue to exist beyond their reproductive span, why the forces of natural selection would favor the emergence of such survival and thus of a species that was characterized by it. Salmon, after all, are the ultimate in age-structural efficiency; the support of the elderly is no burden to their spawn.

In the absence of immediate postreproductive death, as in the example of the salmon's nuptial couch, it is not always easy to tell whether males die soon after they cease mating. It is easier to know that for females. With rare exceptions in captivity, there are apparently few examples of living, post-menopausal female primates outside the human species. Primate females, other than humans, reproduce until they die, and these deaths are apparently "natural," not only attributable to predation. Indeed, because the chimpanzee infant is breast-fed for about three years and dependent on its mother for another two or three, the last-born child of a chimpanzee almost always dies, predeceased by its mother in the first few years of its life. Not so with human females, who may nurture not only their children but their grandchildren and even their great-grandchildren for the half or more of their adult life that now falls beyond menopause. And so also with their mates, whose longevity is only modestly less but the termination of whose reproductive activity has often been as much a matter of inquiry for their spouses as for analysts such as ourselves. The prolongation of life is not simply a function of modern health care systems; it is found as well among hunter-gatherers such as the !Kung, where the expectation of life at the age of menopause is about another quarter century.


108

To what may we attribute this extraordinary change in primate, mammalian, and animal demography? To the existence and selective importance of culture, of course. The experience and skills of a lifetime, the knowledge of heath and meadow, the wisdom that adjudicates dispute, that forges alliances with neighbors, are not to be discarded with the cessation of genetic transmission. The elderly, defined in a Darwinian sense as those over about 50, are the first transgenetic resource bank in the animal world—a place where surplus knowledge is stored.

As with any bank, the resources of this one have to be discounted. In those ancient times when the growth of new knowledge was modest, the rate of cultural inflation was low, perhaps something like the rate of population growth. [1] The stock of knowledge was much the same for cohorts of elderly separated by many years of historical time. Even if their own learning rates were less than those of the contemporary young, there was not so much new to learn, and the knowledge of the elderly was not subject to much discounting. As the rate of knowledge production grew, the value of the knowledge of the elderly, even in the presence of their continued learning, would have to have been more heavily discounted. It would have been all they could do to keep up with those whippersnappers.

Thus, if we think of this knowledge economy and its value to the species, we may conclude that in the old days no one knew very much, but the elderly knew most of that, while in modern times, many people know much more, but the elderly know much less of it, even net of the specialization of labor and expert knowledge that diminish the relative stock of any participant, regardless of age. We can only conclude from this examination that the elderly are worth less to society than they used to be.

But we should inquire into the support structure that has enabled the elderly to contribute to the development of the species. In the old days, when they knew almost everything that anyone else did, they were supported in families and households. One does not have to travel back into the Paleolithic with Dr. Wonmug and Alley Oop to find those conditions; they are encountered in much of the less-developed world today and perhaps everywhere just a few centuries ago. The knowledge that they had, again net of specialization of labor (for what farmer was a goldsmith even in his youth?), was knowledge useful to the same group that nurtured them. Thus the elderly may never have been of much use to society at large (except in some derivative Darwinian sense by the aggregation of small familial advantages) but only to the familial and household group in which they resided. Now, even today, the elderly continue to be of value in those same contexts. [2]

So, from this reexamination, we may conclude that although it might seem that the elderly were once worth more than they are now, their value in the context of their support groups may not have changed so much,


109

after all. It leads us to ask, in a more theoretically informed way, in what kinds of groups have the elderly been nurtured, and how much difference does it make?

Some Systems of Household Formation

In the history of Europe and its derivative societies, from which comes most of our information on the history of the family and household, there have been a few primary contexts. The first of these is the so-called nuclear family, or at least its residue, in which two elderly spouses care for each other. The second is the solitary household, in which one elderly person is left to care for herself usually, less often himself. Third is the complex household, sometimes called multiple lineal, in which an elderly couple resides with one or more married children. The so-called stem family household, or famille souche of LePlay's notice, is a variety of these. Fourth is the arrangement in which a surviving parent lives with a married child. This last arrangement can occur when one senior surviving spouse in a multiple lineal household dies. It can also occur if on the death of one senior spouse in a nuclear household, the survivor is reincorporated into the household of a married child. In the terminology pioneered so long ago by Peter Laslett, the contrast can be seen as one between systems of the stem family household and those of reincorporation under nuclear hardship.

In this chapter, I examine the mutual effects between two plausible systems of accommodation of the elderly and different levels of mortality between two contrasting and plausible historical demographic regimes. I also examine whether we would be able to distinguish systems of family formation or regimes of demographic rates with the sample sizes ordinarily available in historical censuses.

Microsimulation Modeling

The examination is carried out by computer microsimulation. This technique is useful in the exploration of theoretical relationships, especially when the relationships between components of a model of behavior are very complex, or when one's interest is in the variability of behavior. It is particularly useful in speculating about processes that occur in small samples, such as those typically encountered by historians, in which random error is an important source of differences. Computer microsimulation is extremely helpful in assessing the effects of random sampling error.

Now, microsimulation is a fairly complicated and technical business, the kind of endeavor that many scholars with an interest in family structure and the broad issues raised in my introductory remarks might find uncongenial. I apologize for the technicalities that follow, and I keep them to a minimum


110

in the text itself, relegating what I can to the notes and to appendixes that can be obtained on request. They are necessary to keep the game honest. It is all too easy to wave one's hand with an abracadabra and a whiff of technology to amaze the uninitiated. The technicalities are here so that the initiated can have a legitimate target for criticism.

Briefly, in this microsimulation model, a population of appropriate age and sex structure is entered into a computer and subjected to some set of demographic rates and rules of household formation. The notional individuals in this population have children, marry, divorce, die, and form and dissolve households. The demographic events occur to individuals by chance, governed by the general set of demographic rates. In the long run, populations so subjected to random occurrence of the same rates will exhibit on average the values of those rates, and differences between them will be a function of sample size and random statistical error. This is not a trivial matter, for one of the lessons most frequently learned from such exercises is that at the sample sizes typically encountered in real historical data, it is actually quite difficult to separate true differences from purely random statistical variation.

The behavior of the notional population is rather different with respect to rules of household formation. Whereas in the simulation of demographic events, occurrences are executed by chance under a set of governing rates, the simulation of household formation is here inflexible . For example, if birthrates are such that on average women between the ages of 20 and 25 could expect to have one child, the simulation does not insist that every woman have exactly one child during that age span, no more, no less. Some women have none, some one, some two, and so on, but on average they have one. Conversely, if our household formation rule is that the youngest son should remain in the parental home on marriage, every son who marries and is not the youngest moves out, and every son who is the youngest and marries stays home. Using such a fixed and rigid rule, we have a terra firma for our questions and can ask what are the effects of changes in fertility and the joint survivorship of parents and children on the attainment of stem family household organization. Of course, in reality, even in a society in which stem families were ideal, not every marriage would follow the kind of rule I have described. Some younger sons would leave to seek their fortunes, some older sons would be better farmers than their younger brothers, and so on. We might like to have a flexible set of decision rules with various contingencies pertaining to the characteristics of the household and its members. Or, we might like to have a statistical distribution of decision rules, so that different principles could compete with one another for the formation of households, in just the same way that alternative demographic events like marriage and death compete for execution in the demographic part of the simulation. [3]


111

Why do we not treat household formation in the same, sophisticated way that we treat demographic events? Let it be clear that we have nothing against it in principle. The reason is that we have no good knowledge of the statistical distribution of decisions that lead to household formation, other than the events of birth, marriage, and death that are produced by the demographic part of the experiment. We may have historical knowledge of the statistical distribution of household types in a population, but we cannot simply use those distributions to invent decision rules. It would be an empty exercise to use results as causes. Until we acquire the detailed historical or ethnographic knowledge about the actual decision processes and contingencies in specific historical populations, we must rest with this more rigid approach. At least it provides us with a firm background against which to examine the effects of demographic variation.

When the computer has done its work of simulating demographic events and household formation, censuses are taken of the population and the households at appropriate times to learn the outcomes. This kind of electronic experimentation is carried out under the contrasting conditions of interest and for each of these, a sufficient number of times to assess the importance of sampling error. [4]

It is important to realize that such experiments are not intended to recreate a specific past. Even if one used a known historical population as the one subjected to a set of demographic rates and household formation rules, and even if one used the rates and rules appropriate to that population, the very first event to occur to the notional population would with virtual certainty not be the same as that which occurred to the historical population. Any individual population is unique. The knowledge we seek is knowledge of kinds of populations. What we seek is knowledge of the expectable range of outcomes for classes of populations under classes of demographic rates and classes of household formation rules, within which individual historical populations may be deemed to fall.

In this exercise we contrast two different demographic regimes, one putatively almost modern and another putatively ancient. For each of these we contrast two systems of household formation, both of them designed in principle to offer co-residence and aid to the elderly. The first demographic regime used here is that of the United States in 1900, as a fairly typical western European system at the beginning of the industrial revolution. It is here called "Late Premodern," abbreviated LPM. [5] The second is a medieval regime gleaned from historical evidence. It is here called "Medieval," abbreviated MED. [6]

The first system of accommodation of the elderly is a form of the classical stem family. In it the youngest male child of a household to marry remains at home. It is here called "stem," abbreviated S. The second system is one in which all children leave the household on marriage but in which a


112

widowed parent, on the attainment of some specified age, rejoins the household of the youngest surviving married son. This system is here called "reincorporation," abbreviated R.

The rules of household formation require some explanation and justification. To achieve comparability between the stem and reincorporation scenarios in all but their critical differences, the kind of child whose marriage creates the stem family and the kind of child who takes the lead in reincor-porating an aged parent should be the same. It would not do to have stem families formed on the basis of a son's marriage and reincorporations take place in a daughter's household, because the mortality expectations of sons and daughters are different and could not be separated from other effects. Similarly, it would not do to have stem families formed on the basis of the marriage of an oldest child and reincorporation take place in the household of a youngest child, because mortality chances vary by age. Differences between household formation rules and demographic regimes could not then be clearly attributed; the effects of demography and of household formation rules would be confounded. In this experiment we want to keep them as separate as we can.

The decisions taken were predicated largely on the intuitive ethnographic expectation that in premodern northern European societies aged parents were more likely to be reincorporated into the household of a youngest child than that of an older child. All manner of plausible reasons for this can be imagined. Sentimental ties are usually stronger between parents and younger children. Younger children have accumulated fewer conflicting social obligations than older ones. And so on. Now this expectation is in conflict with that of primogeniture, which of course obtained in some parts of Europe and under which the eldest child (usually the son) would remain on marriage to form the next generation of the stem family. Nevertheless, the decision was to standardize on the youngest children and thus on ultimogeniture.

One must also decide whether these youngest children are sons or daughters. The choice is difficult. Because stem family formation was intimately connected with inheritance of real property, parents most frequently co-resided with sons. However, there are good reasons to anticipate that elderly parents, especially mothers, might be more likely to be reincorporated in the households of their daughters than of their sons, to avoid conflicts between mothers and daughters-in-law. The choice was to opt for sons; the youngest son, in principle, remains in the parental home, and reincorporated parents in principle join the youngest son.

The rules for stem family formation are thus as follows. Sons are preferred over daughters, but if there are no sons, the youngest daughter remains. The effect of having the youngest rather than the oldest child co-reside is to elevate the proportion nuclear, as early-marrying and on the


113

whole older children are ejected one by one, and to shorten the existence of the ultimately formed stem family households since, although younger children are less likely to predecease their parents, parents are older when their youngest child marries than when their older children do and thus more likely to die. [7]

The rules for reincorporation of widowed parents have to take into account the possibility of competition between the parents of husbands and wives and are as follows. A widowed parent seeks his or her youngest surviving ever-married child, preferring sons over daughters. If a widowed parent (A) finds an available child and has no other qualifying children but a parent (B) already resides in that household with the child, then if B has other ever-married children with whom he or she might live, B leaves to co-reside with that other child. A does not displace B if B has no other place to go. Surviving parents bereft of already co-residing children or children-in-law will seek another place to live even if they have co-residing grandchildren. If the bereft parent does not find a qualifying child but remains in the household with grandchildren, that household is then classified as a special household, namely, one without any nucleus (see below). [8]

It is also necessary to decide whether the reincorporation of widowed parents takes place immediately on widowhood or whether the parents remain independent for a time. For example, it seems strange to imagine the reincorporation of a 45-year-old widow, who might more realistically remain independent. Thus I subdivide the reincorporation scenario into two, one of early and one of late reincorporation. For simplicity, I use two critical ages at which reincorporation takes place. The first is simply the age at widowhood, whenever it occurs, which for notational convenience I here call age 0. The second is age 65, which I select somewhat arbitrarily. Any other reasonable age could have been chosen, but 65 provides a useful comparison point. The reader will later note that for purposes of symmetry, the same distinction is made in explication of stem family formation but that it has no effect under that scenario whatever.

Experimental Results—Means

The average results of the simulations are presented graphically. The graphs incorporate a great deal of information in highly condensed form, so that I must spend some time decoding them.

First I distinguish between the two ages of reincorporation (widowhood and 65, abbreviated 0 and 65, respectively). Then I distinguish within each of these the two demographic regimes, Late Premodern (LPM) and Medieval (MED). Within these I distinguish the rules of household formation, stem formation as S and reincorporation as R. There are three kinds of binary distinctions being made (age at reincorporation, demographic regime,


114

household formation), with the result that there are 23 , that is, eight different categories of information. Thus there emerge results for each rule of household formation under LPM and MED and under both critical age constraints (0 and 65). In the graphs these are labeled LPM(0), LPM(65), MED(0), and MED(65) for Late Premodern with no age constraint, Late Premodern with critical age 65, Medieval with no age constraint, and Medieval with critical age 65. Each of these four sets occurs under the reincorporation and the stem family formation regimes. Of course, as already noted, the distinction between the two critical ages of reincorporation has no bearing on the formation of stem family households. It was simply more convenient in the computing work to set up the design in a completely cross-classified way, but the design gives us the additional advantage of actually doubling the sample size for the stem family scenario and allows us to see the results of purely chance variation between the inconsequentially different subsets of stem family formation, for any demographic regime, that are normally distinguished by the critical ages of 0 and 65.

The results are first presented graphically in the form of proportional distributions of households by type, out of all households in the population (fig. 3.1). Since some complex types of household occur only rarely under these systems of household formation and are of lesser theoretical interest when the focus is on the elderly, for example, fraternal joint households, I concentrate here on only four types. These are nuclear (NUC), solitary (SOLE), multiple lineal (MLN), and extended lineal (XLN). [9] The data of figure 3.1 are found in table 3.1. [10]

Let me explicate the graph. Along the horizontal axis are first distinguished four situations in which elderly parents are absent. These are nuclear families and solitaries (NUC and SOLE), each of two varieties. Then there are distinguished four situations in which elderly parents are present. These are multiple lineal and extended lineal (MLN and XLN), each of two varieties. Within each set of four (as just given) and for each household type (NUC, SOLE, MLN, XLN), there are distinguished the two basic regimes of household formation that lead to the constellations indicated. These regimes are stem (S) and reincorporation (R). Thus we see along the horizontal axis nuclear households occurring in stem family systems [NUC(S)], nuclear households occurring in reincorporation systems [NUC(R)], and so on, to extended lineal households occurring in stem family systems [XLN(S)] and in reincorporation systems [XLN(R)].

The reader will note that each of these eight combinations of household type of interest and household formation regime of origin has space for four columns in the bar graph directly above it. Thus, for example, NUC(S) has four bars above it. Each of these bars represents a different combination of demographic regime and critical age of reincorporation. Thus, as the legend shows, we have LPM(0) for Late Premodern with critical age 0 (widow-


115

figure

Fig. 3.1.
Household proportions for all households—in the four time periods.

hood), LPM(65) for Late Premodern with critical age 65, and so on. These bars are differently patterned. LPM(0) is clear. LPM(65) is striped diagonally. MED(0) is striped vertically. MED(65) is striped horizontally. These structures and conventions enable us to compare conditions very conveniently. The reader can compare the heights of bars within one of the eight categories on the horizontal axis, for example, comparing the effects of demographic regimes within NUC(S), to see the effect of such different regimes on the formation of nuclear families in a stem family system, or across the eight categories, for example, to see whether NUC(S) or NUC(R) is more prevalent under Late Premodern conditions.

For each of the bars in the graph, the vertical axis indicates the proportion that that kind of household, under those conditions, makes up of the totality of households under those conditions. Let us first contrast the achievement of parent-child co-residence (MLN and XLN) with its absence (NUC and SOLE) across the two regimes of stem formation and reincorporation. We see that the proportion of households that are nuclear is consistently higher across all demographic regimes under a scenario of hardship reincorporation than under one of stem family formation. That is, each of the NUC(R) bars is higher than the corresponding NUC(S) bar with the same


116

TABLE 3.1
Proportions of Household Types by Demographic Regime and Household Formation Scenario

 

All Households

Households of the Elderly

Type

LPM(0)

LPM(65)

MED(0)

MED(65)

LPM(0)

LPM(65)

MED(0)

MED(65)

NUC(S)

0.482

0.503

0.490

0.477

0.205

0.230

0.118

0.150

NUC(R)

0.658

0.667

0.665

0.661

0.349

0.367

0.113

0.107

SOLE(S)

0.229

0.216

0.274

0.293

0.288

0.281

0.343

0.335

SOLE(R)

0.282

0.297

0.289

0.324

0.496

0.535

0.695

0.825

MLN(S)

0.113

0.112

0.053

0.058

0.183

0.168

0.083

0.093

MLN(R)

0.000

0.000

0.000

0.000

0.000

0.000

0.000

0.000

XLN(S)

0.121

0.116

0.153

0.141

0.277

0.272

0.402

0.356

XLN(R)

0.039

0.019

0.031

0.006

0.123

0.079

0.165

0.068


117

hatching. The same pattern holds for solitary households; each of the SOLE(R) bars is higher than its corresponding SOLE(S) bar. Conversely, the proportion of households that are multiple lineal (MLN) or extended lineal (XLN) is consistently less under reincorporation than under the stem family scenario. MLN(R) is by definition of zero occurrence, since reincorporation affects only widowed parents, and thus there can be no multiple lineal households under a reincorporation scenario. However, extended lineal households can occur under either stem or reincorporation scenarios, and all the XLN(R) bars are lower than all the XLN(S) bars.

These results show that the reincorporation scenario, which is in principle an explicit effort to include elderly parents in the households of their children, is less successful in achieving that accommodation than simple stem family formation. Why should a household formation system (R), striving to maximize the co-residence of parents and children by the explicit reincorporation of isolated parents, yield a lower proportion of such co-residence in a census? The answer is, on reflection, simple. Under a system of stem family formation, the number of person years a parent will spend in the household of a married child is greater than under a system of reincorporation , for the parent does not separate from the child and then rejoin. Thus a census is more like to capture a parent while living with a child rather than in a nuclear family after the child has left but before the parent has rejoined the child.[11] As has so often been pointed out in sensitive studies of family formation, a census is but a time slice through a process and may not reveal it.[12]

Similarly, under stem family formation rules, a widowed parent would remain with his or her child-in-law if the co-resident married child died. Under reincorporation rules, the widowed parent would rejoin a surviving married child, but not the widowed spouse of that child, if the child had been widowed before the parent.[13] Thus, under reincorporation rules, the formation of extended lineal households is depressed by those situations in which married children predecease their widowed parent.[14]

I submit that these relationships of timing and their results are perfectly obvious after they have been detected but that they were not obvious before they were exposed by these experiments.

Smaller differences are induced by changes in demographic regime (i.e., LPM vs. MED), for any system of household formation. This can be seen by comparing the heights of the bars within each block of four. Demographic regime makes almost no difference in the proportion nuclear. Under stem family rules, the Medieval demographic rates yield relatively lower proportions of stem households and higher proportions of sole and extended lineal households than the Late Premodern rates. With their higher mortality, they break up stem family households, making some people solitaries and others extensions in the households of their children. Under the reincor-


118

potation scenario, the effect of reducing the critical age of reincorporation from 65 to the actual age of widowhood gives the increase in extended lineal households that we would expect; the 0 bar is always higher than the corresponding 65 bar.

Now we shift the point of view. Up until now, the examination of the prevalence of different household types has been that of proportional representation among all households. Such data address questions like, What proportion of all the households in a census, under a particular demographic and household formation regime, are of Type X? But our focus can shift to inquire not holy all persons in the population live but rather holy the elderly live.

Figure 3.2 changes the view to that of the households of the elderly, that is, those households containing persons over age 65. Under LPM demographic conditions, the proportion of the elderly living in nuclear households is higher under the reincorporation than under the stem scenario. The result is just as it was for all households taken together (compare fig. 3.1), again because under the reincorporation scenario parents and children are separated for that period of their joint lives after the marriage of

figure

Fig 3.2.
Household proportions for households of the elderly only-in the 
four time periods.


119

the child but before the widowhood of one of the parents. However, the relationship changes under MED demographic conditions, where the proportion nuclear is noticeably lower either under the S or the R regime, and nuclear structure is somewhat more likely under stem family formation rules than under reincorporation rules. Both of these features are a reversal of the pattern for all households taken together (fig. 3.1). The reason is that the higher mortality rates under MED conditions break up elderly conjugal units, lowering their level of occurrence, both in general and also during the period of parent-child separation under the reincorporation scenario. The proportion solitary is also higher under the reincorporation scenario, for the same reasons. Some widowed persons over 65 remain solitary because they have no surviving children. This circumstance is more likely under reincorporation rules because there is a chance for married children to die during the period of separation. The effects are stronger under MED demographic conditions because of the higher mortality levels. The proportion in extended households is higher under the stem than under the reincorporation scenario, again for the same reasons; co-residence is not interrupted other than by death under the stem scenario. Comparing across demographic regimes, we see that the proportion of the elderly living in stem families is lower under Medieval conditions for the same reasons of mortality, and generally the proportion in extended lineal households is higher, because one of the spouses will be more likely to have died.

These data show us that the point of view of the classification is important for interpretation. Asking in what kinds of households people live under various combinations of demographic and household regimes is different from asking in what kinds of households the elderly live. Although some patterns persist despite a change in classification, others change markedly.[15]

Experimental Results—Variability

The estimates of household proportions achieved through these microsimulation experiments can be regarded as statistically stable because the sample sizes were so large.[16] What happens to our view of patterns and differences if sample sizes are realistically smaller? To answer this question, the large samples were used as a sampling frame, and the statistical technique of bootstrapping was employed; that is, small samples were drawn repeatedly and randomly from the large sample. In this way, without going to the trouble of running the simulations again at small sample sizes, one arrives at a good estimate of what the results would have been like if the simulations had actually been done again. Drawing samples of 100 households each 100 times with replacement produced bootstrap estimates of the means and standard errors of the means for samples of 100 households. The bootstrap means are of course almost identical to the simulation ("observed") means,


120

and the more bootstrap trials we did (say, 1,000), the closer they would be. It is the sampling errors that are of greatest interest and based on them, the confidence intervals. We should be cautious about how we interpret a confidence interval. A confidence interval of, let us say, 95 percent does not mean that we are 95 percent confident of the result. It simply specifies a range about the sample mean within which we would expect 95 percent of a large number of sample means to fall, if we sampled repeatedly and randomly from the same population. The idea is that if the mean of one sample falls within the confidence bounds of another sample, the two samples would in 95 percent of such instances actually be from the same population and are thus not truly distinguishable. This just tells us that sometimes what looks like a difference is not a difference.

Figures 3.3 and 3.4 show, for sample sizes of 100, the range from which sample proportions of household types, under the various demographic rates and formation scenarios, could be expected to come 95 percent of the time by chance alone. Figure 3.3 is for nuclear and solitary households, and figure 3.4 is for the multiple and extended households. They are divided in this way to keep the graphs from being cluttered. The categories along the horizontal axis are different from those employed in figures 3.1 and 3.2. Because figures 3.3 and 3.4 are more complex and must show ranges of data,

figure

Fig. 3.3.
Ninety-five percent confidence intervals for household 
proportions—nuclear families and solitaries.


121

figure

Fig. 3.4.
Ninety-five percent confidence intervals for household
 proportions—multiple lineal families and extended 
lineal families.

the categories on the horizontal axis are combinations of demographic regime and household formation rules. There are eight of these categories in each figure. The first four are Late Premodern (LPM), and the last four are Medieval (MED). In each set of four, the first two categories are for rein-corporation regimes, and the second two are for stem regimes. For each such pair the first member is for critical age 0, and the second is for critical age 65. Above each category are displayed the information about two different kinds of household: NUC and SOLE in figure 3.3, and MLN and XLN in figure 3.4. Each such display of information is represented by a bar, at the center of which is the mean proportional occurrence of the household type, under the stated conditions, surrounded by the 95 percent confidence interval. For example, in figure 3.3, under LPM(R)0, which means under Late Premodern demographic regimes with a system of reincorporating elderly parents as soon as they are widowed, the nuclear households constitute about 66 percent of all households, and solitary households constitute about 29 percent. We would expect the proportion of nuclear households to fall 95 percent of the time between about 57 percent and 75 percent and that of solitary households to fall 95 percent of the time between about 19 percent and 38 percent, in samples of 100. Of course, we know from statistical theory that if sample sizes are larger, the confidence intervals will be


122

narrower. Thus, if sample sizes were increased by a factor of 10 to 1,000, we would expect the confidence intervals to be reduced by a factor of the square root of 10, thus about a third of what is shown in the figure.

The question to be addressed is thus whether we could expect to be able to distinguish the demography and household formation regimes from the evidence of their results, at particular sample sizes. We would ask whether the mean for a household type under some scenario did or did not fall within the confidence interval of that same household type under some contrasting scenario. If it did not, then at these sample sizes we could use evidence of the household type proportions to argue for the presence of an underlying demographic regime, or household formation scenario, or some combination of the two. If a mean fell within the confidence interval of another regime, we could not reliably use such evidence.

In figure 3.3 we see that it is never possible, on the evidence of nuclear and solitary households, to distinguish between regimes that differ only by virtue of the critical age of reincorporation. That conclusion comes from the observation that the adjacent members of the successive pairs in the figure, such as LPM(R)0 and LPM(R)65, always show the mean of one member falling within the confidence interval of the other. Critical age of rein-corporation, although it makes a difference, does not make very much difference, and that difference would be reliably detectable only at very large sample sizes, such as the totality of the simulations presented in figures 3.1 and 3.2. Although figures 3.1 and 3.2 are a reliable guide to some large universe of results, they are not a reliable guide to historical reality as we recover it in the activity of being real historians.

We further see that it is always possible even at sample sizes of 100, on the basis of the proportion of households nuclear, to distinguish between stem and reincorporation scenarios under the same demographic conditions, for example, LPM(R)0 and LPM(S)0. However, it is never possible to use the proportion solitary in the same way, for example, MED (R)0 and MED(S)0. If comparisons are made for the proportion nuclear across regimes and scenarios that differ both in their demographic regime and in their critical age, for example, LPM(R)0 and MED(R)65, the only factor that permits clear distinction is that between household formation scenarios. The stem system and the reincorporation system always give different results. However, the proportion solitary cannot be used in this way. Thus the evidence from nuclear households is a more reliable guide to underlying conditions and process than is the evidence from solitary households. The differential quality of the evidence would not have been obvious before undertaking these experiments.

The same exercise can be conducted with figure 3.4. Of course, we must ignore the proportion MLN under nonstem scenarios, since it is by definition 0. Again, by definition, under the stem scenarios it is never possible


123

to distinguish between systems differing by age of reincorporation, for example, LPM(S)0 and LPM(S)65, because age at reincorporation is irrelevant to stem family formation.

It is usually possible from the proportion MLN to decide between the LPM and MED demographic regimes. Notice that we can only compare the stem (S) regimes and that age or reincorporation is irrelevant. Thus we should compare the error bars for the third and fourth against the seventh and eighth categories of the horizontal axis. The square plotting points that represent MLN in the seventh and eighth categories are just at the limits of the error bars for the third and fourth, while the plotting points for the latter are well outside the confidence limits for the former.

It is scarcely ever possible to distinguish between critical ages of reincorporation on the grounds of the proportion lineally extended (XLN), except under the reincorporation scenario under Medieval demographic conditions. Only in this last instance does the mean fall outside the comparable confidence interval; the open circle for XLN under MED(R)0 falls well above the upper limit of the confidence interval under MED(R)65.

It is always possible to distinguish between the reincorporation and stem scenarios on the grounds of the proportion XLN; it is always higher under stem family formation than under reincorporation rules, for the reasons already noted in the discussion of figures 3.1 and 3.2. Speaking in general, we get less reliable evidence about underlying process from observing proportions of extended and multiple households than we do from observing nuclear and solitary households; the separation of results is rather better in figure 3.3 than in figure 3.4.

Discussion

The conclusion from this exercise is that demographic differences per se are only weakly influential as against differences in household formation rule systems. This is a further confirmation of conclusions reached earlier with respect to stem family formation under plausibly different demographic regimes for historical England[17] and for joint family systems under early medieval rates.[18] The results given in this chapter are also based on ordinary and expectable historical rates. The differences between these plausible rates, however, are rather substantial, with the expectation of life at birth being only about 25 under the Medieval regime but about double that under the Late Premodern one. However, we must note that mortality differences and changes in the past have rested principally on improvements in infant and child mortality, and these have less effect on stem family and reincorporation scenarios than do changes in adult mortality. The results here presented make all the more striking the situation prevailing today in many northern European and American societies, as well as in Japan, in which the


124

mortality rates for higher ages are sufficiently different from historical experience to provoke sharp changes in household and kinship network constitution. Even for modernizing China, we have shown elsewhere that mortality changes since traditional times have been sufficient to make it more likely that traditional household structures can be reached and maintained now than earlier.[19] Whereas our customary view of demographic differences and their effects on age structure and social structure has focused on fertility and infant mortality, we see that mortality differences at higher ages can also be productive of major effects.

These explorations led to some surprises. The first was that the two household formation regimes could often be distinguished at relatively small sample sizes around 100. In earlier work, devoted to the detection of differences between stem family and nuclear scenarios without reincorporation, sample sizes of close to 200 were required. This is a welcome result for historians, who often deal with samples not much larger than about 100. But this helpful result holds only for the ability to distinguish regimes by examining some household types. For example, the required sample size necessary to distinguish each scenario from any other is only about 30 for the proportion nuclear and between 200 and 300 for the proportion solitary under most regimes but almost 4,000 under the Medieval demographic regime. The minimal size required for extended lineal proportions varies between about 20 and 60. Only trivial sample sizes are needed to distinguish rule systems on the basis of the proportion of multiple lineal households, since they should only occur under stem family formation rules, except as freak accidents of household recombination through remarriages. Working historians should be prepared to follow different strategies to make their points, depending on the household types on which they choose to focus. All households do not give answers of the same reliability.

Also unexpected was the depression of the proportion of extended lineal households, the very goal of reincorporation as an alternative to stem family formation, under an explicit system of hardship reincorporation. The residue of stem families on the death of a senior member yields a higher proportion of extended lineal households than a behavioral system that brings the elderly back from nuclearity when widowhood leaves them solitary, because of the effects of prior mortality on married children.

