Intergenerational household structure and economic change at the turn of the twentieth century.

The historical living arrangements of the U.S. elderly have been shaped by two dimensions of household structure. The first, household composition, has undergone radical shifts over the past two centuries. The proportion of older people who lived with adult children was high - and perhaps reached a peak - at the end of the nineteenth century.(1) By the mid-twentieth century, the rate of intergenerational coresidence declined so dramatically that the decline has been termed a "demographic revolution."(2) In contrast, there has been little change in a second dimension of household structure, intrahousehold hierarchy, or headship. Most U.S. elderly, past and present, live and have lived as heads of households or as the spouses of heads.(3)

Both trends, taken together, reveal a little-examined historical pattern' while headship or residential autonomy for the elderly is currently associated with living apart from younger kin, it was associated with extensive intergenerational coresidence in the not-too-distant past. This reversal of behavior is important because headship for the elderly is a status that denotes authority and control over household decision making.(4) It is a status that may discourage intergenerational coresidence today. This article examines both dimensions of household structure from the vantage point of older men, asking why intergenerational living arrangements occurred.

Two general perspectives, the household economy and cultural-ideational approaches, may explain intergenerational living arrangements and kin relations at the turn of the twentieth century. Living arrangements also may have reflected the transformations in community-household linkages, although the mechanisms and the directions of association between these linkages are debated.(5) I therefore use the 1910 Public Use Sample and linked decennial and other census data to examine a wide range of influences on old-age living arrangements in a sample of males age sixty-five and olden I focus on men because elderly men and women faced vastly different opportunity structures in this era. Men had choices in late life that shaped their living arrangements, which elderly women simply did not, such as remarriage and work. I have closely examined the living arrangements of older, turn-of-the-century, U.S. women elsewhere.(6)

HISTORICAL AND THEORETICAL BACKGROUND

The western elderly, past and present, have occupied potentially precarious social structural positions. Chronological aging never has been sufficient to cause old-age dependency, but the likelihood of economic, health, and social dependencies (e.g., through widowhood and/or deaths of other kin) increases with age, and prudent elders are well aware of these potential troubles.(7) They generally have few options. Depending on time and place, older men can work, live on savings if available, or else command kin or public support.(8) However, marshaling support in old age has been difficult in the western European and U.S. past because kin responsibility for elderly dependents has never been universal.(9) In fact, kin and community support has often reflected, as it does today, the possession of property, wealth, or mobilized resources at hand.(10) Thus, as a household is an important social resource, the western elderly tend to maintain residential autonomy, or household headship, for as long as possible.(11)

This long-standing problem of marshaling support has led to a number of socially structured mechanisms of intergenerational assistance that have - at least for short periods of time, from U.S. colonial times to the present - somewhat ameliorated old-age hardship. For example, prior to the early nineteenth century, when households, more than families,(12) were the key forms of social organization, dependents of all ages were allocated into households within communities.(13) Yet even with this social safety net, elderly men preferred to head households and to not coreside with young adult kin, although kin lived nearby and were expected to provide old-age assistance in the form of food and chores.(14) Moreover, elderly men did not rely on kin goodwill to provide such support but employed sociolegal institutional mechanisms of titles, deeds, and wills to enforce the provision of care or used community pressure as a last resort to secure the cost of old-age maintenance from kin.(15)

By the early nineteenth century, households ceased to be the arms of community-based control and were losing core functions (e.g., education, job placements, vocational training, sick and dependent care). This transformation had implications for turn-of-the-century elders. Those without pensions or savings would need to remain economically active, turn to kin, or turn to new forms of (dreaded) public support, often in almshouses.(16) Thus, older men in 1910 needed a different kind of cooperation from younger kin than did earlier cohorts of the aged. They needed adult children to remain voluntarily at home and contribute wages or, if not, to provide access to newly formed households, if old-age dependency arose.