Conclusion

This chapter began by inquiring into the contribution of the elderly to their nurturant environment. Which social environment permits them their greatest contribution, and what social contract facilitates both it and their nurturance? Who should be responsible for this nurturance? The answer, if one wants to have old Mum at the hearth, is to start early, so that she remains


125

longer and more continuously in the bosom of the family. From a policy point of view, George Homans's villagers of the thirteenth century had it right; they kept Mum home from the start.

What of responsibility? The contribution of the elderly is surely more to their narrow than to their broad social environment, more to the family and household than to society at large. Surely those who benefit the most from their presence should take the major responsibility for their nurturance. Now, this nurturance need not involve co-residence, and the analyses presented here are only a special case, for the historian bound to four walls by the nature of the scribblings of busybodies. In the modern era we must attend to the cellular telephone, the answering machine, and perhaps even to grandmothers who send electronic mail and faxes. But the general point is the same, even if eased by technology. The care of the elderly is less likely to be a general social burden if it can be made the responsibility of their heirs—not only of their biological descendants but of their social descendants, thus not only of their children but also of their children-in-law. The filial conjugal estate should carry the social debts of both of its coparceners, the younger spouses. Might we realistically expect daughters-in-law to enter with joy into such contracts?

As levels of mortality diminish, of course, the children whom Mum might rejoin through reincorporation are less likely to be widowed, so that Mum is not robbed of a locus through the death of a child and the survival only of a child-in-law. However, as mortality declines, Mum lasts longer herself. The depression of extended lineal households under a system of reincorporation is thus a function of the ratio of the survivorship of adult children to that of their parents, or in the phrase employed in macroanalytic approaches to this same problem, the size of adjacent generations. Although the situation is not precisely analogous to the effects of shifting cohort sizes in social security systems at the macrolevel, it involves, even as there, an element of intergenerational social contract. At the microlevel the situation is conditioned by the cultural fact (at least in Euro-American societies) that blood is thicker than water. The care of the elderly, under these contrasting household formation scenarios, becomes a general social responsibility where reincorporation is the rule because some social descendants may escape an obligation that could be construed as properly theirs by virtue of their survival as coparceners of the conjugal estate that had that obligation. Under the stem family formation system, they would have continued to hold it. The outcome is obvious, but like many obvious outcomes, only after the experiments.

This exercise conducted with computers and focused on a problem from the distant past indeed has a connection to the broad problems raised at the beginning of the chapter. It is, after all, culture that differentiates us from our animal forebears, and that culture with implications for attitudes be-


126

tween the generations and for the maintenance of the cultural stock itself is closely related to mortality systems. It is these shifting mortality systems that drive the coexistence of the generations and that demand the cultural accommodation. The world now emerging is not like the old one in which parents were never free of their children but is one in which children are never free of their parents until they are themselves old. Childhood, in the psychological sense, reaches unimagined proportions. The responsibility of women as caregivers now extends across generations in two directions, and the usually earlier demise of men unites in a curious dyad the rivals who were linked to them, one as mother, the other as wife.

References

Acsádi, Gyorgy, and J. Nemeskéri. 1970. History of human life span and mortality . Budapest: Akademiai Kiado.

Coale, Ansley, and Paul Demeny, with Barbara Vaughan. 1983. Regional model life tables and stable populations . 2d ed. New York: Academic Press.

Hammel, E. A. 1972. "The zadruga as process." In Household and family in past time , ed. Peter Laslett, 335-373. Cambridge: Cambridge University Press.

——— 1980. "Household structure in 14th-century Macedonia." Journal of Family History 5: 242-273.

———. 1990. "Demographic constraints on the formation of traditional Balkan house-holds." Dumbarton Oaks [Washington, D.C.] Papers 44: 173-180.

Hammel, E. A., and Peter Laslett. 1974. "Comparing household structure over time and between cultures." Comparative Studies in Society and History 16: 73-109.

Hammel, E. A., Carl Mason, and Kenneth Wachter. 1990. SOCSIM II: A sociodemographic microsimulation program. Revision 1.0: Operating manual . Special Publication of the Graduate Group in Demography and Program in Population Research, University of California, Berkeley.


128

Hammel, E. A., Kenneth Wachter, and Chad K. McDaniel. 1981. "The kin of the aged in 2000 A.D. " In Aging . Vol. 2. Social Change , ed. James Morgan, Valerie Oppenheimer, and Sara Kiesler, 11-39. New York: Academic Press.

Hammel, E. A., Kenneth Wachter, Carl Mason, Feng Wang, and Haiou Yang. 1991. "Rapid population change and kinship: The effects of unstable demographic changes on Chinese kinship networks, 1750-2250." In Consequences of rapid population growth in developing countries , 243-271. London: Taylor and Francis.

Laiou, Angeliki. 1977. Peasant society in the late Byzantine empire: A social and demographic study . Princeton: Princeton University Press.

Reeves, Jaxk H. 1982. "A statistical analysis and projection of the effects of divorce on future U.S. kinship structure." Ph.D. dissertation, Department of Statistics, University of California, Berkeley.

———. 1987. "Projection of number of kin." In Family demography: Methods and their application , ed. John Bongaarts, Thomas K. Burch, and Kenneth W. Wachter, 228-248. Oxford: Clarendon Press.

Ruggles, Stephen. 1987. Prolonged connections: The rise of the extended family in nineteenth-century England and America . Madison: University of Wisconsin Press.

Russell, J. C. 1958. "Late ancient and mediaeval populations." Transactions of the American Philosophical Society , new series, vol. 48, pt. 3.

Wachter, Kenneth, E. A. Hammel, and Peter Laslett. 1978. Statistical studies of historical social structure . New York: Academic Press.


129

Four
Household Systems and the Lives of the Old in Eighteenth- and Nineteenth-Century Hungary

Rudolf Andorka

Where did the elderly find a place to live in past centuries in Hungary? Or, more specifically, in what type of households did they live? This question is not of historical interest only; it is relevant for present-day social policies. Like other advanced societies, Hungary is facing the problems caused by the aging of the population. But these problems are somewhat different in Hungary. As the decline of fertility below the level of simple replacement occurred more or less a decade earlier (at the end of the 1950s) in Hungary than in western European societies, these problems became acute about a decade earlier in Hungary. And, as the welfare state is much less developed in Hungary than in most western European societies, the elderly must necessarily rely much more on help from family, kin, and other personal contacts than is the case in the West. It is therefore interesting to know how the elderly were cared for in past centuries in Hungary.

The conditions of the elderly in Hungary are also important from an international comparative perspective. Earlier historical sociological research, most notably by Peter Laslett (1977, 1983), has shown that western Europe was unique in past centuries in many demographic and social indicators and institutions, including those relating to the conditions of the old population. The famous Trieste-St. Petersburg line of John Hajnal (1965), with the "late marriage, low celibacy" eastern European pattern lying to the east, passes along the western frontier of Hungary. Hungarian historians have long focused on the question of whether Hungary belongs to western or to eastern Europe, a question given new prominence by recent political events. Neglecting here the political implications of the answer to this question, I would like to point to what is probably the most serious work on this subject, by Ernö Tárkány-Szücs (1981). Tárkány-Szücs argues that after belonging to western Europe from 1000 to 1500, Hungary slid after its defeat


130

in the battle of Mohacs by the Ottoman Turkish army to a special "intermediate" eastern-central European region, together with other parts of the Habsburg empire and Poland.

It ought to be added that more recent research has shown that the demographic and social patterns in Europe were more complicated than a simple east-west division suggests. There were variations within western Europe; most notably, southern Europe seems to have been different from the West in many respects, while in the east the Balkans were different from Russia and the Ukraine, and there were "western" islands in eastern Europe and "eastern" islands in western Europe, most of all in central Europe (Laslett 1983; Viazzo 1989). There are signs that Hungary itself was rather differentiated, some regions or ethnic or denominational groups being more western European and others more eastern European.

A Short Outline of the Demographic, Economic, and Social History of Hungary in the Eighteenth and Nineteenth Centuries

The nearly two centuries of Turkish wars and occupation of the majority of the present territory of Hungary (from 1526, the date of the battle of Mohacs, to 1711, when the Habsburg army defeated the Hungarian army of Rákoczí fighting for independence) caused great population loss, most of all in the central part of Hungary, which was dominated by the Ottoman Empire. Estimations of population size based on tax lists in 1715 and 1720 vary from 2.6 to 4 million in the territory of the Hungarian Kingdom and Transylvania; less than one-third of that number lived in the present territory of Hungary. In the following decades of the eighteenth century, the relatively peaceful conditions and the gradual ending of the plague (the last great epidemic was in 1738-1741) favored rapid population growth. The first census in Hungary in 1784-1787, performed under the reign of Joseph II, found a population of 8.1 to 8.2 million in Hungary and Transylvania, of whom 2.7 million lived in the present territory of Hungary. Thus, the population more than doubled in about seventy years.

A civil servant made and published a calculation of the birth- and death rates in 1777. Although these are estimations, the very high birthrate— 55.2—and the much lower death rate—40.4—support the conclusion drawn from the census population numbers that the second half of the eighteenth century was indeed a period of very rapid population growth.

The population of the present territory of Hungary was 5 million at the time of the next reliable census, 1870. Thus in more than ninety years, the population had almost doubled again. The growth rate was clearly lower than in the eighteenth century. The birthrate was 45.4 and the death rate 40.2 in 1870-1879, when regular official vital statistics began to be kept.


131

Thus the birthrate seems already to have begun to decline in the first half of the nineteenth century, much before the onset of industrialization, but the death rate did not improve until the 1870s.

For the decades before the 1870s, only sample studies can be used to calculate vital rates. Estimations of life tables on the basis of the data from a sample of parish registers demonstrate a life expectancy at birth of about 30 years (table 4.1). Life expectancies at the age of 30 for married males and females, who figured on the completed family sheets, could be calculated from family reconstitution studies (table 4.2). Infant mortality rates were established from parish registers used for family reconstituting (table 4.3).

TABLE 4.1
Select Indicators from Mortality Tables Calculated from a Sample of Parish Registers, 1821-1830

 

Life Expectancy at Age 30 (e30 )

Number of Survivors per 1,000 Births (160 )

Category of Settlement

Male

Female

Male

Female

Villages

 

Population < 2,000

31.3

28.8

251

226

 

Population > 2,000

32.6

30.9

258

253

Agricultural towns

 

Population < 2,000

31.5

29.5

265

219

 

Population > 2,000

32.6

31.4

255

253

Towns having "royal right"

28.6

30.5

172

214

   

Total

31.6

30.8

238

242

TABLE 4.2
Life Expectancy at Age 30 of Married Males and Females in Six Hungarian Villages in the Eighteenth and Nineteenth Centuries, by Marriage Cohorts, Based on Family Reconstitutions

   

Life Expectancy at Age 30 (Years)

Village

Marriage Cohort

Male

Female

Vajszló and Besence

1791-1820

36

31

 

1821-1850

35

33

 

1851-1880

40

41

Alsonyek[*]

1760-1790

34

28

 

1791-1820

32

31

Sárpilis

1752-1790

37

28

 

1791-1820

34

29

Átány

1730-1789

33

29

 

1790-1819

29

28

Pocemegyer

1759-1790

29

29

 

1791-1820

25

28


132

TABLE4.3
Some Demographic Rates Found by Family Reconstitution in Hungarian Villages

     

Religious Denomination

Age at First Marriage of Women 1850-1895

Illegitimate
Births (%)

Infant Mortality
per 1,000 Births

Total marital Fertility, 20-49, Marriage Cohorts of

Village

Region

Ethnicity

1790-1820

1850-1895

1790-1820

1850-1895

1790-1820

1850-1895

Vajszló and Besence

Southern Transdanubia

Hungarian

Calvinist

19.9

1.0

4.4

282

130

4,760

2,760

Alsonyek[*] and Sárrpilis

Southern Transdanubia

Hungarian

Calvinist

17.5

0.4

4.5

188

215

3,725

2,725

Kerkáskápolna

Southern Transdanubia

Hungarian

Calvinist

19.1

5.9

7.8

168

182

4,415

2,500

Bakonya

Southern Transdanubia

Hungarian

Roman Catholic

20.5

       

3,595

 

Töttös

Southern Transdanubia

Hungarian

Roman Catholic

19.9

       

4,705

 

Velem

West

Hungarian

Roman Catholic

22.1

       

6,635

 

Rábakecöl

West

Hungarian

Roman Catholic

 

1.4

 

290

 

7,395

 

Bük

West

Hungarian

Roman Catholic

25.0

4.6

7.0

128

207

   

Bük

West

Hungarian

Lutheran

23.2

0.9

4.9

189

207

   

Bárna

North

Slovakian

Roman Catholic

19.5

0.2

2.3

 

225

7,795

7,375

Felsovadasz[*]

Northeast

Ruthenian

Greek Catholic

20.1

1.9

3.6

243

108

6,249

8,665

Átány

Great Plain

Hungarian

Calvinist

22.8

1.5

2.4

173

232

6,230

6,775


133

From these data the following tentative conclusions can be drawn: (1) mortality was relatively low around 1800 but did not decline in the subsequent decades; (2) regional differences seem to have been very important; and (3) female mortality was worse than male mortality, the disadvantage of females being higher in smaller settlements and in the earlier decades. Infant and child mortality of girls was also somewhat worse than that of boys, so it might be hypothesized that some form of discrimination against females prevailed.

Total marital fertility rates (i.e., the number of children ever born to 1,000 married women) calculated from a national sample of parish registers are available for two marriage cohorts, 1830-1839 and 1850-1859; the rates are 6,015 and 5,400, respectively (Dányi 1965). Results from family reconstitution studies of individual parish registers (Andorka 1978, 1988, 1990) provide data on earlier decades and permit us to evaluate the differences by regions and by ethnic and denominational groups. These data seem to suggest that while birth control was nonexistent or very rare in the marriage cohorts of the pre-1790 decades, it began to spread in some southern Transdanubian parishes in the marriage cohorts of 1790-1820, while in others marital fertility remained at an unchanged high level until the end of the nineteenth century (table 4.3).

Age at first marriage was relatively young in Hungary. The female average age at marriage was 20.3 years in the marriage cohort of 1830-1839 and 19.4 in the marriage cohort of 1850-1859 (Dányi 1965). The family reconstitution studies show considerable regional variations, the southern Transdanubian parishes having the lowest and the western Transdanubian parishes the highest age at first marriage (table 4.3). It is tempting to hypothesize that western Hungary, never occupied by the Ottoman Empire and having always had the closest contact with western Europe, was influenced by the western European marriage pattern.

From the family reconstitution studies, it is difficult to find reliable data on lifelong celibacy. If the family status of women is given in the death register, the percentage of never married above the age of 50 can be estimated. In no study was this found to be higher than 5 percent. One explanation might be the frequent remarriage of widowers. Almost all of them remarried, often much younger brides. In this way remarriage of widowers cleared the market supply of single women who had not found a single bridegroom.

Thus the elderly in Hungary in the eighteenth and nineteenth centuries faced somewhat different demographic conditions than in western Europe: marriage was earlier and almost universal; marital fertility began to decline at least in parts of Hungary earlier than in many parts of western Europe (certainly in an earlier phase of the modernization of the economy than in western Europe); mortality was not conspicuously worse at the end of the eighteenth century but did not improve until the last decades of the nineteenth century (Flinn 1981).


134

Household structure, investigated by the method and typology proposed by Laslett (1972), displays a pattern that is somewhere between the two extremes of England and Russia. The percentage of extended and multiple family households is higher than in American colonial Bristol, English Ealing, French Longueness, Serbian Belgrade (Laslett 1972), Flandrian Lampernisse (Danhieux 1983), and German Grossenmeer (Laslett 1983) but similar to Estonian Karuse (Palli 1983), Italian Alagna (Viazzo 1989), and Fagagna (Laslett 1983) and lower than in Italian Casalecchio (Kertzer 1989) and Russian Mishina (Czap 1983) (table 4.4). In terms of the characteristics of household structure, Hungary seems to be in an intermediate position between the central European, the Mediterranean, and the eastern European regions proposed by Laslett (1983).

Regional and ethnic variation were important in Hungary: German and Slovakian villages near to the center and to the largest town, Buolo, of Hungary were more similar to the central European type than the Hungarian villages of southern Transdanubia. However, household structure seems to have changed over time. In both cases, where two population listings—the German Nagykovácsi and the Hungarian Sárpilis—could be analyzed by the Laslett typology, the proportion of complex households increased (Andorka and Faragó 1983). The analysis of macrodata from the census of 1784-1787 and from the nonnoble enumerations of 1804, 1819, and 1828 demonstrates similarly that the size of households and the number of adult males per household increased (Andorka and Faragó 1983; Faragó 1977).[1]

By matching the households and the persons in the two listings of Sár-pilis in 1792 and 1804 and adding the data from the parish register, it was possible to find out what happened to the households and to the persons living in Sárpilis (Andorka and Balázs-Kovács 1986). During these twelve years the population increased from 458 to 555, but the number of households only increased from 85 to 100. In consequence, the average household size increased from 5.39 to 5.55. This tendency was more marked if we consider only the Calvinist serf peasants, that is, the autochthonous and relatively well-to-do population. Their population increased from 431 to 515, the number of their households only from 74 to 77; consequently, their average household size increased from 5.83 to 6.69. Three types of household changes can be identified: (1) in a few cases a complex household became a simple family household; (2) most often a simple family household became a multiple family household or an extended household; (3) some households fissioned into two and in one case into three parts.

In addition, a clear "household life cycle" can be demonstrated by classifying the population of Kölked by age in 1816 into the household types in which they lived (Andorka and Faragó 1983). The percentage living in extended and multiple family households was high around the age of marriage, that is, in the 20 to 29 age group, declined in the 30 to 49 age group, when some of the married couples having young children separated from


135

TABLE 4.4
Household Structure: Proportion of Households, by Type, in Some Hungarian Villages

         

Percentage of Households by Types


Mean
Size of Households


Village


Year


Region


Ethnicity

Religious Denomination


Solitary

Non-
family


Simple


Extended


Multiple

Not Classifiable

Perbál

1747

Central

German-
Hungarian Slovakian

Roman Catholic

1

1

85

6

5

2

4.67

Pilisszáintó

1747

Central

Slovakian

Roman Catholic

3

1

71

8

17

4.61

Nagykovácsi

1747

Central

German

Roman Catholic

6

79

7

7

1

4.82

 

1769

     

3

77

6

14

5.28

Fajsz

1762

Southern Great Plain

Hungarian

Roman Catholic

1

1

56

10

32

5.91

Sárpilis

1792

Southern Transdanubia

Hungarian

Calvinist

2

72

5

21

5.53

 

1804

     

1

54

9

36

5.77

Alsonyek[*]

1792

Southern Transdanubia

Hungarian

Calvinist

2

44

15

39

5.80

Kölked

1816

Southern Transdanubia

Hungarian

Calvinist

47

13

36

4

5.75

Mezocsoknya[*]

1800

Southern Transdanubia

Hungarian

Calvinist

47

15

38

6.01


136

the parental households (and thus fewer of the children lived in extended and multiple family households), and increased again after the age of 50, when the children of the old persons married and at least one of them tended to remain in the parental household.

Thus the complex household was not "perennial" in Hungary, neither on the macrolevel of historical time nor on the microlevel of individual persons and families. It seems that families tended to react to the changing demographic and economic conditions by changing their household structure.

One important economic factor, influencing household structure and probably also demographic behavior, was the growing density of the population, caused by the high growth rate of the population in the eighteenth century. This caused scarcity of land, at least in the more densely populated regions of Hungary, which was aggravated by the existence of manorial estates of landlords. Peasants were mostly serfs and were allotted a "serf parcel" of land by the landlords.[2] Families had serious difficulties increasing the land that they could cultivate, as landlords usually wished to increase or at least maintain the land cultivated by their manors. Peasant families having several heirs thus usually had to subdivide their farm. This obviously meant pauperization, as, given the lack of markets and transportation, intensification of production and sale of the agricultural products at more distant markets were almost impossible. Allowing the households to become larger and more complex was essentially a strategy, to avoid the fragmentation of the peasant farms and/or the pauperization of part of the family.

The fragmentation of serf land was favored by existing inheritance practices. The inheritance customs in the eighteenth and nineteenth centuries seem to have been rather varied in Hungary (Tárkány-Szücs 1981). For example, impartible inheritance and primogeniture predominated in the villages of German ethnicity.[3] Hungarian peasants in the backward but densely populated regions usually divided the familial property equally among the sons, and daughters received a substantial dowry at their marriage. It seems that inheritance inter vivos and retirement contracts described for northern and western-central Europe by David Gaunt (1983) and for Austria by Lutz Berkner (1972) were exceptional in Hungary. Therefore it is understandable that many married sons tended to remain—at last temporarily—in the households of their parents and farm together.

Another important institutional characteristic of eighteenth- and nineteenth-century Hungary, especially the villages, was the lack of any transfers from charitable institutions, the municipalities, or the state for poor persons not having family members or other kin to support them. Therefore, in the case of "nuclear hardship," Hungarians had to rely on families and kin groups. This fact necessarily had a strong influence on


137

household structure and kin relations and caused important differences as compared to Western societies, where support for poor individuals from the collectivity was available (Laslett 1988; McIntosh 1988; Pullan 1988).

At the same time, there was a vague but deep feeling in the Hungarian peasantry until at least the first half of the twentieth century that it was appropriate for married children to help their parents in farming and to stay at least for some years with their parents in the same household. Even when the young couple built their own house, the mutual helping relationship between parents and children and between married brothers and sisters was maintained (Fél and Hofer 1969). It might be concluded that the Hungarian peasant culture favored—but did not necessarily prescribe—mutual help and common or nearby residence of parents and adult children. Thus the elderly in Hungary in the eighteenth and nineteenth centuries could usually rely on their descendants and their families, in most cases by living together with them in the same households.

Sources of Data on the Household Conditions of the Old

To study the household contexts of older people, listings of households and populations that provide ages are needed, similar to those used by Laslett (1977) for the investigation of the history of aging and the aged. Such listings are relatively rare in Hungary in the precensus decades and centuries. Only a few listings of the first census of 1784-1787 are available in different local archives (Dányi 1965). The census forms of the regular decennial censuses beginning in 1870 were largely destroyed. Therefore, to study the household context in which old people lived, we have to search for population and household listings in the local archives and in the individual parishes. Four such listings, all found in local parish archives, are used here: (1) the status animarum of the Roman Catholic village of Fajsz in 1762, prepared for the administration of the bishopric of Kalocsa (Barth 1975); (2) the listing of the Calvinist village of Mezocsoknya[*] in 1800, prepared by the local pastor to replace the parish register, which was burned together with the local church building two years earlier, giving the year of birth and marriage of each inhabitant; (3) the listing of the Calvinist village of Sárpilis in 1804, prepared by the local pastor for the administration of the department (megye ) of Tolna in the framework of a conscription of the population of the department;[4] and (4) the listing of the Calvinist village of Kölked in 1816, prepared by the honorary principals of the local congregation, giving the age of each inhabitant (Mándoki 1971).

These listings are of varying quality. For example, Sárpilis's does not give the age of all women, so ages were determined by trying to find them in the parish register of burials. The problem of completeness, that is, of the eventual omissions, causes more serious concern. The possibility of


138

figure

Fig. 4.1.
Location of the four Hungarian villages investigated.

TABLE 4.5
Characteristics of Population Aged 60+ in Four Hungarian Villages

   

Number Aged 60+

   

Village

Population Size

Male

Female

Total

Percentage Aged 60+

Sex Ratio 60+

Fajsz

1,022

43

18

61

6.0

239

Mezocsoknya[*]

547

9

11

20

3.7

82

Sárpilis

525

19

10

29

5.5

190

Kölked

643

17

15

32

5.0

113

 

Total

2,737

88

52

142

5.2

163

omissions is suggested by the very high sex ratio of the populations over age 60 investigated here (table 4.5). One reason for the higher number of males over age 60 might be the shorter life expectancy of females (tables 4.1 and 4.2). The very high sex ratio in Fajsz and Sárpilis, however, cannot be explained simply by mortality differences of men and women. In the case of Sárpilis, it was possible to compare the parish registers and the listings and to verify in this way that no old women are missing from the listing of 1804. The surplus of men might be caused partly by the migration balance by sexes of the previous years. From 1792 to 1804, 45 men out-migrated and 33 immigrated, while 58 women out-migrated and 40 immigrated. Thus the loss of migration was 12 men and 18 women (not detailed


139

by age). If the migration balance was similar in previous decades, then in addition to the higher mortality of women, the greater loss of women by out-migration might explain the very high sex ratio. No similar data are available to explain the even higher sex ratio found in Fajsz. It might be assumed that—as the listing dates from an earlier period than that of Sár-pilis—the surplus mortality of women was higher.

All four listings present the names of the inhabitants and their ages ordered by households. Households are divided by serial numbers. It is, however, not absolutely clear what the serial numbers mean, namely, whether one household or one house is given under one serial number. It might be assumed, however, that the enumerators used the definition of the census of 1784-1787, according to which all those belong to one "familia" that cooks and eats together.

Usually the oldest male family head is given as the first person in the household and might be defined in consequence as the household head. In some exceptional cases, the male head of the younger family is first mentioned. Widowers seem to be treated similarly, that is, they are usually listed as household heads if they are the oldest in the household. Widows are sometimes mentioned first and are therefore defined as household heads; in other cases, they are mentioned as a family member of the household head, who is usually their son. The household position of the other members is defined in relation to the first-mentioned person of the household (e.g., spouse, child, mother of the household head).

Cotters (landless peasants) and servants are mentioned explicitly in the listings of Fajsz and Sárpilis. Two types of them were distinguished in the analysis of the data and were treated differently: (1) solitary servants and cotters, who are always assigned in the listings to households, are treated as members of these households; (2) cotters and servants having a family were treated here as separate households, even if they were assigned in the listing to another household. Ethnographic evidence suggests that these cotter and servant families cooked and ate separately. As, however, the number of cotters and servants assigned to other households was relatively small (much lower than in contemporary England), their treatment in the analysis does not influence the results very much.

As is visible from the data presented earlier in this chapter, in the case of Sárpilis, additional sources were used. A listing of the population in 1792, not containing complete age data, was used to study household structure. The comparison of the listings of 1792 and 1804 and of the parish register could be used to investigate migration and the changes of household structure. A listing of the economic resources of the households from 1793 made it possible to investigate the characteristics of households and families by social strata (Andorka and Balázs-Kovács 1986). The family reconstitution


140

based on the Calvinist parish register made it possible to investigate the demographic characteristics and changes from 1752 to the end of the nineteenth century (Andorka 1978).

Some Data on the Economic, Social, and Demographic Characteristics of the Four Villages

The four villages are situated near one another in a well-defined geographic and ethnographic region of Hungary, namely, southeastern Transdanubia and the opposite bank of the Danube (Fajsz). This region was not poor, the soil being fertile, but land became scarce in the second half of the eighteenth century, and the area was far from any markets and from centers of protoindustrialization. The geographic conditions—marshy areas around the Danube, lack of roads—also contributed to the isolation of the region from European and even national influences. It might be concluded that these populations lived at that time in archaic conditions.

Relatively little is known about the family economy of these four villages. A monograph on the villages of the western bank of the Danube gives more information on Sárpilis and Kölked (Andrásfalvy 1975). The economic conditions were probably similar in Fajsz on the eastern bank. The economy of these villages was dominated by the fact that, being on the floodplain of the Danube, they had relatively little arable land. In consequence they had to combine cultivation of arable lands with animal husbandry (for which purpose the meadows that were periodically inundated were very suitable), fishing, hunting in the forests and at the riverside, and long-distance transportation with horse-drawn carts. The economic register of Sárpilis in 1793, which was matched to the population listings of 1792 and 1804, as well as two similar economic registers demonstrate that these peasants had little arable land. Most households had less than one unit of land held in villenage but a relative abundance of animals (horses, oxen, cows). These fragmentary data suggest that the peasants of these villages had to employ a very flexible family economy, continuously adapting their labor supply to the sources of income open to them and expanding and reducing their productive activities according to the number of adults in the household. For example, plowing of arable land could be done either by oxen or by horses. Oxen were stronger and therefore could plow the soil deeper, but plowing with one or two pair of oxen required two adult men, because in addition to the man directing the plow, one man was needed to guide the oxen. Therefore either the number of adult males had to be adjusted to the oxen or, if only one adult male lived in the household, horses had to be used instead of oxen for plowing. It might be hypothesized that peasants sought to employ a certain household structure to utilize efficiently the possibilities of production.

The peasants of these villages were socially rather differentiated. The data from Sárpilis in 1793 illustrate the social differences: of the 85 house-


141

holds, 32 comprised rich peasants, having 2 or 4 oxen or 2 to 6 horses; 31 comprised middle-level peasants, having 0 to 2 horses and some cows; 11 comprised poor peasants, having no large animals; and another 11 comprised very poor peasants, not included in the listing of resources. Richer peasants tended to have larger and more complicated households.

As compared to other villages in Hungary, all four villages investigated here had larger and more complicated households (table 4.4). It is doubtful that the demographic indices of Sárpilis, calculated on the basis of family reconstitution, can be assumed to represent the historical demography of the other three villages as well.

Mortality conditions in Sárpilis can be characterized by the indicators of infant and child mortality.[5] In the period from 1792 to 1820, 382 deaths of children age 0 to 9 per thousand births occurred. The rate was somewhat higher for girls (410) than for boys (350), the opposite of the usual pattern. Similarly, the mortality of adult males was more favorable than that of adult females (table 4.6).

In the period 1752-1790, the average age at first marriage for women was 19 years, and in the period 1791-1820, it was 18.4 years. The average age of men at marriage was 24.4 and 19.9, respectively, in these periods. It is difficult to estimate the frequency of remarriage in Sárpilis, as until 1805 the marital status of bridegrooms is not given. Of the 100 marriages celebrated between 1792 and 1804, 13 involved brides who were widowed and 2 involved brides who were divorced. In the case of the 66 marriages over the next ten years, between 1805 and 1814, 18 bridegrooms were widowers, 9 brides were widows, and 1 bride was divorced. From these data, it might be guessed that remarriage of widowers was more frequent than remarriage of widows.

Sárpilis is one of the villages where signs of early birth control were observed in the marriage cohort of 1752-1790. However, marital fertility was still very high, amounting to a total fertility rate of 7,180 from the ages of 20 to 49.[6] Illegitimacy rates were low (table 4.3).

The high level of fertility and the moderate rate of mortality resulted in a high rate of natural increase. In the period from 1779 to 1803, 630 bap-

TABLE 4.6
Survival of Couples Married in Sárpilis before 1791

 

Percentage Attaining the Given Age

Age

Male

Female

30

100

100

40

100

89

50

91

71

60

74

50

70

46

34


142

tisms and 414 burials are mentioned in the parish register of Sárpilis, and the population (458 in 1792 and 555 in 1804) increased by 216 as a result of the surplus of births over deaths.

Data on migration were obtained by matching the listings of 1792 and 1804 and the parish register of Sárpilis. From a population of 458 in 1792, 103 persons out-migrated in the following twelve years. From a population of 555 in 1804, 73 persons had in-migrated during the previous twelve years. Thus part of the natural increase of 216 over those years was "discharged" from Sárpilis through a negative migration balance. This fact itself points to the difficulties caused by the high rate of natural increase for the society of Sárpilis.

The out-migration of 22 percent from Sárpilis in 1792-1804 is lower than the migration intensity from Clayworth and Cohenhoe in seventeenth-century England and Hallines in eighteenth-century France and more or less similar to the population turnover in Longuenesse in eighteenth-century France (Laslett 1977). It is, however, higher than out-migration in Pinkenhof, Latvia (Plakans and Wetherell, this volume), and Krasnoe Sobakino (Czap 1983) in nineteenth-century Russia.[7]

An important factor in both out-migration and in-migration was marriage with residents of neighboring villages. The residence of the new couple was mostly determined by the principle of living in the household or at least the village of the father of the bridegroom. A few cases were found in which the new couple resided with the father of the bride.

Another important factor in the turnover of population was the out-migration and in-migration of complete households. These households belonged to three types: (1) most were landless cotter families who frequently migrated from one place to another; (2) the households of the local Calvinist pastor, notary, and schoolteacher from time to time changed their place of work from one Calvinist village to another; and (3) two economically strong farmer households and a nuclear family component of an economically strong multiple family household out-migrated from 1792 to 1804, demonstrating that there was pressure, probably caused by land scarcity, to find new "farm niches" in less densely populated parts of Hungary.

In addition, there were some "individual" migrants. Five were found in the listing of 1804. Their demographic and social characteristics illustrate the causes of these movements. These in-migrants were the following: a 35-year-old discharged soldier who returned to the household of his father and brother; a widowed father who came to live with his married son; a widowed woman who came to live with her married sister; a 77-year-old widower (the only one who was above 60 among these individual migrants) who came with his son, who had married the daughter of a local peasant; and a widowed mother who joined the household of her married son. Thus it seems that in


143

exceptional cases of hardship, the "reincorporation strategy" (Hammel, this volume) was utilized. Nevertheless, Plakans's hypothesis that old age or "post-labor force" migration might have been rare in preindustrial rural places in Europe seems to be confirmed by the data from Sárpilis.

All these data point to the fact that these were most probably not average Hungarian villages but villages living in archaic conditions. The findings on the conditions of the elderly should not be generalized, therefore, to the whole of Hungary, where demographic and social conditions, including the conditions of the elderly, might have been rather varied. These data nevertheless give some insight into the question of the type of household conditions in which the elderly lived in archaic Hungarian villages.

Household Composition and Household Position of the Population Over Age 60 in the Four Villages

On the basis of the four listings containing age data, the indicators and tables used by Laslett (1977) were calculated and compiled. The data are compared to Laslett's.

The percentage of the population over age 60 is lower than in the English, French, German, Italian, Icelandic, and Japanese localities given by Laslett but is near those found in Austria, Estonia, and Serbia. The sex ratio is among the highest in all these places (table 4.5). The problems of the elderly, which in present-day Western societies are predominantly the problems of old women, were in these four villages about two hundred years ago predominantly the problems of old men.

The most conspicuous difference in terms of marital status between the Hungarian villages and the places presented by Laslett is the absence of single persons in the former.[8] The percentage of married men is somewhat higher in the Hungarian villages than in most places given by Laslett, while the percentage of married women is similar or somewhat lower in Hungary than in England and France but much higher than in Serbia and Japan (table 4.7). In consequence, the percentage of widows among the women is higher in Hungary than in all the English and French locals but much lower than in Serbia and in Japan.

The high sex ratio and the much higher percentage of married persons among the men than among the women over age 60 suggest that the age difference between spouses might have been important. The comparison of the ages of married men with the ages of their wives demonstrates that, indeed, more than half of the wives were ten or more years younger than their husbands (table 4.8).[9] Husbands' second marriages were distinguished only in the listing of Mezocsoknya[*] . Here both husbands over age 60 who had wives who were younger by more than ten years were in a second marriage.[10] It might be assumed that in the other villages, marriages involving husbands


144

TABLE 4.7
Proportion Married among Those Aged 60+

Village

Men (%)

Women (%)

Fajsz

81

61

Mezocsoknya[*]

89

9

Sárpilis

79

60

Kölked

53

27

 

Total

76

41

TABLE 4.8
Number of Married Men Aged 60+ by Age Groups and the Age Groups of Their Wives in Four Hungarian Villages


Age of Husband

Age of Wife

 

< 39

40-49

50-54

55-59

60-64

65-69

70-74

75-79

80+

Total

60-64

1

12

9

3

4

29

65-69

1

1

1

1

1

5

70-74

2

1

1

2

3

2

11

75-79

2

1

1

1

4

80+

1

1

1

1

1

2

1

2

11

 

Total

5

16

13

5

7

6

5

1

2

58

who were much older than their wives were second marriages as well. Therefore, it might be hypothesized that the relatively low proportion of widowers among the old men was due to a large extent to the fact that widowed men usually remarried, sometimes marrying much younger women.

As in the English communities presented by Laslett, most married men and women over age 60 were household heads and spouses of heads (table 4.9). The majority of widowed men were also household heads, but only about half of the widows were heads of households. Altogether, 93 percent of the men and 70 percent of the women over age 60 were household heads (table 4.10).

A closer look at the persons who were not household heads helps explain the high headship rates. Only two married men were found who were not heads of households. One of them was 93 years old and lived with his 88-year-old wife in the household of a married son. The other was an 85-year-old servant-lodger who lived with his 38-year-old wife in the household of an unrelated farmer. The two widowed men who were not heads and lived with married children were 77 and 76 years old. The two who lived in a household of nonkin were a 76-year-old lodger and a 60-year-old servant. One of the widows who was not a head lived as a servant in a nonkin-related household.


145

TABLE 4.9
Proportion of Persons over Age 60 Who Are Household Head or Spouse of Head

 

Male

Female

Village

Married (%)

Unmarried (%)

Married (%)

Unmarried (%)

Fajsz

94

75

91

100

Mezocsoknya[*]

100

100

100

30

Sárpilis

100

100

100

0

Kölked

100

100

100

60

 

Total

97

81

96

52

No systematic difference could be found between the widows who were heads of the households in which their married and/or unmarried children lived and the widows who lived in households headed by their married children. Men over age 60 were not heads only in exceptional cases of old age or poverty. No clear rule is evident for women remaining heads after widowhood or giving over the headship to one of their children.

The most conspicuous difference as compared to the English communities analyzed by Laslett is that in these Hungarian villages, no person over age 60 lived in an institution, only two unmarried men and one unmarried woman lived as a servant and/or lodger in a nonkin household, and no person lived alone. It is clear that no institutions for the care of old and poor persons existed in these villages, and apparently no children were willing to let their widowed parents live by themselves. Old widowed parents either continued to live in their households together with their unmarried and married children or entered the households of married children after widowhood. Not only the parents of the household head but in some cases also the widowed mother of the head's spouse lived in the household. Comparison of the listing of 1804 and Sárpilis with the listing of 1792 indicates that widowed parents in some cases were accepted into the households of their married children, that is, the hardship reincorporation household scenario (Hammel, this volume).

Both married and unmarried old people lived much more frequently with their married children than was the case in England (table 4.10). Seventy-six percent of the men and 82 percent of the women over age 60 lived together with married children and in most cases with grandchildren. In consequence, the generational depth of the households in which old people lived was great: in 73 percent of these households, members of three generations lived together (table 4.11). Most old persons thus had everyday contact with their grandchildren, and many children had everyday contact with their grandparents.[11]


146

TABLE 4.10
Number of Persons Aged 60+ by Sex, Marital Status, and Household Position in Four Hungarian Villages

 

Fajsz 1762

Mezocsoknya[*] 1800

Sárpilis 1804

Kölked 1816

Total

Marital Status and Household Position

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Married, household head,
having in their household

 

Unmarried children

7

3

3

10

3

 

Married children

6

3

1

9

3

6

1

22

7

 

Unmarried and married children

17

6

2

6

3

3

28

9

 

Others

3

1

3

1

 

Only spouse

2

1

2

1

Married, not household head,
living in the household of

 

Unmarried children

 

Married children

1

1

1

1

 

Unmarried and married children

 

Others

1

1

Not married, household head,
having in their household

 

Unmarried children

2

1

1

3

1

 

Married children

2

5

1

2

1

7

5

11

12

 

Unmarried and married children

2

1

1

1

1

3

3

 

Others

1

1

Not married, not household head,
living in the household of

 

Unmarried children

 

Married children

1

7

1

2

3

2

12

 

Unmarried and married children

-

1

1

2

 

Others

l

1

1

2

1

   

Total

43

18

9

11

19

10

17

15

88

54


147

TABLE 4.11
Distribution of Households by Generational Depth in Four Hungarian Villages

Number of Generations in Household

Fajsz

Mezosoknya[*]
1800

Sárpilis

Kölhed

Total

1

1

1

2

2

16

4

3

4

27

3

31

14

18

24

87

Solitary servants

2

2

4

 

Total

50

19

23

28

120

NOTE: In Fajsz, 2 persons aged 60+ lived in two-generation households and 9 persons in three-generation households; in Mezocsoknya[*] , 2 persons lived in 1 one-generation household; in Sárpilis, 2 persons lived in 6 three-generation households; in Kölked, 2 persons lived in 4 three-generation households.

The other aspect of the same phenomenon is that most of the households in which old people lived were of complex structure: following the typology proposed by Laslett (1972), only 13 percent belonged to the simple family household type, 26 percent to the extended family household type, and 57 percent to the multiple family household type (4 percent were not classifiable). Of the 57 percent who lived in multiple family households, 26 percent belonged to the "stem" type, 18 percent to the "joint" type, and 13 percent to the "frérèche" type.[12] Thus the joint-and frérèche-type households, which were infrequent in Hungary and very rare in western Europe, were relatively frequent among the households of the elderly in these villages (table 4.12).

It seems that older people lived largely in extended and multiple family households. It would be premature, however, to draw the conclusion that a stem and/or joint family system was the general rule in these four villages. To understand the formation of these complicated households, the changes of household contexts of the elderly have to be analyzed.

The comparison of the listings of 1792 and 1804 gives some insight into the changes of the household context of the persons over age 60 in 1804 in Sárpilis. Five of these older persons were not found in the listing of 1792; they probably immigrated in the meantime. Two of them were widowed mothers of household heads. It might be assumed that they lived in a neighboring village before the death of their husbands and came to live after widowhood with their married sons in Sárpilis. The three others were old men who immigrated with their families.

For those found in both listings, we can distinguish the following transitions: (1) becoming a cotter or a servant; (2) living in the same simple household; (3) the household changing from a simple to a more complex pattern, in consequence of the marriage of an unmarried child; and (4) splitting of the household, the elderly parent living in 1804 in a "successor"


148

TABLE 4.12
Household Structure of Persons Aged 60+: Number of Households by Type in Four Hungarian Villages


Household Type


Fajsz 1762

Mezocsoknya[*] 1800

Sárpilis
1804

Kölked 1816


Total

3. Simple family household

 

3.a. Married couple alone

1

1

2

 

3.b. Married couple with child(ren)

8

3

1

12

 

3.c. Widower with child(ren)

1

1

 

3.d. Widow with child(ren)

1

1

4. Extended family household

 

4.a. Extended upward

8

6

3

11

28

 

4.b. Extended downward

2

2

 

4.c. Extended laterally

1

1

5. Multiple family household

 

5.a. Secondary unit(s) up

 

5.b.1. Secondary unit(s) down "stem"

16

4

6

5

31

 

5.b.2. Secondary, unit(s) down "joint"

6

10

6

22

 

5.c.d. Units on one level

4

5

2

4

15

 

5.e. Other multiplea

1

1

6. Not classifiable (solitary servants in household of nonkin)a

2

2

4

 

Total

50

19

23

28

120

a The conjugal families of two married brothers and of one married son of one of the brothers living in one household.

household. There is no case of an elderly person who lived in 1792 in a complex household living in 1804 in a simple household.

One person, who was head of a simple family household in 1792, was widowed; his son married into another household, and he himself lived as a cotter in a nonkin household. A widower who lived in the stem household of his son in 1792 lived after the death of this son as a servant in a nonkin household in 1804.

One person lived as the head of the same simple family household in both years with his wife and his unmarried young children. One person lived as the head of the same stem household with his wife and the family of his married son. Two persons, an old couple, lived as heads of the same stem household with the family of their eldest son, the two younger daughters having departed to a neighboring village when they married.

Seven people—three couples and a male household head—lived in simple family households in 1792 which became stem households in 1804 as a result of the marriage of one of their sons. One male household head


149

lived in a simple family household in 1792 which became a joint household in 1804 due to the marriage of two sons. One male household head lived in a simple family household in 1792 which became a multiple household (type 5.c.d.) after the death of his spouse and the marriage of two sons.

The remaining nine people over age 60 in 1804 lived in households in 1792 which had split by 1804 into several households. One woman had lived with her husband in a simple family household. Her husband died, one of her sons married, and she then lived with her younger son in the stem household of this married son, who was the head of the household. Meanwhile, a daughter married and entered another simple family household.

One old man lived with his wife, his younger unmarried son, and his older married son in a stem household in 1792. The married son separated and lived in his own simple family household. The younger son also married and lived with his parents, who remained household heads, in a stem household.

Two people lived in a stem household with one married son and three unmarried children. Two of the unmarried sons married between 1792 and 1804; one of them established a separate simple family household, the other remained in the household of the parents together with the oldest married son, so that the household was of a joint type (5.b) in 1804.

One male household head lived in a joint household in 1792 with two married sons. One of the married sons established a separate 3.b household, so that the original household of the head became a stem household.

One male household head lived in 1792 in a frérèche-type (5.c.d) household with the family of his younger married brother. The two families split, and the older brother lived in a simple family household, while one of the sons of his younger brother married, so that they ended up living in a stem household.

One male household head lived with his family and two younger married brothers in a frérèche household. The three families split and established separate households, the younger brothers forming simple family households, the oldest brother a stem household in which a married son lived with his parents.

Finally, a couple lived in 1792 with two married sons and the widowed sister of one of their daughters-in-law and her child in a joint household. The widowed sister died, and her son married into another household, where his father-in-law was the head. The older married son established his own separate simple family household, while the parents lived with their younger married son in a stem household.

Thus the household context of the old people changed in various ways between 1792 and 1804 under the influence of demographic events. One type of change, however, never happened: no elderly person who lived in a complex household with his or her children later lived in a simple family household, if married children were present in the villages. No clear corre-


150

TABLE 4.13
Co-Residence by Age and Sex in Kölked, 1816

 

Percentage Living in Extended and
Multiple Family Households

Age

Male

Female

0-9

55

62

10-19

51

63

20-29

85

88

30-39

71

57

40-49

49

59

50-59

68

83

60+

89

100

 

All

64

67

lation could be found between the changes of these households and their economic resources (measured by the number of oxen, horses, and cows in their possession in 1793). The two households that disappeared were, however, poor.

It seems that the life of the older population was organized in a rather different way in these four Hungarian villages from that in western Europe and even in western-central Europe (e.g., Austria). Most of the old people lived in households with their married children or at least with one married child (table 4.13). They shared with them not only housing but also the costs of living. In case of need, adult children were able to care for their old parents. This resulted from a household system characterized by a large proportion of extended and multiple family households. This household system was well adapted to the economic circumstances of these villages and was supported by cultural values and norms. Nevertheless, this did not re-suit in "perennial" multiple family households. The household structure adapted continuously—both on the micro- and macrolevel—to the changing demographic and economic conditions of the individual families and of the village community.

Implications for the Handling of Current Social Problems in Hungary

The conditions of the elderly population (18.7 percent of the total population was over age 60 in 1989) are problematic in several respects. As pensions are not indexed, only the lowest pensions are increased so that they do not fall below the official minimum. The older pensioners, more or less those above the age of 70, often face serious financial problems. The health conditions of the elderly are very bad; many suffer from


151

chronic illnesses. Many old persons, especially widows, are isolated. As long as the old people continue to work after retirement in a part-time job or in the second economy, their daily life is dominated by these income-supplementing activities. When they cannot find such work, they, especially the old men, do not find rewarding and meaningful leisure activities for themselves (Andorka 1990).

The Hungarian state is not able to solve the problems of the elderly. It does not seem to be possible to assure the indexing of pensions in the framework of the present social security system. The health care system, particularly the care provided for the elderly, is considered to be of rather low quality, so that those in the medical service and the old patients themselves are very dissatisfied. The institutional homes for the elderly provide housing and care for only a small segment of the elderly population and are considered to give very poor services, so that older people and their younger family members tend to take advantage of them only in case of extreme need.

In these circumstances, families, especially children, aid their older relatives as much as possible in Hungary today. The simplest way to provide financial support, daily health care, and human contacts to the elderly is for elderly parents and their married or unmarried adult children to live together in the same household. Although the share of complex households has declined, it has remained relatively high in Hungary: in 1980, 13.5 percent of the population lived in households that might be defined as "extended," and 8.3 percent lived in "multiple family" households, following Laslett's typology (table 4.14). A survey in 1969 showed that 40 percent of the men of pensionable age (60+) and 49 percent of the women of pensionable age (55+) lived together with at least one child (Andorka et al. 1972). A comparison of the households of the elderly in the United States and in Budapest in the 1980s demonstrates that a much higher percentage of Hungarian elderly lived with adult children in the same households (Farber et al. 1990).

TABLE 4.14
Household Structure: Distributions of Persons and of Households by Type in Hungary, 1980

Household Type

Persons (%)

Households (%)

Solitary

7.1

19.6

Nonfamily household

2.6

3.4

Simple family household

68.5

63.5

Extended family household

13.5

9.2

Two-family, in vertical relation

7.4

3.8

Two-family, in lateral relation

0.1

0.1

Two-family, nonkin

0.4

0.2

Three-family

0.4

0.2

 

Total

100

100


152

All available survey data suggest that the old people living with or at least in the proximity of younger family members, usually children, have fewer income problems, are better cared for in case of illness, and are more satisfied with their human relationships and with life in general than those who do not have frequent contact with their children. The decline of the average number of children per family, the growth of the divorce rate, the decline of the rate of remarriage of divorced persons, urbanization, the high level of migration, and the small size of new dwellings in urban apartment houses all tend to diminish the cohabitation of parents and adult children, increasing the loneliness of old couples and unmarried old persons.

These are tendencies of modern life; nevertheless, it might be hypothesized that long-standing national characteristics, cultural values, beliefs, and norms continue to condition the living conditions of the older population even in advanced societies (Altergott 1988). From the data presented here, it seems that the cohabitation of elderly parents with their adult married children and the intensive help provided to elderly parents by their children are long-standing norms in Hungarian culture. It might be hoped that at least as long as the present hardships of an important part of the old population continue, their children will be willing to provide the financial help and the personal care that the underdeveloped welfare state is unable to provide.

References

Altergott, Karen. 1988. "Daily life in later life: Concepts and methods of inquiry." In Daily life in later life , ed. Karen Altergott, 11-22. Newbury Park, Calif.: Sage.

Andorka, Rudolf. 1978. Determinants of fertility in advanced societies . London: Methuen.

———. 1988. A családrekonstituciós vizsgálat módszerei [The methods of family re-constitution]. KSH Népességtudományi Kutató Intézet Történeti Demográfiai Füzetek, no. 4.

———. 1990. "The role of the family in the care of the elderly in Hungary." In Aiding and aging: The coming crisis in support for the elderly by kin and state , ed. John Mogey, 35-48. New York: Greenwood Press.

Andorka, Rudolf, Laszlóné Babarczy, László Cseh-Szombathy, Béla Ghyczy, and András Laktos. 1972. Az öregek helyzete és problemai [The conditions and problems of the aged]. Budapest: Central Statistical Office.

Andorka, Rudolf, and Sándor Balázs-Kovács. 1986. "The social demography of Hungarian villages in the eighteenth and nineteenth centuries, with special attention to Sárpilis, 1792-1804. "Journal of Family History 11: 169-192.

Andorka, Rudolf, and Tamás Faragó. 1983. "Pre-industrial household structure in Hungary." In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 281-307. Cambridge: Cambridge University Press.

Andrásfalvy, Bertalan. 1975. Duna menta népének ártéri gazdálkodása Tolna es Baranya vármegyében az ármentesités befejezéséig [The economy in the inundated area of the population living at the riverside of the Danube in counties Tolna and Baranya until the completion of the anti-inundation works]. Szekszard: Tolna megyei Tanacs Leveltara.

Barth, Janos. 1975. "Fajsz népessége a 18. század közepén" [The population of Fajsz in the middle of the eighteenth century]. Bacs-Kiskun megve multjabol 1: 81-131.

Berkner, Lutz K. 1972. "The stem family and the development cycle of the peasant household." American Historical Review 77 (2): 398-418.

Czap, Peter. 1983. "'A large family, the peasant's greatest wealth': Serf households in Mishino, Russia, 1815-1858." In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 105-151. Cambridge: Cambridge University Press.


154

Danhieux, Luc. 1983. "The evolving household: The case of Lampernisse, West Flanders." In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 409-420. Cambridge: Cambridge University Press.

Dányi, Dezsö. 1965. "Városi háztartások és családok a 18. század végén, Györ, 1787" [Urban households and families at the end of the eighteenth century]. Történeti Statisztikai Évkönyv 1963-1964 3: 5-204.

Faragó, Tamás. 1977. "Háztartásszerkezet és falusi társadalom-fejlödés Magyarorszá-gon 1787-1828" [Structure of households and development of rural society in Hungary, 1787-1828]. Történeti Statisztikai Tanulmányok 3: 105-214.

Farber, Bernard, John Mogey, lone DeOllos, and Robert A. Lewis. 1990. "Household composition among the elderly in the United States and Hungary: A comparison." In Aiding and aging: The coming crisis in support for the elderly by kin and state , ed. John Mogey, 11-33. New York: Greenwood Press.

Fél, Edit, and Tamás Hofer. 1969. Proper peasants . Chicago: Wenner-Gren Foundation.

Flinn, Michael W. 1981. The European demographic system 1500-1820 . Brighton: Harvester Press.

Gaunt, David. 1983. "The property and kin relationships of retired farmers in northern and central Europe." In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 249-279. Cambridge: Cambridge University Press.

Hajnal, John. 1965. "European marriage patterns in perspective." In Population in history , ed. D. V. Glass and D. E. C. Eversley, 101-143. London: Edward Arnold.

———. 1983. "Two kinds of pre-industrial household formation systems." In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 65-104. Cambridge: Cambridge University Press.

Kertzer, David I. 1989. "The joint family household revisited: Demographic constraints and household complexity in the European past." Journal of Family History 14(1): 1-15.

Laslett, Peter. 1972. "Introduction: The history of the family." In Household and family in past time , ed. Peter Laslett, with Richard Wall, 1-89. Cambridge: Cambridge University Press.

———. 1977. Family life and illicit love in earlier generations . Cambridge: Cambridge University Press.

———. 1983. Family and household as work group and kin group: Areas of traditional Europe compared. In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 513-563. Cambridge: Cambridge University Press.

———. 1988. "Family, kinship and collectivity as systems of support in pre-industrial Europe." Continuity and Change 3: 153-176.

Mándoki, Lászlo. 1971. "A kölkedi népszámlálás" [The conscription of Kölked in 1816]. Janus Pannonius Muzeum Evkonyve 13:215-224.

McIntosh, Marjorie K. 1988. "Local responses to the poor in late medieval and Tudor England." Continuity and Change 3: 209-246.

Palli, Haldur. 1983. "Estonian households in the seventeenth and eighteenth centuries." In Family forms in historic Europe , ed. Richard Wall, with Jean Robin and Peter Laslett, 207-216. Cambridge: Cambridge University Press.

Pullan, Brian. 1988. "Support and redeem: Charity and poor relief in Italian cities from the fourteenth to the seventeenth century." Continuity and Change 3: 177-208.


155

Tárkány-Szücs, Ernö. 1981. Magyar jogi népszokások [Hungarian legal folk customs]. Budapest: Gondolat.

Vásáry, Ildiko. 1989." 'The sin of Transdanubia': The one-child system in rural Hungary." Continuity and Change 4: 429-468.

Viazzo, Pier Paolo. 1989. Upland communities: Environment, population, and social structure in the Alps since the sixteenth century . Cambridge: Cambridge University Press.


156

Five
Migration in the Later Years of Life in Traditional Europe

Andrejs Plakans and Charles Wetherell

The spatial mobility of persons in the later years of life in traditional Europe remains virtually unexplored. Although the subject should be part of research on migration and old age, the current agendas in both fields do not lend themselves to merging for several reasons. For one, historians of migration and old age focus primarily on demographically and economically "modern," "modernizing," "industrial," or "industrializing" societies, a tendency attributable to the prominence of modernization theory in early debates on aging as well as to the indisputably profound level of migration in the Western world during the last three centuries (Moch 1992). Thus historians have ignored the migration of the elderly in the distant past, where discussions of "traditional Europe" should properly be lodged, in part because they perceive the action to have been in the recent past.

For another, in mapping the demographic landscape of both traditional and modern Europe, demographers have devoted more attention to mortality and fertility than to migration generally and far more to the movements of those who were actively involved in productive labor and the search for it than of those who were withdrawing from it or had completed the process altogether (Knodel 1988; Smith 1981; Wall 1984). We also know more about the western and central parts of Europe than about southern and eastern areas, although recent changes in archival access in eastern Europe promise to rectify some of the geographic imbalance. Moreover, demographers of modern Europe have focused on older segments of populations because these became proportionately larger after the demographic transition and now generate immense concern as the population of the Western world "ages" in the twentieth century.

Although historians have found the analytical imperatives of modernization theory and the demographic realities of the modern world less corn-


157

pelling in recent years (Hareven 1976; Jackson and Moch 1989; Moch 1983, 1992), simply shifting attention from the recent to the distant past cannot instantly fill the void in our historical understanding of migration among the elderly in traditional European societies. A disaffection with the explanatory power of structural differences between the traditional and modern worlds (as yet temporally undefined in any meaningful way) and a preference for attitudinal change that characterizes much current research handicap historians of inarticulate, traditional European populations. Yet major structural differences, such as those embodied in the institution of serfdom, do separate the distant and more recent European past. Whether they are important to the history of the mobility of the elderly remains to be seen. At the very least, the meager knowledge historians possess about the old in traditional Europe, particularly the migrating old, suggests the need to avoid simplistic or anachronistic conceptualizations. Moch's (1992) recent survey of migration in western Europe since 1650 goes a long way to correct earlier notions, but historians of the elderly still face a formidable task.

Historians dealing with the migration of older Europeans in the twentieth century, for example, draw heavily on labor economics and view movement among the elderly principally as retirement migration (Cribier 1974, 1982; Law and Warnes 1982). While a useful perspective for societies in which age grading is prominent, the notion seems decidedly unhelpful for traditional Europe, where withdrawal from the world of work was piecemeal and the elderly worked until infirmity or death (Plakans 1989). Any sharp contrast between the independence of the traditional and modern European aged may also need to be rethought, as has proved to be the case for the United States (Achenbaum 1978; Fischer 1977; Gratton 1986). In a study of retirement migration in contemporary France, Françoise Cribier (1974: 361) argued that "the proportion of the elderly in the population of a region, which used to be a function only of the birth rate and the exodus or in-migration of young people, today depends also on the actual behavior of the elderly" (emphasis added). Cribier never says why the elderly in societies that used to be did not exercise as much choice, but her premise seems to be that decision-making talent marked only the better-endowed modern old.

Finally, the notion of migration itself poses analytical difficulties for any analysis of the traditional past. Geographers and demographers routinely distinguish between migration and mobility . Conceptually, they base that distinction on the distance people move and on the social impact of that movement. Changing residence constitutes mobility if work, for example, remains close enough not to disrupt other social activities; it constitutes migration if work is not close enough to avoid disruption. Practically, however, the distinction often translates into movement across adminis-


158

trative boundaries, which can pose enormous strategic problems for historians forced to deal with records from separate secular or ecclesiastical divisions.

Beyond the seemingly simple yet often intractable distinction between migration and mobility, researchers also commonly differentiate among types of movement. In addition to local migration in which the distance moved is generally short and the accompanying level of social dislocation is low, Charles Tilly (1978: 51-57; see also Moch 1992: 16-17) distinguished among circular (distance unimportant, but return involved), chain (persisters following movers), and career (long-distance, no return) migration. Despite their considerable heuristic value, the underlying context of each type of migration is economic (specifically, labor markets) and so primarily useful for evaluating migration in the recent past. David I. Kertzer and Dennis P. Hogan's (1989) and Leslie Page Moch's (1992) studies of the nature and context of local, regional, and continental patterns of migration in western and southern Europe have greatly advanced our understanding of the phenomenon by showing how larger, often state induced, changes in the "fundamental structures of European economic life: landholding, employment, demographic patterns, and the development of capital" (Moch 1992: 6) affected migration. Yet if movement was always a central feature of European life, what Tilly and others term "local flows" (Tilly 1978: 63) remain the least investigated—and perhaps most common—in traditional European societies where the spatial worlds of work and community life were constrained by law and habit.

The major obstacle to a sharper image of migration in the traditional European past lies, of course, in the nature and extent of available historical sources. Few communities exist with records as extensive as Casalecchio in Italy (Hogan and Kertzer 1985; Kertzer and Hogan 1985, 1989, 1990), and therefore analysis almost always has to be data rather than problem driven. As existing studies of the more distant past make abundantly clear, refining the question to ask who moved at certain ages compounds the difficulties because of the nearly complete absence of age-specific migration information in standard historical sources such as parish registers and census enumerations. And if we pose the question in an even more specialized form to deal only with people in the later years of life, comparative opportunities narrow even further. We find, for example, that R. S. Schofield's analysis (1987) of the unique population listing for Cardington parish in eighteenth-century England deals only with the mobility of persons to about age 30 because the list permits no more; that Peter Clark's (1987) study of migration in England in the late seventeenth and early eighteenth century groups the elderly with everyone over 40; and that Jan Lucassen's study (1987) of European migrant labor in the period 1600-1900 contains no age-specific in-


159

formation. If the elderly appear at all in work on historical migration, it is most often as an afterthought.

A Case Study: Pinkenhof, 1833-1850

Our dissatisfaction with existing work stems from our continuing investigation of an eastern European serf estate, Pinkenhof, in the Russian Baltic province of Livland, now Latvia, between 1790 and 1850 (Plakans and Wetherell 1988a , 1988b , 1988c , 1990, 1992). Until 1819, the peasants of Pinkenhof were serfs, and their movements across estate boundaries were severely restricted. The Peasant Emancipation Law of 1819 (Schwabe 1928; Tobien 1899) introduced personal freedom and expanded the right of movement, both of which the Livlandic nobility viewed as necessary prerequisites to a free labor market. Yet the new law did not permit absolute freedom of movement (that right was introduced gradually to different segments of the peasant population) and formally deprived the peasantry of even those usufruct rights to land they had enjoyed under the old estate regime of serfdom. In principle, peasants were free to sell their labor to the highest bidder. In reality, now landless, peasants did not move away but continued to occupy their old farmsteads in exchange for money rents. Although the population possessed both traditional and modern demographic attributes that indicate it was moving swiftly into the demographic transition (Plakans and Wetherell 1988a ), the decades immediately following emancipation in Pinkenhof were ones of gradual adjustment and not rapid change.