This background must be kept in mind while reviewing the two major theoretical perspectives most used by family theorists to explain turn-of-the-century household composition and kin relations. Cultural-ideational, or affiliation, theorists stress that warmer, less formal interpersonal relations arose with the demise of home-based market production. Home life became marked by domesticity, withdrawal from market and community functions, and privacy and companionship.(17) Living arrangements thus reflected social change rooted in new attitudes that spurred people to move away from obligatory, rigid social forms regarding intrahousehold responsibilities and relations.(18) The core component of change was the belief that domestic units should be set off from the rest of the community, that they should shelter kin only, and that they should foster within an emotional climate conducive to interpersonal relatedness and self-expression.(19)

A second perspective, an exchange or resource-pooling model, stresses the rise of the wage-based household economy during this era, which served as a "mediating institution between individuals and the processes of large-scale social change."(20) If households were no longer units of economic production by the end of the nineteenth century, they remained units of economic decision making in the context of newly privatizing, wage-based local economies. Decisions made at the household level pertained to household members' labor market activities, levels and types of household consumption, and the serial allocation of members into ancillary activities, such as school and leisure. Kin lived together to pool their talents and resources and so acted jointly to maintain collective stability in a turbulent wage economy.(21)

The perspectives are not mutually exclusive. Affiliation theorists note that the demise of the colonial system of production and the production-centered household after 1820 freed people to adopt living arrangements based on preference rather than social duty. Similarly, resource-exchange theorists note that living arrangements cannot be reduced to economic factors. But each perspective emphasizes a different process of social organization: kin affiliation versus resource pooling and exchange.

THEORETICAL GAPS

There are problems with both perspectives, as they stand, as explanations of turn-of-the-century intergenerational coresidence. First, neither perspective has been well developed with regard to aging over the life course and in light of the very conditions they espouse. For example, resource theorists remind us of the new wage economies and the need to pool resources. But family members' contributions to the resource pool would vary by their functional abilities. While many older men could contribute, poor health or other negative impacts on productivity might arise. Therefore, resource pooling as a basis for intergenerational coresidence is inherently problematic and should not be assumed. In a similar vein, affiliation theorists note the changing public preference for companionate living and household-level privacy. But these beliefs were likely differentially adhered to by different cohorts of adults who were socialized in different eras. The need for companionship likely was felt most strongly by the young in 1910, not their parents who were born in or before 1845, a traditional era. How were "new attitudes" filtered through parents and parental control when younger kin lived in parental homes? Parents who entered children's homes might resist new norms and rules.(22) Thus, it is not clear how parents and adult children accommodated resource pooling or new expressions of individualism and self-development when living arrangements were shared.

Second, neither perspective, as formulated, addresses the population pattern of intergenerational coresidence. A sizable proportion of older married and widowed men did not live with a child (presented shortly). Both perspectives explain living arrangements, post hoc, rather than predict which subsets of adults coresided.

Finally, neither perspective addresses the critical second dimension of household structure, especially for the aged: hierarchy, as expressed in residential autonomy or headship. For U.S. elders of the past, as today, even a child's home can never be the same as one's own.(23) The potential importance of residential autonomy for the western elderly cannot be underestimated. Household headship was not a position devoid of authority but one that correlated highly with decision-making powers. And, as this article will show, older men who lived with children in 1910 tended to be household heads. Given this status, it is not clear how several generations of kin pursued strategic decision making and/or companionship in the light of old-age headship, especially if headship was highly correlated with interpersonal control. Both perspectives sidestep questions pertaining to intergenerational relations based on dependency/need and power,(24) forces that could engender intergenerational tensions and shape coresidence.

In light of these very theories, we might have hypothesized that early-twentieth-century intergenerational coresidence rates would be low. But, as demographic research finds, the demise of the colonial household economy and the rise of the modern private household did not result in residential isolation of the elderly but indeed gave rise to extensive intergenerational coresidence from the perspective of the elderly. Why? And in what contexts?

THEORETICAL MODEL

It is interesting that ideational and household economy theorists find coresidence motivated by endogenous processes of social cooperation generated at the family level, whether for the purposes of affiliation or exchange. Thus, rather than consider the possibility of intrafamily power differences, both perspectives assume states of balanced relationships as pertains to joint strategies, companionship, privacy, or resource pooling. I use exchange theory to relax this assumption. Exchange theory posits that individuals are quite aware of the give-and-take involved in relationships. They strive to reach a maximization point with regard to the costs and benefits involved in interactions or the maintenance of relations, including coresidence.(25) Individuals who are far from a maximization point are in unbalanced relationships: those who benefit less than their cost of maintaining the relationship are in positions of dependence, and those who benefit more than their cost are in positions of power in relationships. The differentials in the relative contributions made by parties to sustain relationships can be used to detect differentials in the balance of power.(26)

The turn-of-the-twentieth-century household economies of the aged with the most assets and wealth were less likely to contain adult children.(27) Yet a substantial proportion (probably about one-third to one-half) of late-middle-aged and elderly …