The Russian Imperial head tax censuses, or "revisions of souls," for Pinkenhof, which provide detailed enumerations of the human groupings at the farmstead level from 1782 onward, reveal no massive in- or out-migration in either the pre- or the postemancipation periods, although both kinds of movement existed to varying degrees. The 1850 revision, however, was more than a simple nominal listing of 1,569 residents living on 123 farmsteads, the main estate farm, or Hof , and several smaller, functionally specialized places; it indicated where each of the 1850 residents had lived in 1833. Moreover, if an 1833 inhabitant had left the estate before 1850, his or her departure was noted, together with a date and a destination; if he or she had arrived since 1833, that was noted also, together with the place of origin although not always with the year of arrival. The 1850 revision, therefore, allows us to explore external migration and, to a lesser extent, internal mobility.

As in most Baltic landed estates, everyday rural life in Pinkenhof transpired on spatially separated farmsteads, not nucleated villages. Each of the 123 fixed residential farms in 1850 bore a name that recurs in estate documents as far back as the latter part of the seventeenth century. New entrants


160

TABLE 5.1
Age-Specific External Migration Rates, Pinkenhof, 1833-1850


Age Cohort

Midperiod Population

Number of Migrants

External Migration Rate (CMRe )

0-14

490

7

0.8

15-19

162

17

5.8

20-24

140

50

19.8

25-29

145

84

32.2

30-34

129

21

9.0

35-39

104

7

3.7

40-44

85

3

2.0

45-49

63

2

1.8

50-54

61

1

0.9

55+

93

2

1.2

 

Total

1,472

194

7.3

SOURCE: The Ninth Imperial Revision for Livland, Central National Historical Archive, Riga, Latvia (Baltic Microfilms, D112, Oekonomie Expedition d. Stadtkassakollegiums IV E. 4, Revisionsliste Gut Pinkenhof, J. G. Herder Institut, Marburg a.d. Lahn, Germany).

into the estate, therefore, augmented the labor force of particular farmsteads, and those who left diminished it. Correspondingly, internal mobility, including that of the elderly, took place between farmsteads, rather than between farmsteads and institutions reserved for the aging, or between a farm's main residential quarters and outbuildings set apart for the aged as was frequently the practice in central Europe (Mitterauer and Sieder 1977: 162-163). The main building of a farmstead was therefore the residential site of all members, including the aging and aged; and judging by the architecture of these buildings, the living space within could be readily adjusted to accommodate any increase or decrease in residents (Kundzins 1974; Veveris and Kuplais n.d.). As such, housing the marginally productive elderly—if indeed the elderly can be thought of in this way at all—was not a serious problem, and we have to seek the reasons for their movement elsewhere.

But Pinkenhofers did move. Between 1833 and 1850, 192 men and women left and entered Pinkenhof, for a net migration rate of -17.7. For females, the rate was 4.1 and for males, -21.7, a discrepancy attributable to a high level of conscription among males (Plakans and Wetherell 1988a ). Table 5.1 displays age-specific migration rates for both males and females; figure 5.1 gives a stylized age profile.

Overall, the crude external migration rate, CMPe , was 7.3, but the incidence of migration was greatest for those in their 20s (30.3) who were moving in and out of the estate to marry.[1] At the same time, the neighboring estate of Bebberbeck and others in the adjoining province of Kurland, which


161

figure

Fig. 5.1.
Stylized age profile of external migrants, Pinkenhof, 1833-1850. 
Data from the Ninth Imperial Revision for Livland, Central
 National Historical Archive, Riga, Latvia (Baltic Microfilms, 
D112, Oekonomie Expedition d. Stadtkassakollegiums IV E. 4, 
Revisionsliste Gut Pinkenhof, J. G. Herder Institut, Marburg 
a.d. Lahn, Germany).

were the sources of half (44 of 84) of all immigrants and half (38 of 74) of all emigrants (excluding 36 conscripted males), were so geographically close to Pinkenhof that any hard and fast distinction between migration and mobility may blur the historical reality. Nonetheless, the record indicates that external migration, defined as movement across the estate's boundaries, was limited almost exclusively to young adults; for those over 40, it was virtually nonexistent.

Internal mobility is more difficult to evaluate for two reasons. First, the revision only documents the presence of a person in one farmstead in 1833 and in another in 1850; it does not record any intervening moves that might have occurred between those two years. Second, the source does not allow us to say at what ages internal migrants moved. All we know are the ages of 459 people who were at least 18 years old in 1850 and who had changed their farmsteads of residence at least once between 1833 and 1850. At the same time, it seems safe to conclude that movement within the estatemobility —was much greater than movement across estate boundariesmigration . For one thing, the ratio of recorded internal to external moves was more than 2:1. For another, the record keepers were especially careful to document movement into and out of the estate. Indeed, 2 of the 192 migrants both entered and left Pinkenhof between 1833 and 1850. Finally, we also know that


162

TABLE 5.2
Hypothetical Age-Specific Internal Migration Rates and Proportions of 1850 Population Internally Mobile, Pinkenhof, 1833-1850


Age Cohort


1850 Population

Number of Migrants

Internal Migration Rate (CMR1 )

Percentage of Population Mobile

18-19

76

30

21.9

35.9

20-24

139

79

31.6

56.8

25-29

136

75

30.6

55.1

30-34

125

72

32.0

57.6

45-39

106

53

27.8

50.0

40-44

104

43

23.0

41.3

45-49

85

39

25.5

45.9

50-54

72

31

23.9

43.1

55+

126

37

16.3

29.4

 

Total

969

459

26.3

47.4

SOURCE: The Ninth Imperial Revision for Livland, Central National Historical Archive, Riga, Latvia (Baltic Microfilms, D112, Oekonomie Expedition d. Stadtkassakollegiums IV E. 4, Revisionsliste Gut Pinkenhof, J. G. Herder Institut, Marburg a.d. Lahn, Germany).

the labor force in Pinkenhof was far from stationary and that adult farmhands and their children, who accounted for 43 percent of the population in 1850, traditionally moved about the estate on a regular basis (Plakans and Wetherell 1988b , 1992; Svarane 1971). Accordingly, we take the incidence of external migration in the 1850 revision to be a good reflection of the historical reality and the corresponding level of internal mobility as an absolute minimum.

Table 5.2 displays hypothetical age-specific internal migration rates for those 18 years of age and older in 1850 and figure 5.2, a stylized age profile. The age-specific rates are hypothetical and cannot be taken at face value because they reflect only movement sometime between 1833 and 1850 and do not represent standardized rates of mobility for persons in each age cohort. Only if the 63 peasants who were between 45 and 49 years old in 1850, for example, had actually moved while they were 45 to 49 would the reported age-specific rate be valid. Yet a crude internal migration rate, CMR1 , of 26.3 suggests that spatial mobility was a common experience in Pinkenhof. Indeed, table 5.2 also reveals that nearly half (47.4 percent) of the 1850 population at risk had moved at least once within the estate since 1833. We also know that internal migrants did not move very far, on the average only 2.5 kilometers. Thus local flows, although common, were short (Plakans and Wetherell 1988a ; Wetherell, Plakans, and Wellman 1994).

The views of external and internal movement that the 1850 revision provides suggest two scenarios that define the probable extremes of the migratory experience of the elderly in Pinkenhof. Both, however, indicate low levels of movement. On the one hand, if the age profile of internal migrants


163

figure

Fig. 5.2.
Stylized age profile of internal migrants, Pinkenhof, 1833-1850. 
Data from the Ninth Imperial Revision for Livland, Central
 National Historical Archive, Riga, Latvia (Baltic Microfilms, 
D112, Oekonomie Expedition d. Stadtkassakollegiums IV E. 4, 
Revisionsliste Gut Pinkenhof, J. G. Herder Institut, Marburg a.d.
 Lahn, Germany).

that table 5.2 reveals is grossly wrong and the real pattern of internal mobility actually resembles that of external migration presented in table 5.1, then the spatial mobility of the elderly (those 55 and over) was extremely limited. Although many may have moved during their lives (and a CMR1 > 25 attests to this), those moves would have taken place twenty to thirty years before they reached old age.[2] So in one scenario, the elderly rarely, if ever, moved. On the other hand, if a substantial minority of Pinkenhofers moved regularly from farm to farm in the course of their lives as nonquantitative research on these peasants maintains (Strods 1972; Svarane 1971), then the pattern of mobility that table 5.2 presents may well reflect the collective migratory experience of this particular peasant community. In this scenario, the elderly moved (CMR1 << 20) but still not as often as young adults (CMRi > 30).

Despite the weaknesses in the record, several things seem clear about mobility and migration in Pinkenhof in the decades immediately following emancipation:

1. Most spatial movement consisted of local, short-distance moves within the estate or to neighboring estates.

2. Internal mobility was common. Perhaps half of all peasants changed their residence at least once in their adult years.


164

3. External migration was confined to young adults who moved largely to and from neighboring estates in order to marry.

4. Movement of those over 30 was virtually always within the estate.

5. The elderly either (a ) moved frequently within the estate but less often than those in their 20s, 30s, and 40s or (b ) moved very infrequently, if at all, after reaching their 30s.

6. By virtue of their number, internal migrants, far more than external migrants, created the need for any social and psychological adjustments that may have accompanied movement.

In light of these findings, it is unsurprising that mid-nineteenth-century social commentary has little to say on the subject of migration among those in the later years of life.

Pinkenhof Patterns in Comparative Perspective

Placing the migratory experience of the Pinkenhof elderly in comparative perspective poses two basic problems: one involves evidence, the other actual behavior. A logical case could be made against finding much usable information of any kind about the old in the historical sources common to traditional European societies.[3] When they did so at all, record keepers enumerated populations for specific purposes, such as assessment of taxes and labor obligations, and had little incentive to make a careful record of those persons who had withdrawn from roles that entailed such involvements. Although we have no evidence of such carelessness in Pinkenhof, age heaping of the elderly is a well-documented attribute of historical European sources.[4] Moreover, we can easily imagine that, by virtue of a growing dependency on the young, older people were increasingly less likely to cross boundaries of administrative units such as parishes, departments, estates, or provinces. Thus the likelihood of the elderly showing up in those few sources that were primarily concerned with boundary crossers diminishes even further.

At the same time, the record of external migration in Pinkenhof between 1833 and 1850 suggests a fundamental immobility among those over 55 that is not without modern parallels. In 1968 in France, for example, 86.5 percent of persons 55 and older still resided in the same commune as they had in 1962 (Cribier 1974: 362). The 13.5 percent who had moved to another commune during that period, moreover, represented a 35 percent increase over the years from 1956 to 1962. Although external migration (55+ CMRe = 144.7) among the elderly in contemporary France was much greater than in Pinkenhof (55+ CMRe = 1.2), it still suggests a basic propensity among those in the later years of life not to migrate. If less than one-fifth of the elderly in a population with access to modern transportation, communica-


165

tion, and information systems chose to migrate, it seems likely that in traditional societies without such systems, where mental maps extended only to the edges of landed estates, older people chose to stay close to or at the sites of their past productive labor or to the residences of their family and kin. They may still have moved, but their movements could be measured only with household-level sources that serially enumerated (at a minimum by name, age, sex, and residence) the same local population. Accordingly, the "inattention" to the elderly that Gerald N. Grob (1986: 33) argues has characterized historical migration research may in fact be a matter of definition. By focusing exclusively on migration across administrative boundaries, historians may have overlooked that domain in which most elderly in the traditional European past moved, if they moved at all. In Pinkenhof at least, mobility was limited almost entirely to the bounded community. The elderly may have been mobile , but they did not migrate .

Studies of larger patterns of migration in the twentieth-century Western world provide another point of comparison that suggests a fundamental structural difference between the recent and the distant European past. Figure 5.3A presents a stylized, three-part migration schedule for post-World War II Europe and the United States (Rogers and Castro 1986). The underlying model predicts that the rate of migration will start high but decrease rapidly in the first fifteen years of life as children (the prelabor force) move with young adult parents (the labor force). Migration will then increase sharply to its maximum levels among young adults in their 20s, peak at about age 30, and decline quickly thereafter. Rates will continue to drop until just after retirement and then will show a slight, short-lived upward turn corresponding to retirement migration (the postlabor force). Moreover, figure 5.3B reveals not only that the basic migration schedule applies to both migration and mobility but also that the incidence of local moves exceeds that of long-distance moves. In the modern world, then, (1) the level of local, intracommunity mobility is higher than the corresponding level of external, intercommunity migration, and (2) the age profile of both kinds of movers is the same. Both points allow us to place migration among the elderly in Pinkenhof in a long-term context.

First, the record of residential movement in both the modern and the traditional European past indicates that mobility invariably exceeded migration. Second, the age profile of external migration in Pinkenhof (fig. 5.2) resembles the modern European schedule (fig. 5.3A). Although Pinkenhof's external migration did not possess a prelabor force component of children, the nature of the peasant marriage market, in which brides and grooms moved at the time of marriage, would logically work to minimize the number of children involved. Consequently, we might speculate that in traditional peasant societies when young adults migrated across community boundaries, they did so without spouses or children. Third, the similar age


166

figure

Fig. 5.3.
Stylized migration schedules: A, generalized; B, external
 and internal. Adapted from Rogers and Castro 1986,
 pp. 172 and 162, respectively.


167

profiles of both the modern mobile and the modern migrant suggest that we have to choose between two basic interpretations of the incomplete record of internal mobility in Pinkenhof. In essence, the question is whether the migration schedule in figure 5.3A reflects the historical reality of mobility in Pinkenhof or whether the hypothetical migration schedule in figure 5.2 is a better guess. Clearly the evidence at hand does not allow us to say with any certainty. A look at the motives behind both modern and traditional European moves provides help.

The 31 percent of the population over 55 in Pinkenhof in 1850 who had moved at least once since 1833 no doubt had good reason for doing so, although their motives remain hidden from us. At the same time, we have no reason to believe that decision making by the old—and by those family members and friends whose actions would affect the old—was significantly less patterned than in contemporary populations. Cribier found that no single reason predominated among the economic, familial, and personal factors that elderly couples considered before deciding to leave Paris in the early 1970s.

The nine most commonly stated reasons given by either partner for leaving Paris were (1) nothing to do in Paris after retirement . . .; (2) Paris had become unbearable . . .; (3) Paris is unhealthy . . .; (4) did not want to go on living in a flat . . .; (5) wanted a change of climate . . .; (6) wanted to go back home . . .; (7) for family reasons other than reason 6 . . .; (8) cost of living too high in Paris . . .; (9) eviction. (Cribier 1982: 117)

Although impossible to weigh with much precision, Cribier's respondents seemed to be concerned with family, health, housing, and leisure time. Yet they lived in a period of European history when retirement from remunerative employment had become a well-marked life course transition. They also lived in a modern, industrial society where the choices available to them were more numerous than in the traditional European past. Indeed, Imhof (1981) has argued that all life patterns involving the elderly changed entirely with the arrival of the contemporary world.

The older inhabitants of Pinkenhof, more than a century earlier, may well have shared the same concerns of housing, health, and family that Cribier's respondents voiced, but they certainly had fewer options. Their reasons for moving at all probably involved a mixture of both push and pull factors. Existing studies of estate policy toward peasant farmsteads suggest that push may have dominated. If an energetic estate owner added a particular farmstead's land to the demesne, the peasant holding would be broken up (spridzinasana , lit. "blowing up the farmstead") and its residents relocated. The elderly in these instances would most likely become farmhands elsewhere in the estate. Alternately, elderly persons might move to another peasant farmstead, which would receive some remuneration for serving as


168

a kind of private retirement home, or to the Hof . Similar social welfare functions are well documented for other traditional European communities and certainly existed in Pinkenhof in both the pre- and postemancipation periods. On the basis of our evidence, we cannot say who actually decided that the elderly would move. Assuming a general proclivity among the elderly not to move, however, would suggest that other, younger people probably initiated discussions that led to any movement. Yet hovering over most peasant decisions, particularly those affecting the operation of the estate's farmsteads, was the estate owner, whose own concerns were not likely to have been much influenced by discomfort and anguish among elderly peasants.

What we know about eastern European peasant estates in general and Pinkenhof in particular also suggests that the different life experiences of farmstead heads and their families, on the one hand, and of farmhands, on the other, affected the mobility of the elderly. Of the 817 adults over age 20 in Pinkenhof who lived on the 119 farmsteads with identifiable heads in 1850, half (416) were heads themselves or co-resident kin. Of this privileged group, nearly two-thirds (268) resided on the same farmstead they had in 1833. Conversely, of those 401 adults who were not related to the head of the farmstead on which they lived in 1850, three-fourths (304) had moved at least once since 1833. Among the 121 peasants over age 55 in 1850 on these 119 farmsteads, two-thirds (86) had not moved since 1833. And for those 76 elderly peasants who were fortunate enough to be related to the head of one of Pinkenhof's farmsteads in 1850, 8 of 10 had not moved in the previous eighteen years. If elderly Pinkenhofers valued residential stability, and the emotional and physical support it undoubtedly brought, a kinship tie with a farmstead head clearly helped them achieve the goal (Plakans and Wetherell 1992; Wetherell, Plakans, and Wellman 1994).

For the half of the population who were farmhands, the pressure to move was far more pronounced and regular, for they faced an annual search for new employment. If a farmstead head decided his current farmhands were a drain on resources or were poor workers, he could always let them go and bargain with others to take their places (Plakans and Wetherell 1988b , 1992). In these situations, any change of residence by the elderly would have been involuntary if they were being dismissed and voluntary if they were being recruited from another farmstead. In either event, the elderly would move. Family-linked moves by the elderly could also be the result of married sons improving their lot by assuming headships on better farmsteads. Whether farmhands moved every year or every five or ten years, we can well imagine a cycle of movement that created enough turnover to give rise to the traditional folk view that the population of peasant estates was constantly churning.

The experiences of Pinkenhof's peasants, particularly those of farmhands and their families, suggest that the age profile of internal mobility in traditional eastern European peasant communities was fundamentally different


169

from what it is in the modern world. Enough movement can be either documented or inferred to support the contention that the hypothetical internal migration schedule in figure 5.2 better captures the historical reality in Pinkenhof than the modern schedule in figure 5.3A. Although adults in their 20s and early 30s still moved more frequently than those in their late 30s and 40s, the incidence of mobility among middle-aged peasants was probably greater in the distant past than in the recent past. At the same time, mobility among the elderly was arguably less pronounced in traditional European societies for both structural and attitudinal reasons.

We cannot say with certainty why elderly Pinkenhofers moved, if they did at all, and their reasons may have been more numerous than we have inferred. Certainly Cribier's and other studies of contemporary populations suggest as much. Stanley H. Brandes, for example, found that in the twentieth-century Spanish village of Becedas, "as long as both elderly parents are still alive, it is considered heartless to break the continuity of their lives by asking them to leave their home. After one parent dies, however, the disintegrated nuclear unit no longer justifies the maintenance of a separate household, and the widowed individual must accommodate himself to the homes of his or her children" (1975: 110). Social customs of this kind, easily discovered by questioning living populations but impossible to glean from most traditional European historical sources, warn against oversimplifying behavior in the distant past.

It is no oversimplification, however, to note that more than two-thirds of peasants 55 and older in 1850 had not moved in the preceding eighteen years. Staying in place among the old was consistent with two general structural features of peasant communities: first, neither men nor women ever completely withdrew from farmstead labor (Plakans 1989); second, peasants had probably moved several times before the), reached old age. Describing Russian peasant villages at the end of the nineteenth century, Adele Lindenmeyr (1982: 232-234) observed that old men and old women "performed essential tasks even if they could no longer work in the fields: fetching water, chopping wood, preparing food, cultivating kitchen gardens, rocking cradles, and minding children, chickens and geese." Similarly in Pinkenhof, the elderly would not have remained taskless in view of the dozens of light but important chores farmstead life entailed; and performing them maintained a link with the world of work that was rarely, if ever, severed. Because elderly Pinkenhofers were not strangers to movement, we do not have to idealize their lives to conclude that a fixed abode, more than likely with married children in the vicinity, if not on the same farmstead, was a desirable goal.

Residential stability—retirement immobility—was well within the realm of ways European peasants dealt with the elderly and something the elderly could realistically expect as a form of support. Peter Laslett's (1988) survey


170

of this diversity—cast as an exploration of the different ways family and community acted as a support system for those who required it—need not be repeated here, but one of his observations helps to place the Pinkenhof experience in space and time.

Old people were sustained by a whole range of expedients in which family and collectivity collaborated within the customary framework as the situation required. An important reason for this was that there was no standard situation in which the necessity of support arose, but a whole array of differing situations, differing from time to time, circumstance to circumstance, individual to individual. (Laslett 1988: 168)[5]

Pinkenhofers formed families and lived in them. Yet those families, by virtue of the traditional residential system that prevailed in the Baltic, were situated on farmsteads that often contained two or more conjugal family units with children and other relatives. Pinkenhof's "family system" could not possibly have allowed peasants to realize all their values independently, for they always had to reckon with the economic imperatives of the "farmstead system." The latter tended to feel the managerial hand of the estate owner (in the case of Pinkenhof, the city of Riga), and therefore we find it impossible to believe that mobility of the elderly was in any sense a pure expression of the values that drove either the family system, the estate owner, or the elderly themselves. Indeed, the world of Pinkenhof's peasants was less one of "perennial households" (Laslett 1988: 158ff., after Czap 1982) than one of "perennial farmsteads," which, together with the families in them, were the two "collectivities" that cooperated to provide sustenance and support to the elderly and reduce both their migration and their mobility.

Until we know more about the rural populations of eastern Europe after emancipation, we hesitate to generalize from Pinkenhof to any other part of the traditional European east, let alone all of Europe in the distant past. We would contend, however, that the record of migration and mobility in Pinkenhof provides an initial point of reference, certainly for the Baltic area and possibly for much of the European east. Indeed, movement in nineteenth-century Pinkenhof arguably reflects long-standing behavior because the structural constraints of law and imperatives of the farmstead-based agricultural regime operated well past emancipation. Pinkenhof's population was subject to restrictions typical of postemancipation eastern European societies. Peasants were no longer serfs, but, notwithstanding the free labor market philosophy behind emancipation, they were not absolutely free to move either.

The great increase of rural-to-urban and rural-to-rural migration in the Baltic did not begin until the 1870s, when it increased the region's cities and towns at an unprecedented pace (Anderson 1980; Leasure and


171

Lewis 1968; Winner and Winner 1984). How the elderly participated in that migration transition remains to be seen, but in the decades immediately following emancipation, elderly peasants in the main stayed put. Permanence of place was the hallmark of the traditional elderly that distinguished them from their modern twentieth-century counterparts. The phenomenon of retirement migration did not appear in all European societies with the advent of the modern world, but where it did, it clearly announced the arrival of a new feature in the history of the elderly. In Pinkenhof as in most traditional European communities, there was no retirement migration because retirement in the modern sense simply did not occur.

References

Achenbaum, Andrew W. 1978. Old age in the new land: The American experience since 1970 . Baltimore: Johns Hopkins University Press.


172

Anderson, Barbara A. 1980. Internal migration during modernization in late nineteenth-century Russia Princeton: Princeton University Press.

Brandes, Stanley H. 1975. Migration, kinship, and community: Tradition and transition in a Spanish village . New York: Academic Press.

Clark, Peter. 1987. "Migration in England during the late seventeenth and early nineteenth centuries." In Migration and society in early modern England , ed. Peter Clark and David Souden, 213-252. London: Hutchinson.

Cribier, Françoise. 1974. "Retirement migration in France." In People on the move: Studies on internal migration , ed. Leszek A. Kosinski and R. Mansell Prothero, 361-373. London: Methuen.

———. 1982. "Aspects of retirement migration from Paris: An essay in social and cultural geography." In Geographic perspectives on the elderly , ed. A.M. Warnes, 111-137. London: John Wiley and Sons.

Czap, Peter. 1982. "The perennial multiple family household, Mishino, Russia, 1782-1858." Journal of Family History 7: 5-26.

Fischer, David Hackett. 1977. Growing old in America . New York: Oxford University Press.

Gratton, Brian. 1986. "The new history of the aged: A critique." In Old age in a bureaucratic society , ed. David Van Tassel and Peter Stearns, 1-29. New York: Greenwood Press

Grob, Gerald N. 1986. "Explaining old age history: The need for empiricism." In Old age in a bureaucratic society , ed. David Van Tassel and Peter Stearns, 30-45. New York: Greenwood Press.

Hareven, Tamara. 1976. "Modernization and family history: Perspectives on social change." Signs 2: 190-206.

Hogan, Dennis P., and David I. Kertzer. 1985. "Migratory patterns during Italian urbanization, 1865-1921." Demography 22: 309-325.

Imhof, Arthur E. 1981. Die gewonnenen Jahre [The gained years]. München: C. H. Beck.

Jackson, James H., Jr., and Leslie Page Moch. 1989. "Migration and the social history of modern Europe." Historical Methods 22: 27-36.

Kahk, Juhan. 1982. Peasant and lord in the process of transition from feudalism to capitalism in the Baltics . Tallinn: Eesti Raamat Publishers.

Kertzer, David I., and Dennis P. Hogan. 1985. "On the move: Migration in an Italian community, 1865-1921." Social Science History 9: 1-23.

———. 1989. Family, political economy, and demographic change: The transformation of life in Casalecchio, Italy, 1816-1921 . Madison: University of Wisconsin Press.

———. 1990. "Household organization and migration in nineteenth-century Italy." Social Science History 14: 483-505.

Knodel, John E. 1988. Demographic behavior in the past: A study of fourteen German village populations in the eighteenth and nineteenth centuries . Cambridge: Cambridge University Press.

Kundzins, Pauls. 1974. Latvju seta [The Latvian farmstead]. Stockholm: Daugava.

Laslett, Peter. 1988. "Family, kinship and collectivity as systems of support in pre-industrial Europe: A consideration of the 'nuclear-hardship' hypothesis." Continuity and Change 3: 153-175.


173

Law, Christopher, and Anthony M. Warnes. 1982. "The destination decision in retirement migration." In Geographic perspectives on the elderly , ed. A. M. Warnes, 53-81. London: John Wiley and Sons.

Leasure, J. William, and Robert A. Lewis. 1968. "Internal migration in Russia in the late nineteenth century." Slavic Review 27: 376-394.

Lindenmeyr, Adele. 1982. "Work, charity, and the elderly in late nineteenth-century Russia." In Old age in preindustrial society , ed. Peter Stearns, 232-247. New York: Holmes and Meier.

Lucassen, Jan. 1987. Migrant labor in Europe 1600-1900: The drift to the North Sea . Trans. Donald A. Bloch. London: Croom Helm.

Mitterauer, Michael, and Reinhard Sieder. 1977. The European family: Patriarchy to partnership from the Middle Ages to the present . Trans. Karla Osterveen and Manfred Horzinger. Chicago: University of Chicago Press.

Moch, Leslie Page. 1983. Paths to the city: Regional migration in nineteenth-century France . Beverly Hills, Calif.: Sage.

———. 1992. Moving Europeans: Migration in Western Europe since 1650 . Bloomington: Indiana University Press.

Plakans, Andrejs. 1989. "Stepping down in former times: A comparative assessment of 'retirement' in traditional Europe." In Age structuring in comparative perspective , ed. David I. Kertzer and K. Warner Schaie, 175-195. Hillsdale, N.J.: Lawrence Erlbaum.

Plakans, Andrejs, and Charles Wetherell. 1988a . "The kinship domain in an East European peasant community: Pinkenhof, 1833-1850." American Historical Review 93: 359-386.

———. 1988b . "Unfree labor and family life in a Baltic serf estate, 1808-1816." Paper presented at the annual meeting of the Social Science History Association, Chicago, November 3-6.

———. 1988c . "Land and labor in a Baltic serf estate, 1809-1850." Paper presented at the annual meeting of the American Association for the Advancement of Slavic Studies, Honolulu, November 18-21.

———. 1990. "Transfer of headships in nineteenth-century eastern European serf estates." Paper presented at the annual meeting of the Social Science History Association, Minneapolis, October 18-21.

———. 1992. "Family and economy in a Baltic serf estate in the early nineteenth century." Continuity and Change 7: 199-223.

Rogers, Andrei, and Luis J. Castro. 1986. "Migration." In Migration and settlement: A multiregional comparative study , ed. Andrei Rogers and Frans J. Willekens, 157-208. Dordrecht: D. Reidel.

Schofield, R. S. 1987. "Age-specific mobility in an eighteenth-century rural English parish." In Migration and society in early modern England , ed. Peter Clark and David Souden, 253-266. London: Hutchinson.

Schwabe, A. 1928. Grundriss der Agrargeschichte Lettlands [The basic agrarian history of Latvia]. Riga: Bernhard Lamey.

Shyrock, Henry J., and Jacob S. Siegel et al. 1973. The methods and materials of demography . 2 vols. Washington, D.C.: Government Printing Office.

Smith, Daniel Scott. 1981. "Historical change in the household structure of the elderly in economically developed societies." In Aging: Stability and change in the fam-


174

ily , ed. Robert W. Fogel, Elaine Hatfield, Sara B. Kiesler, and Ethel Shanas, 91-114. New York: Academic Press.

Strods, Heinrihs. 1972. Lauksaimnieciba Latvija parejas perioda no feodalisma uz kapitalismu [Agriculture in Latvia in the transition period from feudalism to capitalism]. Riga: Zinatne.

Svarane, Melita. 1971. Saimnieks un kalps Kurzeme un Vidzeme XIX gadsimta vidu [Farmstead heads and farmhands in Kurland and Livland in the mid-nineteenth century]. Riga: Zinatne.

Tilly, Charles. 1978. "Migration in modern European history." In Human migration: Patterns and policies , ed. William H. McNeill and Ruth S. Adams, 48-72. Bloomington: Indiana University, Press.

Tobien, A. von. 1899. Die Agrargesetzgebung Livlands im 19 Jahrhundert [The agrarian laws of Livland in the nineteenth century]. Vol. 1. Berlin: Puttkamer and Mühlbrecht.

Veveris, Ervins, and Martins Kuplais. n.d. Latvijas Etnografiskaja brivdabas muzeja [In the Latvian ethnographic open-air museum]. Riga.

Wall, Richard. 1984. "Residential isolation of the elderly: A comparison over time." Ageing and Society 4: 483-503.

Wetherell, Charles, Andrejs Plakans, and Barry Wellman. 1994. "Social networks, kinship, and community in Eastern Europe." Journal of Interdisciplinary History 24: 639-663.

Winner, Irene P., and Thomas G. Winner, eds. 1984. The peasant and the city in Eastern Europe: Interpreting structure . Cambridge: Schenkman.


175

Six
Older Lives on the Frontier: the Residential Patterns of the Older Population of Texas, 1850-1910

Myron P. Gutmann

This chapter is about the role played by the moving American frontier in defining the domestic residential experiences of the families of the older population. Between 1850 and 1910, Texas had a great deal of open land, a young population, and substantial ethnic diversity. All these factors shaped the experiences of the older population. Within the state there were significant numbers of African-Americans, Mexican-Americans, recent European immigrants and their children, and white Americans who were the children of parents born in the United States.[1] The residential living arrangements of a population are the result of responses to other changes in their lives: in the case of the older population, the important turning points are the consequences of the aging of parents and children and the end of marriage through widowhood. Economic and cultural contexts shaped the nature of residential arrangements. Some older Texas couples seem to have preferred to live by themselves, while others preferred to live with a variety of other kin. Even where such a preference was held, the nature of working arrangements and the lack of property sometimes made it impossible for an older couple to gather their family around them. All this took place in the context of a growing and maturing regional population. As the regional population grew, older Texas residents had more kin nearby, which opened new possibilities for some of them to live with kin rather than with strangers.

The experience of aging in Texas occurred in a larger American context. The kinds of multiple sharecropper households described for northern Italy by David I. Kertzer and Nancy Karweit elsewhere in this volume were extremely rare in the United States, and the kinds of household arrangements they describe for widows and widowers were also rare. The population of Texas described in this chapter was a rural population that lived in small towns and on isolated farms and was employed in agriculture and in


176

services that supported agriculture.[2] Agricultural populations have more flexibility than urban industrial populations, such as those described by Tamara K. Hareven and Peter Uhlenberg in the American Northeast, with which to accommodate the needs of an older population, and that too is part of the message of this chapter.

During the second half of the nineteenth century there were more large and complex families in the United States than there had been previously, or than there would be since.[3] Moreover, older people in the United States were far more likely than most other members of the population to be living in complex families, most notably, with their married children, with other kin, or with nonkin. That is the conclusion of the research reported here and the research of others, especially Daniel Scott Smith, who directed an important project on the older population of the United States in 1900.[4] Yet even this situation was undergoing change. The results reported here show that as lives lengthened at the end of the nineteenth century, it became possible for men and women to spend more of their later years living together as a nuclear couple, without being forced to live with their married children. In many cases, it was only after the death of one spouse that a surviving widow lived with one of her or his married children.

Later in the twentieth century the residential patterns of the older population changed yet again. Fewer of them lived in complex domestic arrangements, more of them lived with their spouses, and many more of them—especially widowed women—lived alone (Riley 1982). This isolation of the elderly is one of the great changes in residential patterns that have taken place in the twentieth century, and its implications have been the subject of a considerable amount of study (Smith 1981, 1986). Despite its importance, and despite evidence that married older Texans were increasingly likely to live by themselves, the new isolation of the elderly in the twentieth century is not the principal subject of this chapter. Rather, I want to look at developments that led up to the kind of conditions that existed at the turn of the century.

The results reported here are drawn from data collected from the 1850, 1860, 1870, 1880, 1900, and 1910 U.S. manuscript censuses for six Texas counties: Red River, DeWitt, Gillespie, Webb, Angelina, and Jack.[5] These six counties taken together reflect the range of cultural, ethnic, economic, and ecological characteristics of the rural part of the state.

Three of the counties are located in the eastern half of Texas. Red River County, in the northeast corner of the state, was among the earliest areas settled extensively by immigrants from elsewhere in the United States. It had a native-born population of whites and blacks and an economy largely dominated by the production of cotton. Angelina County, in East Texas near the Louisiana border, also had a mixed white and black population, with a cotton economy supplemented by the production of timber and sugar. DeWitt


177

figure

Fig. 6.1.
Location of the six Texas counties analyzed.

County is located in Southeast Texas. It had the most varied population of any of the counties studied here, made up of German immigrants, native-born whites and blacks, and an increasing number of Mexicans in the early twentieth century. Its economy consisted of a mixture of cotton production and cattle ranching.

The remaining three counties are located in the central and western parts of the state. Webb County is located on the Mexican border and has as its seat the most important town in our study, Laredo. Its population was almost exclusively of Mexican origin, and it had a diverse economy of ranching and town and transportation services. Gillespie County, located in the Central Texas hills, was another haven for immigrants. Settled in 1846 by German immigrants, it remained largely German-American well into the twentieth century. Its economy was a diverse mix of agriculture and cattle, sheep, and goat ranching. Jack County was the last settled of those studied here. Located west of Fort Worth in North Texas, it was organized in 1852 and remained only scarcely settled until after 1870. With a population that consisted mostly of native-born whites, Jack County's economy was largely devoted to cattle ranching.

The data used in this chapter were collected as part of a larger project intended to study the dynamics of the rural population of Texas from 1850 to


178

1910.[6] A sample was drawn from each of the six counties' manuscript census returns, designed to permit us to link references to individuals from census to census and from the census to other documents.[7] The sampling proportions differed from county to county, depending on the size of the population (Gutmann and Fliess 1989). The choice of counties was made so as to overrepresent the European immigrant (especially German- and Slavic-speaking) and Mexican origin elements of the population. The analysis reported here uses data weighted to restore the actual statewide ethnic balance in each census year. Despite our efforts to be able to study sub-groups within the population, we have not completely succeeded. Slaves were recorded in a separate census schedule in 1850 and 1860. While those schedules are available and could be transcribed and analyzed, the living arrangements of slaves are not readily comparable to those of the free population. We therefore have no African-American population before 1870 (there were few free blacks in Texas, and virtually all lived in cities). Our sample of Mexican-Americans includes no data for 1850 because of the difficulty of working with the census manuscripts for Webb County, as well as other largely Mexican-American areas, in that year.[8]

The results reported here include many categories of individual, spread over a number of points in time. I have attempted to organize the data so that they yield the largest amount of information while reducing the amount of confusion produced by the volume of data points and categories. The data, which cover six decennial time points from 1850 to 1910, will often be reported in three groups, each consolidating two points in time. This has the advantage of reflecting three different eras in the evolution of the population of the state of Texas. The first period, 1850 and 1860, is the antebellum era of settlement. This is the era in which the population of Texas is small, and our sample is still quite small. The second period, 1870 and 1880, is the post-Civil War era of continued settlement and accommodation to the consequences of the end of slavery. The third period is the early twentieth century, encompassing 1900 and 1910. The consolidation of data also has a secondary advantage of enlarging the number of cases in individual cells, although this advantage is mediated by a partial lack of statistical independence.[9] The strategy for data presentation divides the older population into three age groups, 50 to 59, 60 to 69, and 70 and older. This data presentation strategy avoids the more common division at age 65, because in this population the break at age 70 is the most important. The divisions prior to age 70 are made to present evenly structured groups.

Families and Households in an Aging Population

While there was tremendous variety in the kinds of households in which the older population lived and the kinds of relationship older residents had


179

with the rest of their household, there are also overall patterns. Most older Texans were married, or had been married. Few had never married. For the majority who were or had been married, there was a single preferred living arrangement: they lived in nuclear households with one or more of their unmarried children. Even after their children had grown and left the home, many couples continued to live by themselves. This preference for nuclear family arrangements, consisting of a married couple or a surviving spouse and unmarried child, is unexceptional. For a number of older people, however, the nuclear family living pattern was not the only one they experienced in their later years. A large proportion of all older people spent at least some of their life in a more complex family arrangement, one in which they shared living accommodations with a married child. The older woman or man lived with a married child only after her or his spouse had died; very few married couples shared a dwelling with one of their married children and his or her spouse. Not every widowed older person had married children with whom to live, however. A significant share of older people lived away from their immediate family, either with other kin or with nonkin.

We can recast the preceding generalization in the form of a description of the stages in the residential lifetime of a married couple. While there are a number of possible scenarios, an examination of the lifetimes of many individuals in the six Texas counties reveals that most older people passed through only two or three of the stages that will be described here. In the first stage, a young couple set up a household on its own at the time of marriage, or shortly thereafter. Most, though not all, of these young couples soon had children. The husband and wife lived in this household with their unmarried children, so long as their children remained unmarried. As each child married, he or she left to set up a separate household. This stage was continued, under different conditions, after the death of husband or wife. When one spouse died, the surviving spouse continued as household head, in company with unmarried children. In Texas, this stage of household development could be a lengthy one for a widowed woman because her odds of remarrying were slight. For a man, who was much more likely to remarry, this stage often did not last long, and he returned to the beginning of the development process.

The second category in the development of family residence patterns began when the last child left the household. At that time the older couple or the surviving spouse was left alone. This "empty nest" stage was a common and probably popular alternative to other living arrangements, especially among the German immigrants in our sample who reached old age around the turn of the century. Household after household in Gillespie County in 1910 consisted of an older couple (often in their 70s) living by themselves. There was, nevertheless, an alternative to the "no children" living arrangement. The third category in the development of family residence patterns


180

began when the couple or surviving spouse joined a married child to form a complex two- or three-generational household. This was often the final stage in the process, because it was usually only the onset of very advanced age that provoked the sharing of a residence in this way. Widowed women spent more time residing with their married children than widowed men, and both were more likely to live with married children than were older married couples.

Most older individuals in the six Texas counties passed through only two stages of residential life. In the first stage, they spent the majority of their married (and widowed) lives sharing a residence with their unmarried children. After their children left home, the parents followed only one of two alternatives. Either they lived by themselves or they shared a residence with a married child. Among the individuals captured by our research, at least, it was not common for older people to pass from living with unmarried children to living by themselves without children and then to live in their extreme old age with a married child. Cases of that sort certainly exist. William Herrington of Angelina County lived with his wife and unmarried children in 1850, when he was 50 years old. In 1860, he was still married, living with his wife but independent of his children. By 1870, at age 70, he was widowed and living with a married son. Herrington was probably dead by 1880, because we can no longer locate him in the census. The case of William Herrington is unusual, however. Far more representative is the case of Peter Alberthal of Gillespie County. Alberthal was enumerated in the census as late as 1880 (when he was 49 years old) as living with his wife and unmarried children. In 1900 and 1910 (when he is 79 years old), he is still married and living only with his wife.[10] Equally representative is Carl Barsch of Gillespie County, who was also living with his wife and unmarried children in 1880 (at age 55) but who was widowed and living with a married son in 1900 at age 75. Barsch was not enumerated in 1910 and presumably had died.

Two other alternatives were possible for the married or widowed older population. Some lived in a fourth kind of residential arrangement, with a combination of kin or with kin other than their married or unmarried children; others lived in a fifth category, with families to whom they were not related. Those who lived with nonkin were mostly servants, lodgers, boardinghouse keepers, and older couples in empty nest households in which there were resident servants. These options—living with other kin or with nonkin—were more frequently taken by widowed than married individuals. Men were more likely to live with nonkin, while women were more likely to live with kin other than their children.

The five stages or categories described here were neither obligatory nor ordinal. It was not necessary for every individual or couple to pass through all five stages, and not everyone passed through the five (or fewer) stages in


181

the order given. It was possible to skip a step, or even to double back (as in the case of a woman or man who remarried). It should be added that these categories owe a great deal to the categories of analysis for the study of households for gerontological studies devised by Ethel Shanas and her colleagues (1968).[11] I should add that for this analysis I have decided to forgo the frequently used categories published by Peter Laslett (1972). Those categories, which emphasize the structure of family households, are very useful and could be used to add illumination to this work. But they are less enlightening for this purpose than categories that emphasize the situation in which the older person is living and the people with whom that person lives.

The foregoing discussion of stages in the residential experiences of older Texans has left aside the population that never married. Not many older people in Texas were single. Fewer than 6 percent of older men and probably no more than 3 percent of older women had never married.[12] The small number of older people who never married makes their situation difficult to study in detail, but the differences between men and women are interesting. Unmarried men were more likely than women to live alone or with nonkin, while unmarried women were more likely than unmarried men to live with kin. Both unmarried men and unmarried women were more likely to live with kin in the early twentieth century than they had been in the years after the Civil War. This suggests that over time unmarried older people had more kin with whom to live as the population of Texas grew and matured.

The characteristic residential patterns just described took place in the context of an American population that has steadily aged over the last two centuries. Like the United States as a whole, the population of Texas has aged, although the older population has been smaller in Texas—as a proportion of the whole population—at every census. It is not necessarily easy to date the beginning of old age, and it is therefore difficult to choose an age at which to establish its beginning. Using age 65, a common time at which the beginning of old age is measured, table 6.1 presents results about the size of the older population in Texas and the United States, using both published census data and data from the six sample counties. It is worth repeating here that the data reported in this and other tables have been weighted to correct the overrepresentation of European and Mexican immigrants—and their children—in the data samples.[13] The population of the six counties was slightly older than that reported in published enumerations for the state as a whole.

The proportion of people aged 65 and older in Texas was substantially less than that in the United States as a whole in the mid-nineteenth century. Since then, the gap has partly closed, especially between 1870 and 1910. In 1870, the percentage of people aged 65 and over was nearly twice as large in the United States as a whole as it was in Texas. By 1910, it was only half again


182

TABLE 6.1
Estimates of the Older Population, 1850-1970

 

Population Age 65 +


Total Population (Six-County Sample)


Total Number (Six-County Sample)

Sex Ratio: Males per 100 Females (Six-County Sample)

 

Six-County Sample (%)



Texas (%)


United States (%)

 

Age 50+

Age 65+

Age 50+

Age 65+

1850

0.9

5,721

305

51

178

123

1860

1.0

13,526

815

131

141

166

1870

1.6

1.5

2.9

20,799

1,360

336

126

137

1880

1.8

1.7

3.4

37,443

2,644

664

123

126

1890

2.1

3.8

1900

2.6

2.4

4.0

31,671

2,835

817

122

132

1910

3.0

2.8

4.3

34,962

3,560

1,044

122

106

1930

4.0

5.4

1950

6.7

8.1

1970

8.9

9.9

SOURCES : Texas Historical Demography Project Database; Skrabanek, Upham, and Dickerson 1975: 7; Achenbaum 1978: 60.


183

as large. The population of Texas had aged disproportionately in the nineteenth and early twentieth century for a variety of reasons. A young population entered the state during its time of great expansion, and that population had aged, producing by 1900 its first sizable older population. But another process was at work as well. The population was also aging because of the decline of mortality. Over time, the population of Texas lived longer; this too is especially noticeable by 1900. One important implication of the lengthening of the lives of older Texans after 1900 is the delay in the arrival of widowhood. While this did not occur among all groups (it is most characteristic of German immigrants), for those it affected the lengthening of lives contributed to a modification of living arrangements by permitting surviving couples to live by themselves for a longer time after their children had grown.

The population of Texas had a smaller proportion of older inhabitants than the United States as a whole in the nineteenth and early twentieth century for a variety of reasons, but the most important was its proximity to the frontier and its relatively recent settlement. Immigrants, whether from other American states or from other countries, were younger than a settled population would have been. Texas was Mexican territory until 1836, then an independent Republic of Texas from 1836 until 1845, when it became an American state. Until the 1820s, the Mexican government had discouraged the settlement of Americans in Texas. Only during the early 1820s did the Mexicans permit substantial immigration and colonization, and much of the state was relatively unsettled in the 1840s. The great wave of Texas settlement took place after Texas become independent of Mexico in 1836, especially after it became part of the United States in late 1845. The settlers were relatively young. The first generation of settlers, who arrived before the Civil War, aged during the second half of the nineteenth century; it is their swelling numbers that show in the larger proportions of older people after 1880. Before that time, it is clear from table 6.1, the older population was very small. In our six-county data sample, the numbers are striking. There are only 51 people aged 65 and older in 1850, only 87 in 1860, and only 336 in 1870. It will be very difficult to draw unequivocal conclusions from these few individuals, especially in 1850 and 1860. Table 6.1 shows that taking the population aged 50 and older will give us substantially larger numbers, but these too are relatively small.

In addition to being small, the older population was distorted in at least one other way. There were more men than women. Table 6.1 also reports the sex ratio for the population aged 50 and older and for the population aged 65 and over for the six-county sample. The normal sex ratio for a total population is about 105 men for every 100 women, and the ratio tends to decline as the population ages, because of relatively higher male mortal-


184

ity. Yet the Texas population, because of its history of frontier immigration, continued to have a very high proportion of men.

Living Arrangements of the Older Population: Men and Women, Married and Widowed, Older and Oldest

More older people in Texas lived with unmarried children than in any other residential arrangement. Life with unmarried children was thus the most common of the several distinct patterns and stages in which older people lived. Table 6.2 presents a breakdown of the older population in the six Texas counties into the categories of residential arrangements discussed earlier. Depending on the time period, at least two of three married people aged 50 and older and at least three of ten widows of the same age lived with their unmarried children. The remaining married older people were mostly living by themselves or with married children, with a few living with other kin and with nonkin. Widows were very likely to be living with married children or other kin and less likely to be living alone or with nonkin. Table 6.2

TABLE 6.2
Residential Living Arrangements of the Population Aged 50 and Older, Six Texas Counties

 

Married

Widowed

Single

 

Women

Men

Women

Men

Women

Men

1850-1860

           
 

Number of cases

254

481

192

171

26

19

 

Unmarried children

78%

84%

32%

36%

 

No children

10

6

4

8

0%

39%

 

Married children

6

4

28

23

 

Other kin

5

3

24

12

57

5

 

Nonkin

1

2

11

21

43

55

1870-1880

           
 

Number of cases

846

1,694

901

408

39

108

 

Unmarried children

68%

78%

29%

37%

1%

 

No children

14

9

7

10

5

23%

 

Married children

8

6

33

23

10

 

Other kin

4

3

16

6

40

21

 

Nonkin

6

5

15

24

44

56

1900-1910

           
 

Number of cases

1,617

2,764

1,182

408

80

190

 

Unmarried children

64%

72%

30%

38%

1%

 

No children

18

13

9

13

11

31%

 

Married children

10

7

38

24

3

 

Other kin

6

4

19

11

60

30

 

Nonkin

3

3

5

14

25

39


185

gives us a sense of the important differences between men and women and between married and widowed people.

Over the course of time, the proportion of older married individuals living with their unmarried children declined. This is one among several clear temporal themes in the results reported in table 6.2. Three other categories of residential arrangement compensate for the decline in individuals living with their unmarried children: there is substantial growth in the proportion living without children present and some growth in those living with married children (especially women) and with other kin. There are three possible explanations for this change. It is possible that the taste for parents living with unmarried children changed, either because parents no longer wanted to live with their children or because the children no longer wanted to live with their parents. It is also possible that the supply of children diminished, because of lower fertility, or the age structure of the older population changed, producing relatively more parents in the very old category (70 and older), all of whose children had married. Neither of the two demographic explanations put forward here is likely to have been the case. The older population in 1910 and earlier in Texas had experienced very little fertility decline (Fliess, Gutmann, and Vetter 1990), and the decline in older men and women living with unmarried children is characteristic of all age groups within the over-50 population. We are left with a potential change in taste and in the overall structure of communities.

In addition to changes through time, the experiences of older women were different from those of older men, and the experiences of widowed individuals were different from those of married individuals. Table 6.2 reveals that men were more likely to live with unmarried children than women, all other things being equal. This curious finding, produced by the large average difference in ages between married men and women, will be discussed in greater detail later. Women, in contrast, were more likely than men to live with married children and other kin. This is especially so for widowed women. Widowed men were more likely to live with nonkin. This last characteristic is a consequence of the quite different social and economic roles available to older men and women in nineteenth- and early-twentieth-century Texas. Older men living with nonkin were agricultural laborers; older women were more likely to be living with their families. While both older widowed women and men could work as servants, it appears that more work was available in agriculture than domestic service and that positions as domestic servants went by preference to younger women.

When we subdivide our population by age, we see another dramatic finding: a relatively high proportion of very old women and men—those aged 70 and higher—were still living with unmarried children. Table 6.3 presents the full matrix of the distribution of individuals between the categories of residential arrangements described earlier, by breaking up the population


186

TABLE 6.3
Percentage of the Older Population Living in Various Residential Arrangements

 

Married Women

Married Men

Widowed Women

Widowed Men

 

50-59

60-69

70+

50-59

60-69

70+

50-59

60-69

70+

50-59

60-69

70+

1850-1860

                       
 

Unmarried children

86%

56%

0%

90%

7.5%

62%

44%

28%

7%

50%

28%

11%

 

No children

8

13

44

3

14

10

3

4

10

11

7

2

 

Married children

3

16

0

3

6

13

23

29

41

11

30

47

 

Other kin

l

15

56

1

5

15

22

19

38

5

17

25

 

Nonkin

2

0

0

2

1

0

9

20

4

24

18

15

1870-1880

                       
 

Unmarried children

77%

52%

33%

84%

71%

60%

41%

21%

15%

49%

36%

20%

 

No children

8

23

37

6

13

19

4

8

11

10

9

11

 

Married children

7

12

9

5

7

10

26

41

36

13

26

34

 

Other kin

3

7

9

2

3

5

14

15

24

5

5

10

 

Nonkin

5

6

12

3

7

6

16

15

15

22

24

25

1900-1910

                       
 

Unmarried children

75%

47%

22%

81%

68%

41%

48%

28%

41%

54%

42%

18%

 

No children

10

29

43

8

15

32

6

15

32

11

12

17

 

Married children

7

15

16

6

9

12

28

9

12

14

19

40

 

Other kin

4

7

14

3

4

10

14

4

10

12

9

12

 

Nonkin

3

2

5

3

3

5

5

3

5

10

18

12


187

aged 50 and older into three subgroups: those aged 50 to 59, those aged 60 to 69, and those aged 70 and older. Table 6.3 is limited to the married and widowed population, both because it is already sufficiently complex without adding single people and because the number of single individuals in any cell is quite small. The detailed information in the table shows that more than 60 percent of men 70 and older were living with unmarried children prior to 1900, and 41 percent were living with unmarried children in the 1900-1910 period. The percentages for women are lower (and zero in the very small population aged 70 and over before the Civil War) but still remarkable.

Men were more likely than women to be living with unmarried children at advanced ages because the older Texas population was characterized by large differences in ages between husbands and wives. Men aged 50 and older were, on average, more than seven years older than their wives, and the difference in ages between husbands and wives increased with age. Men aged 70 and older were on average at least eleven years older than their wives. With differences in ages between husbands and wives so large, a man in his 70s may well have had a wife in her 60s. Moreover, this older generation was one that bore virtually all of its children before any transition in fertility took place. Women bore children until after their fortieth birthday. A married woman in her 60s was likely to have unmarried children at home, and so was her 70-year-old husband.

As the proportions of the married and widowed older population living with their unmarried children declined with age and over time, older people increasingly found themselves in other residential arrangements. In the series of stages identified earlier, the second stage was the empty nest, in which the couple or a surviving spouse lived by themselves without children. The percentage of the older population living in this category is much smaller than that living with unmarried children, but we should not underestimate its importance. As married men and women reached their 60s and 70s, they were increasingly likely to live by themselves, without their children. The increase in numbers living by themselves does not compensate for the decline in the number living with unmarried children, but it contributes a lot. Already in the early twentieth century, many couples and a considerable number of widowed women and men were finding themselves in circumstances in which they lived by themselves, without children, other kin, or nonkin.

Understanding the Role of Ethnicity, Economic Sector, and Occupational Status: a Multivariate Approach

Thus far the analysis reported here has emphasized the important role played by age, sex, and marital status in determining the ways in which members of the older population lived. While these demographic characteristics


188

are significant, other characteristics of the older population and the households they lived in are also worth considering. Any further information added to the analysis reported here will require a different approach from that already taken. If we need to look—as we do—at the role of ethnicity in addition to age, sex, and marital status, the tables would be so complex and the cell sizes so small as to render them impossible to interpret. We use instead a multivariate approach that permits conclusions to be drawn with more independent variables and smaller cell sizes.

In the special context of Texas and the data on which this analysis is based, evidence about ethnicity and occupation should be vitally important in determining the arrangements in which the older population lived. Ethnicity is especially important because each of the four most important ethnic groups—those of African, Mexican, and European origin and the native-born children of native parents—was likely to participate differently in the economy and to live in different kinds of communities. The recent settlement of the state meant that all the members of these groups were descended fairly recently from migrants and might be seen to shun strong commitments to the community. That, of course, was not the case. Immigrants to Texas from other states were often likely to move on again once they arrived, and much of the Mexican-American population worked in unstable jobs, but the African-American and German-American populations more often built stable communities.[14] This may have affected the kinds of households in which they lived, although the outcome that I will report may run counter to our intuition about the relationship of community to the residential experiences of the older population.

There is not much previous work on ethnic differences in household arrangements among the older population in the rural South on which to base this analysis. Daniel Scott Smith, Michel Dahlin, and Mark Friedberger (1979) examined the residential patterns of older southern blacks and whites in 1880 and 1900. These authors show the complexity involved in studying black-white differences; our four- or five-category division of ethnicity is still more complex. Nevertheless, their conclusions are worth comment. Smith and his colleagues differentiate between two kinds of motivation for parental co-residence with children, that based on economic production and inheritance and that based on family welfare. They show that for both blacks and whites, family welfare was more important than the alternative in producing co-residence. What this means is that few older people lived with their children so that they could share the responsibility for running a farm or other enterprise or so that they could provide inheritance to the younger generation. Rather, older people lived with their close kin to help support the otherwise needy, whether it was the older or the younger generation that was in need. Smith, Dahlin, and Friedberger conclude that differences between whites and blacks existed in 1880 and 1900—


189

blacks were more likely than whites to provide welfare in a downward direction to children and grandchildren, for example—but that these differences were largely a function of black poverty and were less significant than the enormous change that has taken place since 1900.

Despite strong kin ties within certain groups, and despite the existence of many large and complex households, there is no evidence that any sizable ethnic or economic group in Texas chose to live in multigeneration households. Married parents lived with married children, out of choice and for reasons of economic productivity, but such cases were relatively unusual. Put another way, while there appears to have been a strong preference for co-residence with unmarried children, there was no "family economy" of co-residence in Texas which might have served to bring older parents and married children together. Rather, most co-residence of parents and married children occurred for what Smith, Dahlin, and Friedberger call "welfare" reasons: to provide for someone in need. If this was indeed the case, we should find that most co-residence took place among those who had the most need and among those who had the greatest resources. We can get an idea of propensities to needs and resources in different ethnic groups by looking at overall occupational patterns within different groups.

The principal ethnic groups in Texas each had a distinct pattern of economic participation. Table 6.4 presents evidence for 1860, 1870, and 1900 on the economic activities of the households in which our sample population lived. The data represent the economic and ethnic characteristics of household heads, but the numbers represent the total population, because each individual in the population is given the ethnic and occupational characteristics of the head.[15]

A majority of each ethnic population (except the Mexicans) lived in a household in which the head held a farm occupation, but the internal structure of those farm occupations differed from ethnic group to ethnic group. The European origin population largely consisted of farm-owning households. While the native white population had a majority of farm-owning households, it also had a very significant number of farm tenants.[16] The African-American population was at the opposite end of the farmer-tenant continuum from the Europeans. In both 1870 and 1900, the majority were farm tenants, while a minority were farmers or farm laborers. Among farming Mexicans the distinctive feature is the large proportion of farm laborers. In these generally rural communities, there was nonetheless a significant part of the population that worked outside of agriculture. The differences between ethnic groups are predictable. African-Americans and Mexican-Americans were most likely to hold unskilled nonfarm occupations, while the Europeans and native whites were the most likely to be involved in white-collar work. In all groups there were a significant number of


190

TABLE 6.4
Occupation Distribution of Households by Ethnicity

 

African

European

 

N

%

N

%

1860

       
 

Farm

       
   

Farmer

720

63

   

Tenant

28

2

   

Laborer

23

2

 

Nonfarm

       
   

White collar

76

7

   

Skilled

183

16

   

Unskilled

77

7

 

No occupation

42

4

Total

1,149

1870

       
 

Farm

       
   

Farmer

202

4

848

54

   

Tenant

3,629

65

56

4

   

Laborer

1,017

18

10

1

 

Nonfarm

       
   

White collar

8

0

173

11

   

Skilled

42

1

272

17

   

Unskilled

289

5

127

8

 

No occupation

416

7

76

5

Total

5,603

1,562

1900

       
 

Farm

       
   

Farmer

1,232

19

2,552

69

   

Tenant

3,041

48

250

7

   

Laborer

923

15

67

2

 

Nonfarm

       
   

White collar

146

2

343

9

   

Skilled

128

2

249

7

   

Unskilled

522

8

46

1

 

No occupation

348

5

205

6

Total

6,339

3,712

(Table continued on next page)


191

(Table continued from previous page)

     

Mexican

Native

Other

     

N

%

N

%

N

%

1860

 
 

Farm

 
   

Farmer

40

8

4,566

36

239

60

   

Tenant

0

0

2,297

18

14

4

   

Laborer

42

8

2,487

20

3

1

 

Nonfarm

 
   

White collar

21

4

1,037

8

51

13

   

Skilled

98

20

739

6

28

7

   

Unskilled

271

54

607

5

23

6

 

No occupation

32

6

949

7

39

10

Total

504

12,682

397

1870

 
 

Farm

 
   

Farmer

21

3

4,520

44

281

64

   

Tenant

20

3

2,011

20

26

6

   

Laborer

224

30

443

4

0

0

 

Nonfarm

 
   

White collar

26

4

1,133

11

53

12

   

Skilled

86

12

544

5

47

11

   

Unskilled

220

30

526

5

16

4

 

No occupation

144

19

1,125

11

19

4

Total

742

10,303

441

1900

           
 

Farm

           
   

Farmer

80

6

7,190

42

432

29

   

Tenant

109

8

5,460

32

245

16

   

Laborer

482

35

754

4

135

9

 

Nonfarm

           
   

White collar

68

5

1,635

10

238

16

   

Skilled

172

13

752

4

178

12

   

Unskilled

301

22

582

3

133

9

 

No occupation

155

11

830

5

142

9

Total

1,367

17,204

1,503


192

households in which the head held no occupation; this was especially the case among Mexicans.

It is best to concentrate our attention first on the farming part of each ethnic population. Farm owners were more likely than farm tenants or laborers to have had the resources required for a multigenerational household. Farm tenants may have been constrained by tenancy contracts to limit the number of residents on the farm, so as to ensure its profitability (Ransom and Sutch 1977). On this basis we might assert that the German-origin population, which consisted almost totally of farm owners, had the greatest resources with which to support co-residence of parents with married children but probably also the least need. African-Americans and Mexican-Americans, however, had the fewest resources combined with the greatest potential need. Native whites were somewhere between these other groups, but their majority of farm owners within the farm population leads us to place them closer to the European origin population than to the Mexican-American and African-American populations. Turning to the nonfarm population, it was less likely to support co-residence than the farming population, a combination of overall wealth and space in which co-residing parents and married children could live. The ethnic distribution of nonfarm occupations is distinctive, confirming ethnic status in nineteenth- and early-twentieth-century Texas: Mexican-Americans and African-Americans in lower status occupations, and native whites and European origin whites in the higher status occupations.

The goal of the multivariate analysis of the determinants of domestic arrangements for the older population is to bring together in a single model a number of independent variables. From a variety of independent variables we should be able to see the range of influences that cannot be seen using contingency tables and also to be more sensitive to the interaction of variables so that we see genuine effects. The dependent variables are taken from the method of categorizing the family arrangements of the older population, so that we can see the determinants of living with unmarried children, for example, or the determinants of living without children present. Given the categorical dependent variables, I have chosen a logit regression approach.[17] I report this approach with three different dependent variables: whether or not the older person is living with unmarried children (table 6.5); whether he or she is living without children or others present (table 6.6); and whether he or she is living with one or more married child (table 6.7).

The models are estimated from a database of 12,549 men and women aged 50 and older in one or more of the six censuses we have transcribed for the period from 1850 to 1910.[18] The independent variables place the occurrence in a time period (1850-1860, 1870-1880, or 1900-1910), categorize the age of the older person (50-59, 60-69, or 70+), and measure sex,


193

TABLE 6.5
Logit Model: Living with Unmarried Children as Dependent Variable

Variable

Coefficient

Odds Ratio

Time perioda

 

Period 1 (1850-1860)

0.2390

1.2700**

 

Period 2 (1870-1880)

0.1075

1.1135**

Age groupb

 

60-69

-0.8814

0.4142**

 

70+

-1.7543

0.1730**

Sex and marital statusc

 

Married females

-0.6056

0.5458**

 

Widowed males

-1.4405

0.2368**

 

Widowed females

-1.8108

0.1635**

 

Single females

-5.7685

0.0031**

 

Single males

-6.6135

0.0013**

Nonfarm occupation (head)

-0.1194

0.8874

Ethnicityd

 

African-American

-2.8035

0.0606**

 

East European

0.1288

1.1374

 

Mexican-American

-0.4726

0.6234**

NOTE : Log likelihood = - 6754; number of cases = 12,549; x2 (13) = 3607.10,
p < .0001.

a 1900-1910 is the reference period.

b 50-59 is the reference group.

c Married males are the reference group.

d Native American and Other are the reference groups.

* p < .01.

** p < .001.

marital status, the occupation of the head of the household, and the ethnic group of the older person. All of these variables are categorical. When there are more than two categories in a categorical variable used in logit regression, the variable must be converted to a group of variables, each of which has a value of zero or one. One of these variables is not used in the analysis and is considered the reference. The coefficients for the variables included in the equation then describe the situation relative to the eliminated category.

The most important set of categorical variables included in the analysis in this way combine sex and marital status. Rather than use one variable for sex (male vs. female) and several variables for the three categories of marital status (single, married, widowed/divorced), I combined sex and marital status into six categories and treated married males as the reference category. The other important categorical variable was ethnic group, from which I created four new variables, with individuals born in North America of North American parents, plus the residual "other" category, the reference groups.


194

TABLE 6.6
Logit Model: Living without Children as Dependent Variable

Variable

Coefficient

Odds Ratio

Time perioda

 

Period 1 (1850-1860)

-0.4802

0.6186**

 

Period 2 (1870-1880)

-0.2534

0.7762**

Age groupb

 

60-69

0.8223

2.2758**

 

70+

1.2157

3.3728**

Sex and marital statusc

 

Married females

0.5037

1.6548**

 

Widowed males

-0.2739

0.7604*

 

Widowed females

-0.8973

0.4077**

 

Single females

-0.5499

0.5770

 

Single males

1.1700

3.2219**

Nonfarm occupation (head)

0.9184

2.5054**

Ethnicityd

 

African-American

0.3104

1.3639**

 

East European

0.4465

1.5628**

 

Mexican-American

0.3386

1.4030*

NOTE : Log likelihood = -4167; number of cases = 12,549; X2 (13) = 834.58, p < .0001.

a 1900-1910 is the reference period.

b 50-59 is the reference group.

c Married males are the reference group.

d Native American and Other are the reference groups.

* p < .01.

** p < .001.

In all three of the logit models reported here, the overall results are significant. The tables report the overall results of the analysis, plus the coefficient for each variable included. In addition, I report the odds ratio, which is the coefficient exponentiated, because the exponentiated value of the coefficient can often be more easily understood than the coefficient itself. For categorical variables, the exponentiated coefficient is the ratio of the odds of a value in the stated group to that of the reference group. For example, the value of 8.1 for the exponentiated coefficient for widowed females means that widowed females were slightly more than eight times as likely to be living with married children as were married men (the reference group), all other things being equal. With the exception of the coefficients for single females in two of the three models, the basic independent variables reporting age, sex, and marital status were predictable and significant. Age group, for example, predicted a lower likelihood of living with unmarried children as people aged and a higher likelihood of living alone or with married children as people aged. The same set of conclusions can be extended


195

TABLE 6.7
Logit Model: Living with Married Child as Dependent Variable

Variable

Coefficient

Odds Ratio

Time perioda

 

Period 1 (1850-1860)

-0.2862

0.7511**

 

Period 2 (1870-1880)

-0.1388

0.8704

Age groupb

 

60-69

0.6121

1.8443**

 

70+

1.0247

2.7862**

Sex and marital statusc

 

Married females

0.4448

1.5601**

 

Widowed males

1.3535

3.8708**

 

Widowed females

2.0896

8.0813**

 

Single females

-0.1859

0.8303

 

Single males

-1.8019

0.1650**

Nonfarm occupation (head)

-0.7195

0.4870**

Ethnicityd

 

African-American

-0.3144

0.7303**

 

East European

-0.2323

0.7927**

 

Mexican-American

0.1869

1.2055

NOTE : Log likelihood = -4359; number of cases = 12,549; x2 (13) = 1572.16, p < .0001.

a 1900-1910 is the reference period.

b 50-59 is the reference group.

c Married males are the reference group.

d Native American and Other are the reference groups.

* p < .01.

** p < .001.

from the contingency tables presented earlier in this chapter to these logit models. Similarly, the conclusions we drew about changes over time are confirmed by the multivariate models reported here. As time passed, older people were less likely to live with unmarried children and more likely to live alone or with married children, all other things being equal.

The results about economic sector reported in tables 6.5, 6.6, and 6.7 show that older people who lived in households headed by someone in the nonfarm sector were slightly less likely to be living with unmarried children, rather less likely to be living with married children, and quite a bit more likely to be living alone or with a spouse than those who were living in a farm sector household. These results are consistent with what we should expect. A set of alternative models involving five occupational categories was also attempted, and the results of that analysis did not improve on the results based on a simple farm versus nonfarm categorization.[19]

The results for different ethnicities help us go beyond what was revealed by simpler approaches and allow us to speculate about possible cultural dif-


196

ferences in the approach to family life for older people. When compared with the reference group, European immigrants and their children were more likely to be living with unmarried children, more likely to be living without children, and less likely to be living with married children. African-Americans were less likely to be living with unmarried children, more likely to be living without children, and less likely to be living with married children. Finally, Mexican-Americans were less likely to be living with unmarried children, more likely to be living without children, and more likely to be living with married children, when compared with the reference group. We can put these conclusions in another way and then use these findings to go on to more general conclusions. In all this we need to remember that the results are expressed with reference to the native-born population of native parentage. Compared to that group, the European immigrants and their children (many of whom would have reached age 50 by 1910) were more "nuclear" in their residential preferences and practices. The Europeans were more likely to live with unmarried children and more likely to live alone or merely with their spouse than the reference group. By contrast, the Mexican-Americans were more "extended" in their residential preferences and practices, when possible. They were more likely to live with married children but also more likely to live alone or with a spouse. The African-Americans might be described as "disenfranchised." They were less likely to be living with children than any other group. Whether this is due to their inability to support larger households given their financial condition or whether this is a consequence of their greater propensity to live outside normal family situations, either as servants or lodgers, remains to be seen.

The ethnic differentials in the living arrangements of the older population lead to at least one speculation that is worth putting forward here. Research we have done about out-migration from these counties suggests that the European origin population was very likely not to leave, especially when compared to the other groups described here (Gutmann et al. 1990). White natives and African-Americans were in an intermediate group, white Mexican-Americans were most likely to have emigrated or died between any two censuses. One conclusion this suggests is that by the turn of the twentieth century, at the latest, Euro-Americans in Texas had a well-developed local set of kin contacts, especially married children. There was no need for their parents to co-reside with married children, because a significant group lived nearby which might offer greater support than a single co-residing family could.

Conclusion

Like the rest of the population in America and many other Western nations, a large majority of older Texans lived in nuclear family households with spouse and unmarried children. Rural Americans in the Southwest did not


197

participate—or wish to participate—in a family economy in which several generations shared economic resources and a residence. Despite that conclusion, a significant number of older Texans lived with married children, and a growing number over time lived alone or merely with a spouse and not with children or other kin. I have emphasized two kinds of transitions away from the normative life in which the older person lived with a spouse or unmarried children. For older couples who both survived until their last child had left the family household and for some widows, life alone was increasingly common by the turn of the twentieth century. For widows and a small proportion of surviving couples, life with a married child was possible, although it was not as common.

That older couples in small towns and the countryside increasingly lived away from unmarried or married children by the turn of the twentieth century is one of the principal conclusions of this chapter. In good part, their ability to live by themselves was a consequence of longer lives. Unlike older widows, who may not have been able to maintain even a small household on their own, older couples could live by themselves. The fact that as the frontier filled in they were more likely to have had a number of children nearby and thereby a standby welfare system probably also made it easier. Most important, however, was perhaps a growing sense even a century ago that the generations were better off if they did not share a residence. Although the analysis here is too dependent on evidence from censuses and too filled with unresolved complexities to be absolutely sure of its implications, it is nonetheless suggestive of the values of both the older population and their children. By the early twentieth century, conditions were beginning to be right for separate residence to work. If this is the case, then we see as early as 1900 the origins of the late-twentieth-century tendency for the elderly to live isolated lives.

The differences between ethnic groups in Texas add subtleties and complexity to our analysis of the changing ways in which the older population lived. There is little evidence that anyone in Texas lived in complex households as part of a household economy that combined generations to maximize economic productivity or guarantee inheritance.[20] There were differences between ethnicities, but they seem to have been the consequence of different economic conditions rather than of different value systems. Both poverty and prosperity led older couples not to live with their married children, and this was the case for the African-American and the European origin populations. The European—mostly German—population was kin oriented and prosperous. They could have supported co-residence if they had a strong propensity to do so. They did not, because they appear to have been little inclined to share a residence with their married children until widowhood and because they had the economic resources and the standby kin resources to avoid co-residence. The African-Americans, however, may have had more need but certainly had less re-


198

sources with which to support co-residence. The older population was more likely to live alone.

The Texas counties studied in this chapter are near or just behind the moving American frontier between 1850 and 1910, and we expected their proximity to the frontier to affect the living arrangements of their older inhabitants. Was that the case? The major impact of the frontier was in the overwhelmingly agricultural economic structure and in producing a youthful population that aged relatively rapidly when compared with the population of the United States as a whole. The agricultural characteristic of these rural counties certainly contributed to the large proportion of older and widowed women who lived with their married children or with other kin. This was offset for older men and couples, however, by the availability of land, so that younger and older generations did not need to share property.[21] The aging of the population probably contributed to some sense of increasing complexity of households, but the process is difficult to study because there may have been several things going on at once: aging, which created more need for complex households as well as a longer-lived population and the availability of more kin to live with over time, and the beginning of the tendency for the older generation to wish to live alone.

References

Achenbaum, W. Andrew. 1978. Old age in the new land: The American experience since 1970 . Baltimore: Johns Hopkins University Press.

Agresti, Alan. 1990. Categorical data analysis . New York: John Wiley and Sons.

Fischer, David Hackett. 1978. Growing old in America . Oxford: Oxford University Press.

Fliess, Kenneth H., Myron P. Gutmann, and John E. Vetter. 1990. "The creation of Mexican-American fertility patterns." Paper presented at the annual meeting of the Population Association of America, Montreal, Canada, May.

Gutmann, Myron P. 1991. "Denomination and fertility decline: The Catholics and Protestants of Gillespie County, Texas." Continuity and Change 5: 391-416.


201

Gutmann, Myron P., and Kenneth H. Fliess. 1989. How to study southern demography in the nineteenth century: Early lessons of the Texas Demography Project . Austin: Texas Population Research Center Papers, no. 11.11.

Gutmann, Myron P., Kenneth H. Fliess, Amy E. Holmes, Amy L. Fairchild, and Wendy Teas. 1989. "Keeping track of our treasures: Database management for historical research." Historical Methods 22: 128-143.

Gutmann, Myron P., and Amy E. Holmes. 1988. The Texas census data system . Madison: Wisc-Ware.

Gutmann, Myron P., John E. Vetter, Kenneth H. Fliess, and Gregory Joslyn. 1990. Staying put or moving on? Ethnicity and persistence in Texas from 1850 to 1910 . Austin: Texas Population Research Center Papers, no. 12.03.

Holmes, Amy E., and Myron P. Gutmann. 1989. "Four 'nos': A data entry system for class projects." The History Teacher 21: 439-467.

Houdek, John T., and Charles F. Heller, Jr. 1986. "Searching for nineteenth-century farm tenants: An evaluation of methods." Historical Methods 19: 55-61.

Laslett, Peter. 1972. Introduction. In Household and family in past time , ed. Peter Laslett, 1-90. Cambridge: Cambridge University Press.

Ransom, Roger L., and Richard Sutch. 1977. One kind of freedom: The economic consequences of emancipation . Cambridge: Cambridge University Press.

Riley, Matilda White. 1982. "Aging and social change." In Aging from birth to death , ed. Matilda White Riley, Ronald P. Abeles, and Michael S. Teitelbaum, 11-26. Boulder: Westview Press.

Ruggles, Stephen. 1987. Prolonged connections: The rise of the extended family in nineteenth-century England and America . Madison: University of Wisconsin Press.

Shanas, Ethel, Peter Townsend, Dorothy Wedderburn, Henning Friis, Poul Milhøj, and Jan Stehouwer. 1968. Old people in three industrial societies . New York: Atherton Press.

Skrabanek, R. L., W. Kennedy Upham, and Ben E. Dickerson. 1975. The older population of Texas . College Station: Texas Agricultural Experiment Station.

Smith, Daniel Scott. 1979. "Life course, norms, and the family system of older Americans in 1900." Journal of Family History 4: 285-298.

———. 1981. "Historical change in the household structure of the elderly in economically developed societies." In Aging: Stability and change in the family , ed. Robert W. Fogel, Elaine Hatfield, Sara Bikiesler, and Ethel Shanas, 91-114. New York: Academic Press.

———. 1984. "Modernization and the family structure of the elderly in the United States." Zeitschrift für Gerontologie 17: 13-17.

———. 1986. "Accounting for change in the families of the elderly in the United States, 1900-present." In Old age in a bureaucratic society: The elderly, the experts and the state in American history , ed. David Van Tassel and Peter N. Stearns, 87-109. New York: Greenwood Press.

Smith, Daniel Scott, Michel Dahlin, and Mark Friedberger. 1979. "The family structure of the older black population in the American South in 1880 and 1900." Sociolog and Social Research 63: 544-565.

Vetter, John. 1990. "The Texas Project letter sample." Working paper, Frontier Demography Project, Population Research Center, University of Texas at Austin.


202

Vetter, John E., Jesus R. Gonzalez, and Myron P. Gutmann. 1992. "Computer-assisted record linkage using a relational database system." History and Computing 4: 34-51.

Wall, Richard, with Jean Robin and Peter Laslett, eds. 1983. Family forms in historic Europe . Cambridge: Cambridge University Press.


203

Seven
A Home of One's Own: Aging and Home Ownership in the United States in the Late Nineteenth and Early Twentieth Century

Michael R. Haines and Allen C. Goodman

An essential aspect of aging is the provision for food, shelter, and general economic well-being later in the life course. In closer-knit and less mobile societies, great reliance can be placed on kinship ties and relations of reciprocity and exchange between kin and other community members. In more mobile and rapidly changing environments, the elderly must often be more self-reliant. Consequently, individuals and families accumulate wealth for the anticipated slowdown or cessation of work. This wealth can take a variety of forms, both real and financial, but a principal one has been a home.

Home ownership has been central to the hopes and aspirations of American families. For many Americans, a home was—and remains—the major source of real property holding, especially as the society became more urban and less agricultural. Before the creation of large-scale, comprehensive pension plans, homes were the principal repository of all wealth for older urban individuals and households. Even today the "savings of the elderly are primarily in the form of housing" (Wise 1989: 2). In addition to providing more secure housing services, owner-occupied homes could be a source of income from rentals, boarding, and lodging. This was often of considerable importance later in the life course, especially for widows. Over the life course, mortgages have provided a means to accumulate savings in real property. It is an example of a life course phenomenon in which decisions made early in life have a major impact in later years.

It has become ingrained in American culture that a "home of one's own" is part of the aspiration for the good life—the "American dream" (Morris and Winter 1978; Perrin 1977; Rossi 1980). Home ownership has acquired important symbolic value. As John Adams (1987: 18) notes,


204

The equity in owned housing represents the dominant financial asset of the typical household in America, where 64 percent own the houses they live in. Buying a house is usually the most important financial commitment that a family makes, and for many households—perhaps most—housing decisions are highly emotional and intensely personal. . . . Housing has multiple hidden meanings . . . status, position, power, and personal identity.

Further, Peter Rossi and Anne Shlay (1982: 30) have noted,

American preferences for homeownership and for the spatial segregation of homeowners from renters appears to be so general that they can be regarded as norms deeply embedded in American values. . . . Owning one's home is viewed widely as a measure of achievement, as part of the American dream.

Historically, home ownership has been "one of the basic elements of satisfactory middle class life" in the United States (Warner 1962: 157). It was of importance to both native-born and immigrant workers (Kirk and Kirk 1981). Possession of property, especially homes, seemed desirable as a stabilizing and conservative influence, reinforcing thrift, industriousness, occupational and geographic stability, good citizenship, and other virtues, as well as providing a sense of status and economic security (Kirk and Kirk 1981: 473-475; Tygiel 1979: 92-93).

One manifestation of these aspirations is that home ownership, or at least a fair chance at it, is likely to increase with age. Increasing age may serve as a proxy for increasing income and/or wealth. Allen Goodman (1990) discovers, however, that even when income, housing prices, and other sociodemographic variables are controlled, age still has a positive and significant impact on the probability, of home ownership. This impact may also indicate acceptance of and ability to meet the social norms that characterize home ownership.

One might ask whether this relationship occurred historically. What were the implications of the patterns of age and home ownership? It is the goal of this chapter to explore the issue for the United States in the period from the Civil War to the 1930s.

American society has unquestionably aged since the middle of the nineteenth century. In 1850, the proportion of the population aged 60 and over was 4.1 percent. This had risen to 6.4 percent in 1900 and 8.5 percent by 1930. In 1989, the proportion of those aged 60 and over had doubled again to 18.9 percent (U.S. Bureau of the Census 1975: Series A, 119-134; 1991: table 22). A combination of a positive relationship of home ownership and aging with a rising share of the elderly in the population has resulted in a greater share of housing wealth held by the older population. In 1900, about 17 percent of all nonfarm owner-occupied dwellings were owned by households with heads aged 65 and over. This had risen to over 26 percent in 1987 (table 7.2; U.S. Bureau of the Census 1991: table 1285). This has meanings for residential mobility and patterns of co-residence. Are the elderly more or less likely


205

to co-reside with children of their own or rent? Does ownership convey great financial independence and bargaining influence with family and society?

When the nation was predominantly agrarian, a home usually went along with ownership of the farmstead. By the time of the 1890 U.S. census (the first to ask direct questions about housing), almost 48 percent of all dwelling units were owner occupied (see table 7.1).[1] Predictably, the ownership rate was lower among nonfarm households (37 percent) than among farm households (66 percent). For an earlier period, Lee Soltow has estimated for 1850 that about 50 percent of all dwelling units were owner occupied. This had changed little by 1870, when about 51 percent were owned by their occupants. In addition, an ownership differential similar to that found in 1890 applied between farm (65 percent owner occupied) and nonfarm (38 percent owner occupied) households (Soltow 1975: table 2.5). Indeed, there seem to have been few changes in home ownership incidence over the latter half of the nineteenth century.

By 1970, the national proportion of home ownership had risen to about 63 percent, with 62 percent among nonfarm households and 80 percent among the relatively small number of farm households. Since that time, the overall ownership rates have remained roughly stable at about 63 to 64 percent (see table 7.1). Much of the increase in home ownership rates has, however, taken place since 1940. Given the precipitous fall in the share of farm households, the increase must be explained almost entirely by the change in home ownership among urban households.

Among urban workers, lower rates of home ownership have been observed in the past. Only about 18 percent of worker families in the U.S. Commissioner of Labor Survey of 1889-1890 were home owners (U.S. Commissioner of Labor 1890, 1891), as compared with about 37 percent for all nonfarm households in the 1890 census (table 7.1). By 1901, this had risen to 19 percent for a survey of 25,440 urban families in thirty-two states and the District of Columbia (U.S. Commissioner of Labor 1904). This rate has converged toward the national average as the United States has urbanized and as relative worker incomes have grown.

By international standards, the United States has had relatively high levels of owner occupancy. For example, in the 1889-1890 survey of worker households just mentioned, 1,735 European households (in Belgium, France, Germany, Great Britain, and Switzerland) were also sampled. In contrast to the American home ownership rate of 18 percent, that among the European workers was only about 7 percent (and only 2 percent in Great Britain and 5 percent in France). By the middle of the twentieth century when the nonfarm ownership rate was 61 percent for the United States in 1960, it was 50 percent for Belgium (1961), 33 percent for urban France (1962), 13 percent for urban Germany (1961), 26 percent for urban Sweden (1965), and 43 percent for urban England and Wales (1961) (United Nations 1973: table 203). The greater abundance of land in the United


206

TABLE 7.1
Home Ownership Rates: Total, Nonfarm, and Farm, United States, 1890-1987

 

Total

Non farm

Farm

Year

Unitsa

Owneda

%

Unitsa

Owneda

%

Unitsa

Owneda

%

1890

12,690

6,066

47.80

7,923

2,924

36.91

4,767

3,143

65.93

1900

15,429

7,205

46.70

9,780

3,567

36.47

5,649

3,638

64.40

1910

19,782

9,084

45.92

13,672

5,245

38.36

6,110

3,838

62.82

1920

23,811

10,867

45.64

17,229

7,041

40.87

6,581

3,826

58.14

1930

29,322

14,002

47.75

22,917

10,550

46.04

6,405

3,452

53.90

1940

34,855

15,196

43.60

27,748

11,413

41.13

7,107

3,783

53.23

1950

42,826

23,560

55.01

37,105

19,802

53.37

5,721

3,758

65.69

1960

53,024

32,796

61.85

49,458

30,164

60.99

3,566

2,633

73.84

1970

63,450

39,885

62.86

60,351

37,393

61.96

3,095

2,492

80.52

1980

80,390

51,795

64.43

           

1987

90,888

58,164

64.00

           

SOURCES : U.S. Bureau of the Census 1975: Series N, 238-245; 1991: table 1283.

d In thousands.


207

States played a role, but more recent settlement (allowing for property acquisition by a wide variety of the population) and cultural values were also important. During the late nineteenth and early twentieth century, a relatively high rate of asset accumulation was characteristic of the United States generally in contrast to industrializing European nations (Ransom and Sutch 1989). These phenomena granted the possibility of greater independence in attaining security in old age.

As noted, home ownership has constituted a significant part of asset acquisition over the life cycle. For example, in 1988, equity in owner-occupied homes was 43 percent of total household net worth in the United States (U.S. Bureau of the Census 1991: table 759). One group of economic models aimed at explaining savings uses life cycle accumulation behavior (e.g., Modigliani 1988). The models hypothesize that households and individuals in the middle of the life course save in the form of both financial (e.g., pensions) and real assets (e.g., homes) for expected retirement in the later years of the life course. In this connection several authors have recently noted that housing wealth contributes to the interesting phenomenon of continued positive accumulation among many elderly and retired persons, even in the form of housing (see Ai et al. 1990; Feinstein and McFadden 1989; Kotlikoff 1989: 78-79; Stahl 1989; Venti and Wise 1989, 1990).

Housing also constitutes an important component of consumer budgets, a factor of importance to the elderly on retirement incomes. In the later nineteenth century, housing and housing operations took up to between one-fifth and one-fourth of all consumer spending. This had risen to almost 30 percent by the 1980s (U.S. Bureau of the Census 1975: Chap. G; 1991: table 706). Even though housing costs tend to decline for the elderly (Stahl 1989), the issue of owning one's dwelling can bear significantly on the security of residence in old age.

Age is an explanatory variable for demographers, economists, and others using or testing life cycle models of saving and accumulation (Kotlikoff 1989; Modigliani 1988). It would seem that studies of the age pattern of home ownership would be more common. It appears that they are not, especially historically.[2] The aim of this chapter is to provide a preliminary historical overview of the relationship of aging and home ownership from the late nineteenth to the early twentieth century, a period of rapid growth in ownership, to suggest some means of describing these data more concisely, and to examine some of the implications for the later life course of the rising and changing incidence of home ownership over time.

Data and Sources

Table 7.2 provides a summary of the data and sources used here. These by no means exhaust potential sources. In particular, Soltow (1975) and oth-


208

TABLE 7.2
Age Patterns of Home Ownership in the United States, 1865-1930

 

New York Counties
Sample, 1865
a,d

U.S. Commissioner of
Labor Survey, 1889-1890
a


U.S. Census, 1900a,e

Age Group

Total

Owners

%

Total

Owners

%

Total

Owners

%

Below 25

59

4

6.78

291

12

4.12

658

54

8.21

25-29

196

42

21.43

1,010

102

10.10

1,374

205

14.92

30-34

240

52

21.67

1,192

181

15.18

1,787

413

23.11

35-39

263

81

30.80

1,155

194

16.80

1,869

573

30.66

40-44

237

87

36.71

1,021

201

19.69

1,708

604

35.36

45-49

252

100

39.68

839

183

21.81

1,352

553

40.90

50-54

180

82

45.56

582

144

24.74

1,225

547

44.65

55-59

160

59

36.88

356

88

24.72

887

451

50.85

60-64

135

51

37.78

221

65

29.41

744

409

54.97

65-69

78

31

39.74

82

17

20.73

544

330

60.66

70+

95

41

43.16

33

11

33.33

667

447

67.02

 

Total

1,895

630

33.25

6,782

1,198

17.66

12,815

4,586

35.79

(Table continued on next page)


209

(Table continued from previous page)

 

U.S. Census, 1890b

 

U.S. Census, 1930c

Age Group

Total

Owners

%

Age Group

Total

Owners

%

Below 25

412,708

55,644

13.48

Below 25

1,266,066

130,869

10.34

25-29

949,514

184,980

19.48

25-34

5,878,711

1,516,341

25.79

30-34

1,159,634

316,756

27.32

35-44

7,082,391

3,142,403

44.37

35-39

1,119,561

361,977

32.33

45-54

5,743,244

3,201,077

55.74

40-44

967,557

363,420

37.56

55-64

3,680,822

2,396,679

65.11

45-49

865,962

360,222

41.60

65-74

1,880,969

1,361,618

72.39

50-54

749,591

338,202

45.12

75+

561,223

424,288

75.60

55-59

536,246

269,172

50.20

       

60+

1,162,200

673,298

57.93

       
 

Total

7,922,973

2,923,671

36.90

Total

26,093,426

12,173,275

46.65

SOURCES : Seven New York Counties Sample, 1865 : Five percent sample of seven New York counties (Allegany, Dutchess, Montgomery, Rensselaer, Steuben, Tompkins, and Warren) from the manuscript of the 1865 New York State census. U.S. Commissioner of Labor Survey, 1889-1890 : U.S. Commissioner of Labor (1890, 1891 ). The sample consists of 6,809 households of workers in nine industries (bar iron, pig iron, steel, coke, bituminous coal, iron ore, cotton textiles, woolen textiles, and glass) in 24 states of the United States. U.S. Census, 1900 : Tabulations from the public use sample of the manuscripts of the 1900 U.S. census of 101,438 individuals. U.S. Census, 1890 : U.S. Bureau of the Census 1896: table 77. U.S. Census, 1930 : U.S. Bureau of the Census 1933: table 35.

a Male anti female heads of households.

b Males anti females.

c Male heads of household only.

d Nonfarmers only.

e All persons living in homes; farms excluded.


210

ers have worked with sample data on real and personal property holdings from the U.S. census manuscripts for 1850, 1860, and 1870. Roger Ransom and Richard Sutch (1989) and their colleagues are studying savings behavior in America since the late nineteenth century and are making use of the many state labor department surveys that provide data on wealth and home ownership. There also exist national public use samples of the federal censuses of 1910, 1940, and 1950 (as well as later) which provide opportunities to tabulate and analyze housing tenure status by household head's age and other characteristics.[3]

The sources utilized here were readily available to the authors as micro-data (i.e., the sample of upstate New York counties in 1865; the 1889-1890 U.S. Commissioner of Labor Survey; and the 1900 U.S. census public use sample) or were obtained from the published volumes of the 1890 and 1930 censuses. Although the federal census has systematically collected data explicitly on home ownership since 1890, relatively little has been published along the dimension of the age of the household head. The censuses of 1890 and 1930 were exceptions. (Note 1 discusses the issue of attribution of ownership and household headship.)

Home Ownership and the Life Course, 1865-1930

Some of the basic results from table 7.2 are reproduced in figure 7.1, which gives the age-ownership profiles for the sample of seven New York counties in 1865, the 1889-1890 U.S. Commissioner of Labor Survey, the 1900 census public use sample, and published data from the 1930 U.S. census.[4] It is important to note that these results apply to urban and rural nonfarm households. The 1865 New York data are tabulated only for heads of household who were not farmers. The actual question asked in the New York census was, however, whether the individual owned land, so the results (like those from the federal censuses of 1850-1870) are not strictly comparable to those for later dates when explicit questions were posed on renter or owner-occupancy status of the household. The 1889-1890 survey clearly applied only to industrial, mostly urban, working-class households. The 1900 U.S. census public use sample tabulations were done only for heads of household who owned or rented homes or dwellings and not farms.[5] The tabulations from published data for 1890 and 1930 excluded farm households.

There was, not unexpectedly, an upward shift in the age-ownership relationship among urban and rural nonfarm households in the United States from at least the late nineteenth century. The low level of home ownership in the sample of industrial workers in 1889-1890 is also evident, as compared to national census data for 1900 (or 1890) or even the New York data for 1865. This was partly due to the more urban residence of these workers. Results from the U.S. census of 1890 (fig. 7.5, below) demonstrate that ur-


211

figure

Fig. 7.1.
Home ownership, by age, 1865-1930.

ban areas had lower ownership rates than rural areas and the larger the urban area, the lower the ownership rate. The lower rate for workers was also due to lower incomes relative to the urban middle and upper classes.

Another notable feature in figure 7.1 is the contrast between the smooth upward progression of ownership by age in the national data for 1900 and 1930 (as well as for 1890, as seen in fig. 7.4, below) in comparison with the re-suits for New York State in 1865 and for the 1889-1890 survey data. The curves for the latter two data sets tend to flatten out, or even decline, at older ages. This may be seen more dramatically in figures 7.2 and 7.3. It is interesting to point out that similar shapes for the censuses of 1850, 1860, and 1870 were generated by Soltow (1975: 29, chart 2.1) when he plotted the proportion of free adult males having real estate or total real and personal estate.

It appears that the age pattern of property holding changed in the late nineteenth century from one that peaked in middle age to one that peaked late in the life course. It is a most interesting result. It seems that in New York State in the mid-nineteenth century (as well as in Soltow's national samples for 1850-1870) and among urban working-class families in the late nineteenth-century maximum, home ownership was achieved by about age 40 to 50. That is, there was no greater ownership late in the life course than in middle age. This had changed by the turn of the century for the population as a whole. With the consistent upward movement of the age-ownership


212

figure

Fig. 7.2.
Urban, rural, native-born, foreign-born, and total
 home ownership, by age, N.Y. counties sample, 1865.

figure

Fig. 7.3.
Native-born, foreign-born, and total home 
ownership, by age, U.S. labor survey, 1890.


213

profiles from the 1890, 1900, and 1930 censuses and with the upward shifts in those profiles, it is clear that individuals were able to continue acquiring homes (on a net basis) right into their 60s and 70s. Looking from census to census, it is also apparent that this was true for the same age cohorts over time as well as across age groups at a point in time. Age was proving to be no barrier to the achievement of this part of the American dream. Increasingly, wealth in the form of homes was characteristic of the later years of the life course.

More detailed information from the five data sets is plotted in figures 7.2 through 7.9, which provide a variety of dimensions of the age-ownership profiles. Several salient aspects appear. Urban ownership rates were lower than nonfarm rural rates.[6] This was true in 1865 New York (fig. 7.2) and in the country as a whole in 1890 (fig. 7.3), 1900 (fig. 7.6), and 1930 (fig. 7.8). Within the urban population, the results for 1890 (fig. 7.4) demonstrate that smaller cities (population between 50,000 and 250,000) had higher ownership rates than larger cities (with populations above 250,000). A plausible explanation is that the higher population densities of larger cities raised land values, which, in turn, increased housing prices, reduced ownership rates, and raised the profitability of building and maintaining multiple family rental properties. Current evidence for 1960, 1970, and 1980 indicates that ownership rates continue to be lower in (denser) central cities

figure

Fig. 7.4.
Urban home ownership, by age, U.S. census, 1890.


214

figure

Fig. 7.5.
Home ownership, by age and region, U.S. census, 1890.

figure

Fig. 7.6.
Urban, rural, and total home ownership, by age, U.S. census, 1900.


215

than the remainder of the standard metropolitan statistical areas (SMSAs) and higher outside of SMSAs than within (Adams 1987: 53, table 3.14).

Historically, there were also ownership differentials by nativity and race. Certainly differences in residence and income accounted for some of this. For example, many blacks had lower incomes but lived in rural areas where ownership rates were higher. Many white immigrants lived in large urban areas, most often in central cities where ownership rates were lower. But cultural and other factors may have played a role. Immigrant peasants often viewed property as a sign of social mobility but also as a means of reducing risk in an uncertain economic environment (Bodnar 1985: 180-183). Higher home ownership rates among working-class Irish immigrants have been seen as a consequence of land hunger carried from Europe, with children's earnings used to achieve this goal at the expense of their education (Thernstrom 1964: 154-157). The evidence on this is not clear (Bodnar 1985: 182), but there were a variety of factors, such as specific area of origin, duration of residence in the United States, region and place of residence, occupation, and income, that interacted with race and ethnicity to produce the observed patterns.

For the sample of New York counties in 1865, foreign-born whites had consistently lower ownership incidence than the native born (fig. 7.2). This was still true for the nation as a whole in 1900 (fig. 7.7), although the differences by age were much smaller. The nativity differential in home ownership rates for urban and rural nonfarm whites had virtually disappeared by 1930 (fig. 7.9), indicating that, on this dimension, immigrants were assimilating and eventually sharing in this promise of American life. In addition, this advantage was achieved within a generation. Tabulations from the 1900 public use sample and published data from the 1930 census show that when nativity of parents is considered, there were only small differences in home ownership between native whites with native parents and second-generation immigrants (native whites of foreign or mixed parentage).

Interestingly enough, the home ownership curves for native versus foreign born exhibited a crossover in the 1889-1890 labor survey (fig. 7.3), with the foreign born having had lower home ownership incidence up to age 30 and higher rates thereafter. A breakdown of the data for specific nativity of household head shows that this was due especially to German and Irish immigrants. It should be noted, however, that many migrants were more likely to live in regions of the country (such as the Midwest) where home ownership was more common and that they also had different incomes, occupations, and family compositions. Nevertheless, multivariate analysis of this data set has indicated that the ethnic differentials do not entirely disappear when differences in incomes, residence patterns, industries, occupations, ages, and family composition are taken into account. German migrants were significantly more likely to own a home than the native born,


216

figure

Fig. 7.7.
Home ownership, by age and race, U.S. census, 1900.

figure

Fig. 7.8.
Urban and rural nonfarm home ownership, by age, U.S. census, 1930.


217

figure

Fig. 7.9.
Urban home ownership, by age and race, U.S. census, 1930.

while Canadian and British migrants were less likely (Haines and Goodman 1992). The same analysis also revealed a strong and significant nonlinear relation of age to the probability of home ownership.

In contrast, differentials by race did not disappear over this period. Data from the 1900 (fig. 7.7) and 1930 (fig. 7.9) censuses point to systematically lower ownership rates for urban and rural nonfarm blacks relative to whites, both immigrant and native born. In 1900, the nonagricultural black population only attained an ultimate ownership rate of about 20 to 25 percent, and this was achieved by about age 40. Indeed, the age pattern resembled that of the white population in the middle of the nineteenth century rather than around 1900. It also looked a good deal more like that of the industrial workers in the 1889-1890 survey.

Perhaps this should also not be too surprising, since there existed a number of confounding elements. Average income, occupational attainment, and socioeconomic status of the black population was low relative to the native white population and most immigrant groups, and there were also specific barriers to property acquisition in many areas, limiting opportunities for ownership as well as creating residential segregation. For example, for 1880 in Philadelphia, the proportion of adult males listing occupations as unskilled was 78.4 percent among blacks, 48.7 percent among Irish immigrants, 16.6 percent among German migrants, and 20.4 percent among all


218

native whites (including second-generation migrants). By 1930, 56 percent of adult blacks were classified as laborers or in domestic and personal service, as opposed to 19.7 percent for foreign-born whites and 8.5 percent for native whites. Blacks also had the highest indexes of residential segregation in both 1880 and 1930 (Hershberg et al. 1981: 468, 471, 475).

The urban black age-ownership profile had become steeper and more regular by 1930 (fig. 7.9), but it still lay below that of the white population. While both native and foreign-born urban whites had attained 50 percent ownership rates by ages 45 to 54 and experienced increases for older age cohorts, urban blacks had barely achieved this by the last years of the life course.

Another dimension that was tabulated in the published data for 1890 was region of residence, some results for which are given in figure 7.5.[7] While it does not seem intuitively apparent that regional differences should exist, in 1890, the highest rates of nonfarm home ownership were found in the Midwest (the North Central Region) and the West, with the lowest in the South (the South Atlantic and South Central regions). New England and the Middle Atlantic states (the North Atlantic Region) were intermediate.

Differences between the regions in levels of urbanization and income account, in part, for this. For example, in terms of nonagricultural income per worker in 1900, the West was unquestionably the highest ($803), with the North Atlantic and North Central regions intermediate (at $630 and $650, respectively) and the South the lowest (with $223 and $225 in the South Atlantic and South Central regions) (calculated from Easterlin 1957). This would have promoted the high ownership rates in the West and North and the low rates in the South. However, the higher incidence of large cities in the Northeast and generally greater level of urbanization there relative to other regions would have depressed its ownership rates.[8] Regional differences have persisted, but the relative positions have changed. In 1983, for instance, the Midwest still had the highest incidence of owner occupancy (69.1 percent), followed by the South (67.4 percent), the Northeast (60.4 percent), and the West (59 percent) (U.S. Bureau of the Census 1985: table 1308).

One additional piece of information is available and was tabulated by age for the published census data in 1890 and the public use sample for 1900—the extent of mortgage indebtedness for owner-occupied homes.[9] Tabulations of the percentage of owner-occupied dwellings having a mortgage encumbrance reveal that the age pattern showed a generally downward incidence of mortgaged property with increasing age as individuals and families were able to attain full ownership over the life course. But the age group below 25 years had a lower incidence than the two or three next oldest groups (up to ages 35 or 40). This might have been due to inheritance, that is, a larger portion of the youngest home owners having obtained their property unencumbered via bequest. If such properties were inherited from parents or other older persons, the chance of their being unencumbered was greater. Home ownership thus played a role in intergenerational mobility.


219

Urban home owners were also more likely to have had mortgages than their rural nonfarm counterparts. This probably reflects higher urban site values as well as more developed mortgage capital markets in cities. In addition, there were substantial regional differences in mortgage incidence. A much higher proportion of homes were mortgaged in the Northeast and Midwest as compared with the South and the West. It is known that the South and the West had considerably higher interest rates, on average, especially relative to the Northeast in this period (Davis 1965; James 1978: chap. 1; Snowden 1987), which would have tended to reduce mortgage incidence, reduce site values, and/or discourage ownership. Finally, foreign-born whites were more likely than native whites to have mortgaged property, especially in the middle years of the life course. This likely reflects less inherited property and less wealth overall. Blacks seem to have had least access to the mortgage markets, although the number of cases is too small to permit reliable inferences.

Given the large amount of information about home ownership and age along a number of dimensions in the tables and figures developed here, efforts to condense and summarize these data are indicated. Some of these results are given in table 7.3. Several measures easily suggest themselves: ownership rates earlier and later in the life course (i.e., percent owners age 25-34 and percent owners at older ages—60 and over, 70 and over, 75 and over), as well as the overall ownership rate and an age-standardized ownership rate.[10] It turns out that age standardization does not greatly change things for the native white population or the overall rural or urban populations. The ownership rates for blacks and foreign-born whites are generally increased by standardization because of their somewhat younger age structure relative to the overall, predominantly native white, population.[11]

The last two columns of table 7.3 present elasticities of ownership with respect to age. These are simply the percentage changes in the probability of home ownership for a 1 percent change in age. They are evaluated at ages 30 and 60. They are calculated by fitting a statistical function (in this case, a logit) to the age-ownership profiles and then calculating the elasticities at these two points.[12] In general, it may be said that the propensity to acquire a home at these two moments in the life course increased over time, was higher in urban than in rural areas, was higher in larger cities and in the Northeast (in 1890), increased with age up until about 1900 and thereafter diminished with age, was generally higher among the foreign born relative to the native white population, and rose for the younger black population in the early twentieth century. It is evident that property acquisition was becoming more accessible for the younger population in the twentieth century as well as for the black population as a whole. The substantial appetite of younger immigrants for real property is also supported. The reduced likelihood of acquiring property at older ages by 1930 is interesting. It reflects


220

TABLE 7.3
Summary Measures of Home Ownership, United States, 1865-1930

 

Percent Owners

Elasticity of Ownership

 

Aged 25-34

Upper Ageda

Overall

Age Standardizedb

Age = 30

Age = 60c

N.Y. counties sample 1865

 

(70 + )

       
 

Total

21.56

43.16

33.25

32.46

0.552

0.882

 

Native born

25.00

46.25

37.90

37.04

0.504

0.789

 

Foreign born

14.86

28.57

23.30

22.48

0.488

0.836

 

Rural

30.19

50.00

39.19

38.58

0.329

0.563

 

Urban

16.61

37.25

29.39

28.74

0.701

1.096

U.S. labor survey, 1889-1890

 

(60+)

       
 

Total

12.85

27.68

17.66

19.22

0.815

1.375

 

Native born

13.57

26.40

15.94

17.84

0.635

1.146

 

Foreign born

11.63

28.44

19.72

20.10

0.902

1.430

U.S. census, 1890

 

(60+)

       
 

Total

23.79

57.93

36.90

37.63

0.839

1.051

 

Cities 50,000+

13.94

40.47

24.02

24.72

0.887

1.294

 

Cities 250,000+

11.13

35.82

20.32

20.97

0.938

1.417

 

Cities 50,000-250,000

18.03

47.13

29.34

30.11

0.843

1.164

 

North Atlantic

17.84

55.64

32.98

32.56

1.017

1.289

 

South Atlantic

18.31

42.60

26.89

28.37

0.755

1.095

 

North Central

31.08

69.30

46.34

47.26

0.837

0.904

 

South Central

20.46

45.34

29.22

31.42

0.740

1.055

 

West

34.27

61.40

44.05

45.39

0.562

0.710

(Table continued on next page)


221

(Table continued from previous page)

TABLE 7.3
Summary Measures of Home Ownership, United States, 1865-1930

 

Percent Owners

Elasticity of Ownership

 

Aged 25-34

Upper Ageda

Overall

Age Standardizedb

Age = 30

Age = 60c

U.S. census, 1900

 

(70+)

       
 

Total

19.55

67.32

35.80

36.16

1.157

1.226

 

Urban

16.75

65.19

33.63

33.62

1.228

1.484

 

Rural

26.21

71.29

41.47

43.00

1.018

1.063

 

Native-born white

21.77

70.19

38.06

39.17

1.158

1.273

   

Native-born parentage

16.62

68.85

39.48

39.27

1.124

1.250

   

Foreign-born parentage

12.54

39.22

33.76

38.70

1.301

1.385

 

Foreign-born white

22.45

70.72

37.07

34.53

1.276

1.452

 

Black

20.26

62.50

18.49

20.19

0.757

1.339

U.S. census, 1930

 

(75+)

       

Urban

           
 

Total

23.20

70.18

43.07

43.14

1.117

0.957

 

Native-born white

25.21

71.36

43.57

45.00

1.103

0.924

   

Native-born parentage

25.26

71.37

42.61

44.36

1.083

0.929

   

Foreign-born parentage

25.11

71.32

45.58

46.25

1.144

0.914

 

Foreign-born white

20.36

70.59

47.33

42.63

1.069

0.900

 

Black

11.87

52.94

24.67

27.76

1.419

1.580

Rural nonfarm

           
 

Total

31.98

79.79

50.85

50.37

1.005

0.597

 

Native-born white

33.37

80.36

51.33

51.52

0.981

0.670

   

Native-born parentage

32.28

79.82

49.40

50.09

0.986

0.703

   

Foreign-born parentage

39.25

83.34

59.97

57.76

0.912

0.532

 

Foreign-born white

36.41

85.31

62.98

55.04

0.875

0.518

 

Black

17.42

56.13

31.78

34.08

1.152

1.218

a Upper age group is given at the top of the column for section.

b Standardized to the age structure of all household heads in 1900.

c For 1890, elasticity at age 57.5.


222

the flattening out of the curves and an apparent upper limit to ownership incidence late in the life course.

Conclusion

Overall, examination of the relation between age and property acquisition—in this case, home ownership—seems a fruitful area for further historical research. The importance of home ownership (often unencumbered by a mortgage) to individuals and households late in the life course has become increasingly evident. Homes are a source of more secure shelter as well as a means of providing potential income (from rentals, boarding, and lodging) and collateral for borrowing. Home ownership likely conveys greater independence for the elderly.

For the American case, there were changes in both the shape and level of the age-ownership profile over time. The basic data also revealed differentials by nativity, race, rural-urban residence, city size, and region. Since roughly the middle of the nineteenth century, it has become more likely at all ages that a household would live in a home of its own but particularly later in the life course. This is reflected in the upward shift and the greater steepness of the age-ownership profile. Increasingly, immigrants and the black population began to participate in this process. This aspect of the "American dream" was becoming a reality for many but by no means all. As the society has aged and, given the positive relationship of ownership to age, a greater share of American homes are now owned by the elderly (over 26 percent in 1987, up from 17 percent of nonfarm households in 1900). This is clearly an important aspect of today's economy and society, but an understanding of its historical evolution is important as well.

References

Adams, John S. 1987. Housing America in the 1980s . New York: Russell Sage Foundation.

Ai, Chunrong, Jonathan Feinstein, Daniel McFadden, and Henry Pollakowski. 1990. "The dynamics of housing demand by the elderly: User cost effects." In Issues in the economics of aging , ed. David A. Wise, 33-82. Chicago: University of Chicago Press.


225

Bodnar, John. 1985. The transplanted: A history of immigrants in urban America . Bloomington: Indiana University Press.

Davis, Lance E. 1965. "The investment market, 1870-1914: The evolution of a national market." Journal of Economic History 25: 355-399.

Easterlin, Richard A. 1957. "State income estimates." In Population redistribution and economic growth: United States, 1870-1950 , eds. Simon Kuznets and Dorothy S. Thomas, I:703-759. Philadelphia: American Philosophical Society.

Feinstein, Jonathan, and Daniel McFadden. 1989. "The dynamics of housing demand by the elderly: Wealth, cash flow, and demographic effects." In The economics of aging , ed. David A. Wise, 55-86. Chicago: University of Chicago Press.

Goodman, Allen C. 1990. "Demographics of individual housing demand." Regional Science and Urban Economics 20: 83-102.

Haines, Michael R., and Allen C. Goodman. 1992. "Housing demand in the United States in the late nineteenth century: Evidence from the Commissioner of Labor survey." Journal of Urban Economics 31: 99-122.

Hajnal, John. 1953. "Age at marriage and proportions marrying." Population Studies 7(3): 111-136.

Hershberg, Theodore, Alan N. Burstein, Eugene P. Ericksen, Stephanie W. Greenberg, and William L. Yancey. 1981. "A tale of three cities: Blacks, immigrants, and opportunity in Philadelphia, 1850-1880, 1930, 1970." In Philadelphia: Work, space, family, and group experience in the nineteenth century , ed. Theodore Hershberg, 461-491. New York: Oxford University Press.

James, John A. 1978. Money and capital markets in postbellum America . Princeton: Princeton University Press.

Kirk, Carolyn T., and Gordon W. Kirk, Jr. 1981. "The impact of the city on home ownership: A comparison of immigrants and native whites at the turn of the century." Journal of Urban History 7: 471-498.

Kotlikoff, Laurence J. 1989. What determines savings? Cambridge: MIT Press.

Maddala, G. S. 1983. Limited-dependent and qualitative variables in econometrics . New York: Cambridge University Press.

Modigliani, Franco. 1988. "The role of intergenerational transfers and life cycle savings in the accumulation of wealth." Journal of Economic Perspectives 2(2): 15-40.

Morris, E. W., and M. Winter. 1978. Housing, family, and society . New York: John Wiley and Sons.

Perrin, C. 1977. Everything in its place: Social order and land use in America . Princeton: Princeton University Press.

Ransom, Roger L., and Richard Sutch. 1989. "Two strategies for a more secure old age: Life-cycle saving by late-nineteenth-century American workers." Paper presented at the Summer Workshop of the Development of the American Economy Project, National Bureau of Economic Research, Cambridge, Massachusetts, July 17-21.

Rossi, Peter H. 1980. Why families move . 2d ed. Beverly Hills, Calif.: Sage.

Rossi, Peter H., and Anne B. Shlay. 1982. "Residential mobility and public policy issues: 'Why families move' revisited." Journal of Social Issues 38(3): 21-34.

Snowden, Kenneth A. 1987. "Mortgage rates and American capital market development in the late nineteenth century." Journal of Economic History 47(3): 671-691.


226

Soltow, Lee. 1975. Men and wealth in the United States 1850-1870 . New Haven: Yale University Press.

Stahl, Konrad. 1989. "Housing patterns and mobility of the aged: The United States and West Germany." In The economics of aging , ed. David A. Wise, 93-115. Chicago: University of Chicago Press.

Thernstrom, Stephen. 1964. Poverty and progress: Social mobility in a nineteenth-century city . Cambridge: Harvard University Press.

Tygiel, Jules. 1979. "Housing in late nineteenth-century American cities: Suggestions for research." Historical Methods 12(2): 84-97.

United Nations. 1973. Statistical yearbook, 1972 . New York: United Nations.

U.S. Bureau of the Census. 1896. Eleventh census of population, 1890: Report on farms and homes: Proprietorship and indebtedness . Washington, D.C.: Government Printing Office.

———. 1933. Fifteenth census of population, 1930: Population . Vol. VI. Families . Washington, D.C.: Government Printing Office.

———. 1975. Historical statistics of the United States: Colonial times to 1970 . 2 vols. Washington, D.C.: Government Printing Office.

———. 1979. Twenty censuses: Population and housing questions, 1790-1980 . Washington, D.C.: Government Printing Office.

———. 1985. Statistical abstract of the United States: 1986 . Washington, D.C.: Government Printing Office.

———. 1991. Statistical abstract of the United States: 1991 . Washington, D.C.: Government Printing Office.

U.S. Commissioner of Labor. 1890. "Part III. Cost of Living." In Sixth annual report of the Commissioner of Labor, 1890 . U.S. Congress, House of Representatives, House Executive Document 265, 51st Cong., 2d Sess. Washington, D.C.: Government Printing Office.

———. 1891. "Part III. Cost of Living." In Seventh annual report of the Commissioner of Labor, 1891 . U.S. Congress, House of Representatives, House Executive Document 232, 52d Cong., 1st Sess. Washington, D.C.: Government Printing Office.

———. 1904. "Cost of living and retail prices of food." In Eighteenth annual report of the Commissioner of Labor . Washington, D.C.: Government Printing Office.

Venti, Stephen F., and David A. Wise. 1989. "Aging, moving, and housing wealth." In The economics of aging , ed. David A. Wise, 9-48. Chicago: University of Chicago Press.

———. 1990. "But they don't want to reduce housing equity." In Issues in the economics of aging , ed. David A. Wise, 13-29. Chicago: University of Chicago Press.

Warner, Sam Bass, Jr. 1962. Streetcar suburbs: The process of growth in Boston, 1870-1900 . Cambridge: Harvard University Press.

Wise, David A. 1989. "Overview." In The economics of aging , ed. David A. Wise, 1-7. Chicago: University of Chicago Press.


227

PART THREE
WIDOWHOOD


229

Eight
The Impact of Widowhood in Nineteenth-Century Italy

David I. Kertzer and Nancy Karweit

The historical demography of aging has come only slowly to southern Europe. As elsewhere, historical demographic study has concentrated heavily on problems of fertility, marriage, and, to a lesser extent, migration, all of which have focused attention on youths and young adults. Indeed, even the study of mortality, which might be expected to lead to a special concern with the older segment of the population, has instead been dominated here, as elsewhere, by a focus on the youngest of the young, the death of infants.

The one historical demographic field that has shed some light on the lives of the older segment of the population is the study of household composition. This field has expanded tremendously in Italy over the past fifteen years, influenced in no small part by the work of Peter Laslett (1972) and associates of the Cambridge Group for the History of Population and Social Structure. Yet, even here, systematic work on the lives of the elderly is rare. More important, life course studies that examine the dynamics of household processes for the older segment of the population are almost nonexistent. This is in contrast with work done farther north in Europe, where concern for inheritance systems and problems of peasant retirement have generated a lively literature on later life transitions in co-residential arrangements (Berkner 1972; Gaunt 1983; Plakans 1989).

What makes the study of the co-residential careers of older Italians of special interest—beyond what it tells us of Italian society itself—is the variation in household forms found in Italy. While parts of Italy were characterized by nuclear family systems, large portions of the peninsula were marked in the past by an entirely different family system, involving the co-residence of two or more related nuclear families in a single household. The implications


230

of such a co-residential system for the later life course in general and widowhood in particular are great. For the aging individual living in a nuclear family household in which children leave home at marriage, old age may mean increased isolation and increasing difficulties in economically sustaining the household. Loss of a spouse in such a context can be disastrous, leaving the individual entirely alone.[1]

At the other extreme, in a multiple family household system, such as characterized the sharecropping zone that occupied much of central and a good part of northern Italy, aging does not entail increased isolation, nor does widowhood leave the individual on her or his own resources. As practiced in Italy, this system meant all sons might bring their brides into their parental home. New households were formed not at marriage but on the dividing up of multinuclear households when these became too large to be supported on a single agricultural holding or when tensions among the constituent units led to a breakup.[2] In such a setting, aging meant not a diminishing number of co-residents but an increasing number. Moreover, loss of a husband, for example, did not mean a household without men to take on the male household roles, nor did loss of a wife mean the absence of women to perform the roles assigned to women in the household economy.

This, anyway, is the theory. Just how this system worked as far as the older population is concerned has been little studied. We undertake such a study here, considering at the same time some of the factors that must be included in any larger theory of aging and co-residence. These factors include political economy, demography, gender systems, and culture.

We examine the interrelationship of these forces through the study of a single community, Casalecchio di Reno, lying just outside the city of Bologna. Casalecchio is of interest because it had long been dominated by a sharecropping economy but, in the period in question, 1861-1921, began to be transformed by the spread of capitalism, industrialization, and urban life. Our period of study is also of interest because it is one that saw the beginning of the demographic transition, with the fall of both mortality rates and birthrates. In short, Casalecchio not only illustrates the workings of the multiple family co-residential system associated with sharecropping Italy but also provides a window for looking at the effect the demographic and economic changes associated with "modernization" had on the co-residential lives of the older population.

Casalecchio Di Reno

The commune of Casalecchio lies just outside what for centuries had been the rural belt surrounding the walled city of Bologna, in Emilia-


231

Romagna, located in north central Italy. It had long been part of the classic sharecropping area that extended south through Tuscany, the Marches, and Umbria. Landowners lived in the cities and let their land out in small parcels to sharecropping households, either directly or through middlemen. The produce was split (often evenly) between landowner and sharecropper.

A contract bound the entire sharecropping family; emblematic was the fact that landowner consent was required before a family member could marry. Since landowners stood to gain by maximizing the number of adults on each farm, hence maximizing their half of the produce, households composed of more than one kin-related family were the rule. Indeed, the sharecroppers followed the cultural norm of patrilocal postmarital residence quite closely, with sons (not just a single son) bringing their brides into their natal household and daughters joining their husbands' households at marriage (Kertzer and Hogan 1988).

A sharp rise in the rural population in the Bologna area that began in the late eighteenth century, combined with the move by landowners to place more land on a wage labor basis, resulted during the nineteenth century in a surplus rural population that could not be absorbed into the sharecrop-ping sector (Bellettini 1981). This population was funneled into both agricultural and nonagricultural wage labor. Thus the late nineteenth century and early twentieth century were a watershed period, when wage labor began to replace sharecropping as the backbone of the rural economy (Masulli 1980; Sereni 1968).

Casalecchio reflects these patterns; its population swelled from 2,400 to 6,000 in the six decades from 1861 to 1921. At the beginning of this period, when Bologna had just been liberated from papal state rule and made part of the unified Italy, a majority of the Casalecchio population worked directly in agriculture, and 70 percent of these were sharecroppers. But changes were already evident, for the largest textile factory in the province had just been established in Casalecchio, employing large numbers of women, children, and men. By the end of the sixty-year span, slightly over one-fourth of the workforce was in agriculture, with another one-half composed of nonagricultural wage laborers, alongside a growing number of small merchants and artisans.

That this was also a period of demographic transformation is evident from both mortality and fertility rates. The crude death rate in Casalecchio sank steadily, from 30 per thousand annually in the 1870s to 17 per thousand by the early 1920s. As elsewhere, this was due in good part to a decline in infant mortality rather than to greater life expectancy for adults. The proportion of all deaths occurring to children under age 5 sank from 45 percent in 1869-1883 to 23 percent in 1904-1918.[3] Over these six decades, fer-


232

tility also began its decline, with crude birthrate sinking from about 35 in the first portion of this period to about 21 by the end.[4]

Co-Residence in Casalecchio

In looking at the households in which the people of Casalecchio lived in this period, the sharecropping legacy stands out clearly. In 1871, 40 percent of the population lived in households containing two or more nuclear family components (i.e., multiple family households), virtually the same as those living in nuclear family units (43 percent). An additional 13 percent lived in extended family households, that is, those containing some kin beyond the nuclear family but not containing two nuclear units.[5] A half century later, the situation had changed, though not dramatically. The rise of the proletarian sector led to an increase in the proportion of people living in nuclear family households (52 percent) and a decline in the proportion living in multiple family units (31 percent). The proportion living in extended family households remained virtually unchanged (14 percent).[6]

These communitywide figures, though, mask the fact that in the share-cropping sector, multiple family households continued to dominate (76 percent of those living in sharecropper households lived in multiple family units in 1871, compared to 71 percent in 1921). By contrast, among the agricultural and nonagricultural wage workers, nuclear family households had always constituted a majority, though substantial proportions lived in extended and multiple family households.[7] Indeed, given life course considerations, we could not say without further examination whether or not older widows and widowers lived with their married children among the proletarian population.

A Demographic Profile of Widowhood

Given the higher mortality rates pertaining in the past and, consequently, the greater variability in age at death, it is reasonable to assume that widowhood was less exclusively bound to old age in the past than it is today. Illustrative are the calculations of James Smith (1984: 433), who compares England's preindustrial population with England in 1981. Of women who became widowed in preindustrial England, 20 percent did so before reaching age 45, and another 45 percent became widowed before age 65. In contrast, in contemporary England, very few (2 percent) are widowed before age 35, and only 21 percent more are widowed before age 65. While 77 percent of English women widowed in 1981 became widows only after age 65, the corresponding proportion for the preindustrial period is only 35 percent.[8]

Thanks to our reconstruction of the Casalecchio population, we are able to look at the timing of widowhood more directly than is permitted either


233

TABLE 8.1
Age at Widowhood, by Sex and Period, Casalecchio

 

Men

Women

Age

1865-1889

1890-1914

1915-1921

1865-1889

1890-1914

1915-1921

16-39

17%

13%

17%

16%

15%

33%

40-49

24

11

18

22

19

13

50-59

21

24

24

26

27

21

60-69

22

32

14

27

29

23

70+

15

20

27

9

11

10

 

N

123

157

78

198

225

113

SOURCE : Casalecchio population register.

by census-based methods or by those employing simulation. In table 8.1 we show the results, divided into three periods: 1865-1889, 1890-1914, and 1915-1921. The first of these corresponds to the most traditional of our periods, in which the agricultural sector of the economy remained dominant and birth and death rates remained high. The second period is one of accelerating demographic, economic, and political changes. We separate this from the third period, which saw the disruption caused by the First World War and by the flu epidemic of 1919.

In our first period, we see a relatively flat curve for widowhood, especially for men. Forty-one percent of the men who lost their wives had done so before age 50, while 38 percent of the women widowed became widows before they turned age 50. Only a little over a third of the incidents of loss of spouse occurred to men and women over age 60. The second period sees the expected trend for men, though, curiously, such a trend is not evident for women. The proportion of incidents of men becoming widowed before age 50 sinks from 41 percent to 24 percent, but for women this proportion declines only slightly, from 38 percent to 34 percent. To some extent, this pattern may be attributable to a reduction in the marital age gap, yet this reduction by itself is too modest to explain much. It is also tempting to speculate that the reduction in the mortality of young adults in this period benefited women disproportionately, presumably through a decline in mortality associated with childbirth. This remains to be verified.

The effects of the First World War and the flu epidemic that followed it are evident when we look at the period 1915-1921. The impact of the war is seen clearly from the skyrocketing rates of young women's widowhood, with a third of all the widows being drawn from women under age 40, twice the proportion found in the earlier periods.

Looking at age at widowhood alone cannot tell us how large a proportion of the population was composed of the widowed and how these were dis-


234

TABLE 8.2
Percent of Population Widowed, by Age, Sex, and Census Year

 

Aged 18-35

Aged 36-49

Aged 50+

Census Year

Men

Women

Men

Women

Men

Women

1861

2%

2%

5%

8%

18%

37%

1881

0

1

6

7

22

38

1911

0

1

2

8

18

36

1921

1

3

4

9

18

34

SOURCE : Casalecchio manuscript censuses.

tributed by age. For this view, we turn to the manuscript censuses to see if we can find any historical trends. Table 8.2 breaks down the population of Casalecchio by sex and into three age groups: 18-35, 36-49, and 50+. We look at four censuses over the period 1861-1921.[9]

Looked at in this way, we find that few young adults (ages 18-35) were widowed. It should be recalled that mean age of first marriage for men was 27 and mean age of first marriage for women 23 to 24. Under 10 percent of both men and women aged 36-49 were widowed in any of our census years, although the proportion of women of this age group who were widowed was always higher than that of the men. Strikingly, the decline in mortality that occurred in these sixty years did not result in any change in proportion widowed for those over age 50. Indeed, no changes in these proportions are evident from 1861 to 1921.

There are some reasons to suppose that there should be a direct relationship between declining mortality and declining proportions of people over 50 who are widowed. First of all, as the expectation of length of life for those of marital age extends beyond 50, one would expect a higher proportion of those above age 50 to have a living spouse. This especially holds where, as is generally the case, the transition to lower adult mortality results not only in an increased mean life expectancy but also in a decrease in the variation found in age at death. Countervailing forces might include an increase in the disparity between male and female age at death, but this is not normally associated with this early period.

One factor, however, that must be considered in any analysis of this type is propensity toward remarriage. The likelihood that a widow or widower would remarry shrank by roughly a half over our period and especially affected the younger widowed population. It would thus appear that the decline in proportion of the widowed we might have expected from the declining adult mortality rate was counterbalanced by a decline in the rate of remarriage. The older population of Casalecchio was, as a result, just as likely to be widowed in the first two decades of the twentieth century as they


235

had been in the mid-nineteenth century. Before turning directly to an analysis of the impact of widowhood on co-residence, then, we first pause to look at the effect of remarriage.

Remarriage

The fate of the widowed is directly tied to the prevalence of remarriage. Yet studies of remarriage still remain underdeveloped (J. Smith 1984: 435), despite the notable contribution made by the publication in 1981 of Marriage and remarriage in populations of the past (Dupâquier et al. 1981). A few basic patterns are clear, however: (1) widowers almost everywhere are more likely to remarry than widows; and (2) the younger the individual at widowhood, the more likely he or she is to remarry. This pattern extends beyond Europe to North America (e.g., Keyssar 1974). Typical are the results from early eighteenth-century France, where Jacques Dupâquier found 80 percent of men widowed in their 20s remarrying compared to 67 percent of the women, sinking by age 40 to 49 to 52 percent of men and 20 percent of women (in Hufton 1984: 357).

We expect political economic factors to influence rates of remarriage in different places and different periods. For example, Jacek Kochanowicz (1983: 162) explained that in the Polish feudal economy of the eighteenth century, "if a farm was without some elements of its manpower, it ceased to be fully productive. Therefore, the lord required widowers and widows, and especially the latter, either to remarry or to leave the plot."

That there may be an important family economic component in determining rates of remarriage may be inferred from regional differences in remarriage rates in Italy itself. Massimo Livi Bacci's (1981: 357) study of remarriage in the 1880s found much higher rates prevailing in the south than in the center, at least for those who were widowed before age 50. This is provocative, for it goes against the received wisdom concerning southern Italian culture. According to this view, on losing her husband the widow dons black clothes, wearing these as a sign of devotion to her husband's memory for the rest of her life. Yet 49 percent of southern Italian women widowed before age 50 remarried, compared to only 28 percent of the women in central Italy. We concur with Livi Bacci's tentative explanation for this sharp difference: in central Italy, the sharecropping economy, with its large, multiple family household, provided security for the widow, who was not left alone with her children. In contrast, the nuclear family household that prevailed in much of southern Italy placed greater pressure on widows (and widowers) to remarry.[10]

Italy's pattern of remarriages in the nineteenth century, taken as a whole, was similar to that found in other western European countries, including England and France (Livi Bacci 1981: 348). One important trend found in


236

nineteenth-century, Italy was a continued decline in the tendency to remarry, as measured by the proportion of marriages involving one or more previously married individuals. Athos Bellettini (1981: 260), for example, reports that the proportion of marriages in Italy involving at least one previously married individual sank from 19.4 percent in 1864-1870 to 13.8 percent in 1891-1900. In both periods, almost twice as many men remarried as did women.[11]

In Casalecchio over the period 1865-1921, we count 361 men whose wives died and 539 women whose husbands died. Of course, the older the couple, the more likely one of them was to die. However, 16 percent of the widowers and 20 percent of the widows were under age 40 when their spouse died, with a total of 33 percent of the widowers and 39 percent of the widows under age 50. It was thus not uncommon for a person to lose his or her spouse while still caring for dependent children.

Remarriage was not common in Casalecchio, and in keeping with the Italy-wide pattern, it became even less common over the years.[12] In the earlier period (1865-1882), 12 percent of all marriages involved at least one widow or widower, while by the latter period (1897-1921), this had sunk to 5 percent. Widowers were more likely to remarry, than widows; in fact, re-marrying widowers outnumbered remarrying widows by 3:1 in the nineteenth century and by 2:1 in the twentieth century.

Italian men also waited less time before remarrying than did Italian women. In the mid-1880s, for example, almost one-fifth of all Italian widowers who married did so within six months of the death of their wives, and another one-fifth did so within the next six months. By contrast, widows were considerably more restrained, with widow remarriage within six months of a husband's death extremely rare (0.5 percent). Such rapid widower remarriage, however, was not as common in the Bologna area, including Casalecchio, presumably due to the household economic system described above.

Of the 110 men who remarried in Casalecchio from 1865 to 1921, almost half (46 percent) were still in their 20s or 30s. Yet only 16 percent of the men who lost a wife in Casalecchio did so while they were under age 40. This provides strong, though only indirect, evidence of the relationship between age and propensity to remarry. Put differently, only 58 men in this age category became widowers in this period, while 51 widowers in this age category remarried. By contrast, while 303 men over age 40 lost their wives in these years, just 61 men over age 40 remarried.

Although widows were much more likely to remarry if they were in their 20s or 30s (indeed, 52 percent of all widows remarrying were in this age range), even young widows had only limited remarriage possibilities. Indeed, while 539 women became widows in Casalecchio in these years, there were only 70 widows who remarried.


237

For the great majority of women who lost their husbands, widowhood would be their lot for the rest of their lives. Even those who lost a husband at an early age, although much more likely to remarry than older widows, seldom remarried. This reflects a cultural bias against widow remarriage, together with a social system that provided, through kin co-residence, a means for widows and widowers to survive without remarrying. By contrast, there did seem to be the general expectation that men who were widowed early should remarry, and such remarriages were not hard to arrange.[13]

It is likely that the propensity for remarriage among older widowers was linked to their economic and household positions, but the relationship here is complex. The well-off sharecropper, for example, might be a good bet for attracting a spouse, despite his age and widowed status. But, living in a multiple family household, he was under much less pressure to remarry, even if he had young children, for he was surrounded by female kin. Indeed, the other women of the household might not look favorably on the prospect of a new female authority figure among them. However, the poor proletarian widower might have a difficult time surviving without a wife, especially if he had small children; yet he was much less desirable as a spouse.

The Impact of Widowhood on Co-Residence

With all the studies of household composition in the European past that have appeared over the past two decades, rather few have considered the impact of widowhood on the surviving spouse's co-residential circumstances. The issue, though, is of considerable importance, not only for understanding the social meaning of widowhood in the past and not only for understanding more about the lives of other people but also for understanding the very nature of the household system. Just as the moment of marriage is a key component in defining household formation systems, so, too, is the moment when the marriage is broken. Where marriage entails the formation of a new household, for example, we may still find extended family units as a regular part of people's life course co-residential experiences if taking in one's widowed parent is the norm. Thus, we need to avoid an exclusive focus on marriage as the diagnostic element of household systems. One implication of this is that rather than simply dividing household systems in Europe into nuclear, stem, and joint, based on postmarital residence rules, we need to consider the complication added by rules governing the residence of the widowed.

Perhaps the most active setting for work on the impact of widowhood on co-residence has been England. Here the issues have not only been framed in demographic terms but have also been related to debates over the Poor Laws, with their controversies over where the family's responsibilities end and the community's or state's begin (J. Smith 1984).


238

Laslett (1977: 199-200) concludes that in preindustrial England the loss of a spouse entailed no change in household arrangements for the survivor, though the whole remaining family might move to another community. It is striking, however, that evidence from mid-nineteenth-century Preston reveals another pattern. Looking at widowers and widows over age 65, Michael Anderson (1971: 140) finds that 50 percent of the former and 41 percent of the latter lived with a married child, compared to just 15 percent and 17 percent of married men and women, respectively. He finds a similar pattern in his agricultural village sample (ibid.: 84). If this represents a change from the preindustrial pattern, it would be instructive to know just when, how, and why this change took place. Jean Robin's (1984) finding of a high proportion of elderly living with kin in nineteenth-century Colyton brings into question Laslett's earlier generalization. The lack of separate data on the widowed in Robin's stud),, however, makes conclusions here difficult.

Studies of the co-residence of widows and widowers in Italy are only now getting under way and focus largely on the situation of women (Palazzi 1990). Yet the Italian case is an interesting one given the prevalence of complex family household arrangements in much of Italy. Insofar as people live not in nuclear family households but in larger households surrounded by various kin, we would expect the co-residential impact of widowhood to be considerably reduced. In particular, loss of one's spouse would not entail isolation, nor would it have the same economic impact on the household unit.

A first view of where the widowed population of Casalecchio lived and how they differed from people who were still married is provided in table 8.3. There we look both at 1881 and 1921, comparing widows and widowers over age 50 with their married counterparts of the same age. In both years, a slight majority of the married men and women lived in simple family households, while around two-fifths lived in multiple family households. Fewer than 10 percent of the married men and women lived in extended family households. In comparing 1881 with 1921, what is striking is the lack of change in household arrangements of older married adults during a period of dramatic economic, social, political, and demographic change.

Compared to those still married, the widowed were much more likely to live in extended family households, largely at the expense of simple family co-residence. The changes associated with "modernization" and urbanization did not bring about any notable increase in the proportion living by themselves or in nonfamily situations (a proportion that, in total, hovers around a tenth for both widows and widowers). The lesser proportion generally living in multiple family households can be attributed to the impact of spousal death on households consisting of two conjugal nuclei. Using the standard typology, this has the effect of transforming a multiple family household into an extended family household.


239

TABLE 8.3
Co-Residence of Casalecchio Population over Age 50, Married versus Widowed, by Sex, 1881 and 1921

 

1881

1921

 

Married

Widowed

Married

Widowed

Women

       
 

Solitary and no family

8%

12%

 

Simple family household

54%

19

52%

22

 

Extended family household

9

38

6

43

 

Multiple family household

41

34

42

22

   

N

123

84

321

185

Men

       
 

Solitary and no family

0%

11%

1%

13%

 

Simple family household

54

32

57

19

 

Extended family household

9

21

6

38

 

Multiple family household

37

37

36

30

   

N

185

57

418

103

SOURCE : Casalecchio manuscript censuses.

Another view of the co-residential situation of the widowed is provided in table 8.4, which looks at the relationship of widows and widowers to the head of the household in 1881 and 1921. Again, we find rather little change over time. Most widows live not as household heads but in households headed by a son, generally a married son. A woman is many times more likely to live in her married son's household than in the household of her married daughter. This, by the way, contrasts with the picture painted in nineteenth-century England by Anderson (1971: 56), who found that two-thirds of all widows who lived with married children lived with a married daughter. Robin (1984) found a similar pattern of residence with married daughters rather than married sons in nineteenth-century Colyton. Clearly, in this part of Italy we are dealing with a strongly patrilateral system, and this system affected not only postmarital residence choices but also choices of residence for those who became widowed.

The influence of gender on the impact of widowhood is clear from a comparison of the widows and widowers in table 8.4. Before widowhood, the large majority of women had been the wife of the household head and, with him, codirector of the household. For most of these women, widowhood meant loss of this status, as they became—at least formally—dependents of their sons. The lines of authority between the widow and her daughter-in-law in such situations remains one of the most pressing questions of historical inquiry on widowhood in this part of Italy. The impact of losing a spouse was quite different for a man, for in the great majority of cases, losing one's


240

TABLE 8.4
Relationship to Household Head of Casalecchio Widows and
Widowers over Age 50, 1881 and 1921

 

1881

1921

Widows

   
 

Head

16%

25%

 

Mother of head

50

54

 

Mother-in-law of head

8

9

 

Other kin

13

5

 

Nonkin

13

7

   

N

84

185

Widowers

   
 

Head

79%

74%

 

Father

7

13

 

Brother

5

2

 

Other kin

2

5

 

Nonkin

7

6

   

N

57

103

SOURCE : Casalecchio manuscript censuses.

wife did not mean any formal change in household status. He remained, as he was, a household head.

The database we have compiled for the Casalecchio population allows us to take a more direct look at the impact of loss of one's spouse on household arrangements. We created a file consisting of all those whose spouse died while they were living in Casalecchio at any time during the period 1865-1921. Identifying the household in which they lived at the time of the spouse's death, we asked what impact the death of the spouse had on household composition. We retrieved household composition six months before the death and compared it with household composition six months later. The results are shown in table 8.5.

The majority of women lived in simple family households before widowhood. Death of a husband had the immediate effect of leaving the widow in a household by herself in one-fifth of the cases. Other relatives came to join the household of the widow in only 6 percent of the cases, while 16 percent of the widows (with any co-resident children they may have had) left Casalecchio within six months of the husband's death. The other important co-residential situation in which women found themselves on the eve of their widowhood was in multiple family households (in which 36 percent of the women lived). Husband's death here rarely meant either the immediate departure of the household from Casalecchio or the formation of a simple family household (presuming the departure of the widow). In 69 percent of the cases, in fact, the household continued to have two or more


241

TABLE 8.5
Impact of Widowhood on Househould Composition, Casalecchio, 1865-1921

Household Composition
Six Months after Widowhood

Household Composition Six Months before Widowhood

Simple

Extended

Multiple

Other

Total

N

Women

           
 

Moved

16%

9%

2%

19%

10%

56

 

Solitary

21

10

12

65

 

No family

0

5

0

2

 

Simple

57

13

5

52

36

194

 

Extended

2

74

25

10

14

74

 

Multiple

4

4

69

5

27

147

   

Total

56%

4%

36%

4%

   
   

N

299

23

195

21

 

538

Men

           
 

Moved

9%

3%

16%

6%

23

 

Solitary

20

16

11

39

 

No family

1

0

2

 

Simple

67

4

32

37

134

 

Extended

2

94%

24

10

15

56

 

Multiple

2

6

69

21

30

109

   

Total

50%

5%

40%

5%

   
   

N

184

17

144

18

 

363

SOURCE : Casalecchio population register.

simple family nuclei within it, while in one-fourth of the cases the death meant the widow was left living simply with one of her married children's (almost always a son's) family in an extended family unit.

A look at the men whose wives died reveals that the sex of the deceased had very little effect on the impact of widowhood on household composition. Death of a husband more commonly resulted in the departure of the remaining household from Casalecchio in a short period of time, but, otherwise, there is very little to distinguish the co-residential impact of widowhood for men and women.

How frequently loss of a spouse meant emigration from the community of residence is a question that has seldom been studied empirically, either in Italy or elsewhere. We look into this question in Casalecchio, considering differences both by sex and age at widowhood.

As seen in table 8.6, where we show proportion of widows and widowers who leave Casalecchio within one year of widowhood, loss of a husband was more likely to lead to the widow's emigration than was loss of a wife in the


242

TABLE 8.6
Proportion Leaving Casalecchio within 1 Year of Widowhood, by Age and Sex, 1865-1921

 

Men

Women

Age

%

Total N

%

Total N

16-39

9

55

22

102

40-49

5

61

16

100

50-59

17

83

17

136

60-69

16

88

16

144

70+

17

71

37

54

 

Total

13

358

20

536

SOURCE : Casalecchio population register.

case of the widower (20 percent of the women left within a year of widowhood, compared with 13 percent of the men). Interestingly, the most pronounced gender differences are found for the young and the old, with those widowed at ages 50 to 69 showing no gender difference at all in propensity to migrate. In the case of the young widows and widowers, this might be attributable to patrilocality, with the widow and her young children returning to live with family members in her own community of origin. The young widower, by contrast, would be likely to seek the help of his own kin, who would tend to live in Casalecchio. Why the oldest widows should be so much more likely to leave Casalecchio on the death of their spouse than men who are widowed at the same age is less readily explicable.

A fuller view of the likelihood of emigration for widows and widowers is provided through survival analysis, which is not limited simply to a single year following widowhood and which takes into account the effects of death of the survivor.[14] These results reveal a clearer pattern of age differences in emigration among both widows and widowers and also allow us to compare the propensity of emigration of men and women on widowhood with the propensity of their married counterparts in the population.

Figure 8.1 simply divides all those who experienced widowhood by age, looking at their cumulative survival, that is, the proportion over time remaining in Casalecchio, controlling for death and the end of the period of observation (1921). We note here no difference in tendency to emigrate between the two youngest age groups (16-39 and 40-49 at widowhood) but a progressively sharp increase in migration propensity among the next two groups (aged 50-59 and 60+).

This in itself may tell us nothing about the impact of widowhood on propensity to emigrate, since the results could simply reflect underlying age-specific migration rates in Casalecchio. We thus needed to identify an ap-


243

figure

Fig. 8.1.
Migration, by age widowed, 1865-1921: Cumulative 
survival after widowhood (Total N = 906).

figure

Fig. 8.2.
Migration following widowhood for women versus
 migration of married women after 1881, by age:
 Cumulative survival.

propriate control group to see if the experience of widowhood itself affected migration propensity. We took all those in the 1881 census who were married, following them over time to see when, and if, they migrated.[15] The results are shown in figures 8.2 (for women) and 8.3 (for men). The most striking result here is that older widows (over age 60) are considerably more


244

figure

Fig. 8.3.
Migration following widowhood for men versus migration
 of married men after 1881, by age: Cumulative survival.

likely to emigrate than are older married women. Moreover, age differences in migration propensity among married women are minor compared to the substantial differences found among women who are widowed.

To test the significance of these relationships, we employed the Lee-Desu statistic comparing survival curves. No significant difference (p << .10) in survival curve was found among the three age groups of married women. By contrast, among those who were widowed, all relationships were significant. [16] Identical results were found for the men, with no difference in emigration propensity for married men among the three age groups yet significant relationships found for those who were widowed. Loss of a spouse for older people—whether men or women—meant a significantly greater prospect of departure from the community than that faced by their married age-mates.

Conclusion

Understanding the impact of widowhood in the past presupposes an understanding of the workings of household systems. We would expect that insofar as different household systems prevailed, differences would be found in the impact of widowhood on people's lives. Unfortunately, research on the interaction of such systems with widowhood is not yet well developed.

Theoretical explication of the relationship between household system and widowhood requires an understanding of the influence of demographic, gender, political economic, and cultural forces. Demographic forces affect both age at widowhood and proportion who are widowed,


245

and we would expect changes in such demographic elements as mortality rates and age at marriage to affect the prevalence of widowhood. Yet as we have seen in the case of Casalecchio, the interaction of demographic elements can be complex. Thus, for example, changing remarriage rates complicate the secular trends that would be expected to result from declines in mortality.

Differences between men and women in the impact of widowhood have thus far not been well studied historically, with the exception of remarriage, where the much greater tendency of males to remarry is now well established. Gender differences can be seen in Casalecchio in the greater likelihood of women to be widowed and in that, for women, loss of spouse meant a decline in formal household status, while it did not for men. Yet here we enter into the difficult terrain of the relationship between formal positions of authority and actual power relations. In a multiple family household, a husband's death meant being replaced as wife of household head by one's daughter-in-law. Yet it is not clear what this meant in practice for the older woman's ability to influence household decisions.

Political economic forces clearly have a great deal of influence on the social implications of widowhood. We have pointed out the apparent influence of economic differences between southern and central-northern Italy on the remarriage rates of the widowed. In this case, it seems that such forces outweighed the presumably greater cultural constraints on remarriage found in southern Italy. More generally, the prevalence of multiple family households in central Italy is the product of the political economy of sharecropping, a system that has great implications for the lives of widows and widowers.

Yet we cannot discount the role of cultural norms regarding proper behavior, not only insofar as remarriage is concerned but also with respect to the obligations people felt toward widowed parents and parents-in-law. Richard Smith (1984: 425) has made this point with respect to Britain in the past, and the situation in Casalecchio bears him out. First of all, we have seen that the tremendous economic, political, and demographic changes of the six decades under consideration did not result in equally dramatic changes in the co-residential circumstances of the widowed. Even though by 1921 sharecroppers had become only a minority of the population, two-thirds of the widowed over age 50 still lived with extended kin. Moreover, the patrilateral principle that predominated, again associated with a sharecropping family economy, continued to guide co-residential decisions of the widowed even after proletarianization. The tendency to live with a married daughter rather than a married son, documented for parts of nineteenth-century Britain, is simply not found in this part of Italy.

The rise of proletarianization and the slow demise of a sharecropping household system that offered special support to the widowed population


246

(as to the older population in general) thus did not necessarily lead to the co-residential abandonment of the widowed. However, the sharecroppers' large, multiple family household had offered special protection to the widowed, and the same degree of protection would be hard to duplicate in a proletarian family economy, at least until the economic position of the wage workers would itself improve substantially.

References

Alter, George. 1988. Family and the female life course: The women of Verviers, Belgium, 1849-1880 . Madison: University of Wisconsin Press.

Anderson, Michael. 1971. Family structure in nineteenth-century Lancashire . Cambridge: Cambridge University Press.

Barbagli, Marzio. 1984. Sotto lo stesso tetto: Mutamenti della famiglia in Italia dal XV al XX secolo [Under the same roof: Family changes in Italy from the fifteenth to the twentieth century]. Bologna: II Mulino.

Bellettini, Athos. 1981. "Les remariages dans la ville et dans la campagne de Bologne au dix-neuvième siècle" [Remarriages in the city and the countryside of Bologna in the nineteenth century]. In Marriage and remarriage in populations of the past , eds. J. Dupâquier, E. Hélin, P. Laslett, M. Livi Bacci, and S. Sogner, 259-272. New York: Academic Press.

Berkner, Lutz. 1972. "The stem family and the developmental cycle of the peasant household: An eighteenth-century Austrian example." American Historical Review 77: 398-418.

Bongaarts, John. 1989. "The demographic determinants of the duration and incidence of widowhood." In Later phases of the family cycle , eds. E. Grebenik, C. Höhn, and R. Mackensen, 55-65. Oxford: Clarendon Press.

Da Molin, Giovanna. 1990. "Family forms and domestic service in southern Italy from the seventeenth to the nineteenth centuries." Journal of Family History 15: 503-528.

Dupâquier, Jacques, E. Hélin, P. Laslett, M. Livi Bacci, and S. Sogner, eds. 1981. Marriage and remarriage in populations of the past . New York: Academic Press.

Gaunt, David. 1983. "The property and kin relationships of retired farmers in northern and central Europe." In Family forms in historic Europe , ed. Richard Wall, Peter Laslett, and Jean Robin, 249-280. Cambridge: Cambridge University Press.

Hammel, Eugene A., and Peter Laslett. 1974. "Comparing household structure over time and between cultures." Comparative Studies in Society and History 16: 73-103.

Hogan, Dennis P., and David I. Kertzer. 1985. "Longitudinal approaches to migration in social history." Historical Methods 18: 20-30.

Hufton, Olwen. 1984. "Women without men: Widows and spinsters in Britain and France in the eighteenth century." Journal of Family History 9: 355-376.


248

Kertzer, David I. 1984. Family life in central Italy, 1880-1910 . New Brunswick: Rutgers University Press.

———. 1989. "The joint family household revisited: Demographic constraints and household complexity in the European past." Journal of Family History 14: 1-15.

Kertzer, David I., and Dennis P. Hogan. 1988. "Family structure, individual lives, and societal change." In Social structures and human lives , ed. Matilda W. Riley, 83-100. Newbury Park, Calif.: Sage.

———. 1989. Family, political economy, and demographic change: The transformation of life in Casalecchio, Italy, 1861-1921 . Madison: University of Wisconsin Press.

Keyssar, J. 1974. "Widowhood in eighteenth-century Massachusetts: A problem in the history of the family." Perspectives in American History 8: 83-119.

Kochanowicz, Jacek. 1983. "The peasant family as an economic unit in the Polish feudal economy of the eighteenth century." In Family forms in historic Europe , ed. Richard Wall, Peter Laslett, and Jean Robin, 153-166. Cambridge: Cambridge University Press.

Laslett, Peter. 1972. "Introduction: The history of the family." In Household and family in past time , eds. Peter Laslett and Richard Wall, 1-89. Cambridge: Cambridge University Press.

———. 1977. Family life and illicit love in earlier generations . Cambridge: Cambridge University Press.

Livi Bacci, Massimo. 1981. "On the frequency of remarriage in nineteenth-century Italy: Methods and results." In Marriage and remarriage in populations of the past , ed. J. Dupâquier, E. Hélin, P. Laslett, M. Livi Bacci, and S. Sogner, 347-362. New York: Academic Press.

Masulli, Ignazio. 1980. Crisi e trasformazione: Strutture economiche, rapporti sociali e lotte politiche nel bolognese (1880-1914) [Crisis and transformation: Economic structures, social relations, and political struggles in the Bologna area (1880-1914)]. Bologna: Istituto per la Storia di Bologna.

Palazzi, Maura. 1990. "Female solitude and patrilineage: Unmarried women and widows during the eighteenth and nineteenth centuries." Journal of Family History 15: 443-460.

Plakans, Andrejs. 1989. "Stepping down in former times: A comparative assessment of retirement in traditional Europe." In Age structuring in comparative perspective , ed. D. Kertzer and K. W. Schaie, 175-196. Hillsdale, N.J.: Lawrence Erlbaum.

Robin, Jean. 1984. "Family care of the elderly in a nineteenth-century Devonshire parish." Ageing and Society 4: 505-516.

Sereni, Emilio. 1968. Il capitalismo nelle campagne (1860-1900) [Capitalism in the countryside (1860-1900)]. Turin: Einaudi.

Smith, James E. 1984. "Widowhood and ageing in traditional English society." Ageing and Society 4: 429-449.

Smith, Richard M. 1984. "The structured dependence of the elderly as a recent development: Some sceptical historical thoughts." Ageing and Society 4: 409-428.


249

Nine
The Demography of Widowhood in Preindustrial New Hampshire

Daniel Scott Smith

The Malthusian-Frontier Regime

In recent years, scholars in several fields have revived interest in the fundamental relationship between resources and population and the consequences for social institutions of population growth and density (Lee et al. 1988; Wrigley and Schofield 1981). Instead of being a minor adversary of Marx, Malthus now stands as an important figure in the history of thought (Dupâquier, Fauve-Chamoux, and Grebenik 1983). Certainly, emphasis on this connection is appealing to students of the first two centuries of American historical demography and family history. Frederick Jackson Turner ([1894] 1961) developed the implications of a low labor-to-land ratio in his frontier theory, which organized major features of the American historical experience. In Malthusian terms, the American case represents an extreme: land abundance, labor scarcity, and a very high rate of natural population increase. In contrast to rapid demographic expansion, rates of per capita economic growth before the nineteenth century were irregular over the short run and quite modest over the long run (McCusker and Menard 1985; Smith 1980).

Several important eighteenth-century American commentators also used a framework that linked cheap land and migration to the frontier to early and universal marriage and consequently to high rates of natural increase. Drawing on these writings, Malthus himself ([1798] 1960) cited the rapid population growth in America as an exceptional outcome that still illustrated his general rule that environmental constraints limited the possibility of positive rates of demographic expansion.

Without doubt, the Malthusian-frontier framework provides a parsimonious first approximation to the interpretation of major features of early


250

American history (Smith 1980). This investigation extends the theme to women (Guttentag and Secord 1983; Thompson 1974) and to the phenomenon of widowhood, a particularly interesting status because it is open to the radically divergent impacts of the frontier. A low labor-to-land ratio can lead to high wages and ownership of land for laborers, assuming they are free to make choices. This same economic environment is, however, associated historically with the institutions of slavery and serfdom. If land is cheap and labor expensive, those who want others to work for them must take away their economic choices (Domar 1970). This is the economic base of the American paradox of the coexistence of the two extremes—slavery for blacks and virtually unlimited autonomy for white farmers.

Legally, widows, unlike married women, enjoyed the right to be independent actors. The scarcity of women that is reflected in a high sex (male to female) ratio on the frontier could, all else equal, aid widows in two ways: by increasing the prospects for an advantageous marriage and by allowing for continuing economic autonomy as widows. Thus a favorable frontier effect would be apparent in a higher rate of remarriage for widows and a higher proportion of householders among those who remained widows compared to those in nonfrontier areas.

The Case of Belknap's New Hampshire

To explore this subject, New Hampshire supplies both a suitable setting and the best available data for the period before 1790 in America. It was the only colony in British North America to tally widows in its census. In both 1767 and 1773, towns returned to the colonial government the numbers of unmarried women, married women, and widows in the nonslave population. Additionally, the numbers of female householders and population data from the first census and information on the share of taxes of each New Hampshire town in 1790 were added to this data set.[1]

While census data document the incidence and covariates of widowhood, three data sets constructed from the published probate records of colonial New Hampshire wills provide evidence concerning the economic and familial context of widowhood (New Hampshire 1913-1941). The first data set contains limited information for all of the testators whose wills were probated (N = 1,455), while the second records the sparse published evidence concerning the administration of the property of those persons who died intestate (N = 1,886). The third data set records more detailed information for all women's wills up to the year 1771 (N = 93) and a same-size sample of men's wills that were recorded immediately following those for women. Unfortunately, no information on the rates of remarriage for widows is currently available.


251

New Hampshire provides an appropriate case for the study of the ramifications of a low population-to-land ratio for widowhood. On the northern New England frontier, its population increased more than twenty-three-fold between 1700 and 1790, expanding from about 6,000 to nearly 142,000 people, an astounding annual growth rate of 3.5 percent. From 1767 to 1790, demographic increase was even more rapid—4.3 percent per year, from a total of 52,720 in 1767. Because of the expansion of land in settled towns at a yearly rate of 3.2 percent, density only increased from 16.7 to 21.5 persons per square mile, a rate of 1.1 percent per annum. [2]

In the first history of the state (1784-1792), Jeremy Belknap, a minister at Dover for two decades, outlined the basic elements of the demographic system operative in eighteenth-century New Hampshire.

Land being easily obtained, and labour of every kind being familiar, there is great encouragement to population. A good husbandman, with the savings of a few years, can purchase new land enough to give his elder sons a settlement, and assist them in clearing a lot and building a hut; after which they soon learn to support themselves. The household is generally given to the youngest son, who provides for his parents, when age or infirmity incapacitates them for labour. An unmarried man of thirty years old is rarely to be found in our country towns. The women are grandmothers at forty, and it is not uncommon for a mother and daughter to have each a child at the breast, at the same time; nor for