Search

FRANCES K. GOLDSCHEIDER Brown University LEORA LAWTON TechSociety Research* We test the affluence interpretation of the decline in intergenerational coresidence, which implicitly affirms that close kin would still offer housing to those in need, by examining the factors that influence attitudes about coresidence with young adult children and aging parents. Using national data, we model the effects of living in a multigenerational household in childhood and living independently from parents prior to marriage on respondents' obligation to allow aging parents and adult children who are in need to coreside. We find strong effects of living arrangements experiences on attitudes that differ by the type of intergenerational coresidence. Key Words: affluence, attitudes, coresidence, intergenerational relations, living arrangements. Home is the place where, when you have to go there, they have to let you in. Robert Frost, "Death of the Hired Man" Shelter is a fundamental resource that family members attempt to provide each other when necessary. Although levels of intergenerational coresidence have declined dramatically in recent decades in the United States and in other developed countries, most analysts interpret these changes as the result of increases in affluence that have allowed previously dependent relatives to split off from the nuclear family to live in a home of their own (Kuznets, 1978; Michael, Fuchs, & Scott, 1980). Given that splitting off increases the privacy of both those who split off and their relatives, the increased number of young adults leaving home prior to marriage (Goldscheider, 1997) and the greater likelihood that widowed parents live alone (Ruggles, 1996) are usually seen as strictly beneficial. The affluence interpretation underlies this positive view, and it suggests that although the necessity for coresidence has declined, there has been no reduction in the commitment to provide it. Rather, ascribing change to affluence implicitly affirms that close kin would still offer housing to those in need. However, changes in preferences may have been as important as increases in affluence in the splitting off of extra relatives from American families (Pampel, 1983). The outcry in the popular press (Cowan, 1989; Gross, 1991; Lindsay, 1984) and from some scholars (Schnaiburg & Goldenburg, 1989) against the recent increase in grown children living with parents, despite the dramatic increases in housing costs and youth unemployment (Levy, 1987), suggests that Americans may have become increasingly opposed to parents and adult children sharing a residence, even if the parent or child is in a difficult situation and lacks the physical or financial resources to make separate living feasible. In fact, the commitment to provide housing to close kin in need may be weakening. Many may no longer feel that they have to let you in. These two explanations-changes in affluence and changes in preferences-are not contradictory, given that changes in attitudes and preferences are likely to occur as a result of changes in behavior. The decline in coresidence has been under way for most of the twentieth century (Ruggles, 1994). Through social-psychological behaviorist mechanisms of social learning and through the interpretation of experiences via processes such as those set forth by cognitive theorists, people develop attitudes about their family relationships and obligations (Caspi & Elder, 1988; Sroufe & Fleeson, 1988). Children watch what significant others like their parents do, hear what they say, and learn by imitating many of their parents' attitudes and behavioral choices. If increases in affluence mean that families need to provide intergenerational coresidence less often, children are less likely to observe such living arrangements and are more likely to conclude that such solutions are inappropriate or unnecessary. Of course, not all learning takes place in childhood, and most people find that the economic and family commitments of adulthood restructure their choices and priorities and increase their concern for the needs of others. …

This study investigates the following questions: whether greater affection between adult children and their parents leads to more social contact, whether frequent social contact leads to greater affection, or whether each of these mutually influences the other. Using nationally representative data collected in 1990 by the American Association of Retired Persons, we examine predictors of each dimension of solidarity and then estimate a causal model that tests the indirect and reciprocal influence among these dimensions. After finding a reciprocal influence between contact and affection in the mother-child relationship, but not in the father-child relationship, we conclude that the motivations for contact are different in adult-child relations with mothers compared to those with fathers. These differences are important for understanding the consequences of family disruption for intergenerational solidarity in adulthood. Also, parallels are drawn between parent-child relationships and voluntary friendships. Over the last decade, research on family relations has increasingly taken a multidimensional approach to studying adult intergenerational relationships, focusing on frequency of visits and phone calls, helping behavior, geographic distance between generations, and, more recently, the affection that one generation has for another. Each dimension of family relations is further interconnected with the others in ways that affect the well-being of both generations. For example, geographic mobility increases physical distance between generations, impeding the exchange of social and instrumental support (Dewit, Wister, & Burch, 1988; Litwak & Kulis, 1987). Additionally, because the dimensions are interconnected, the social forces that influence one dimension of the family relationships will indirectly influence the others. Changes in family structure, such as the increase in divorce and remarriage, may thus alter the functioning of intergenerational relationships by reshaping access to family members (Furstenberg, 1981, 1988). Therefore, a more accurate knowledge concerning parent-child relationships can be gained by considering the interrelated and mutually reinforcing dimensions as a system. In this article, we consider the mutual impact of affection, contact, and distance. A question arising in discussions about parent-child relations in later life is: Does greater affection lead to more frequent visiting, does more frequent visiting lead to greater affection, or does each of these mutually influence the other? Despite its theoretical plausibility and relevance to a more accurate interpretation of parent-child interactions in adult life, this question has remained unanswered. We address this question by examining three dimensions considered key to understanding the nature of adult parent-child relationships: affection, frequency of social contact, and geographic distance. First, we investigate each dimension of solidarity independently and then we develop a causal model to understand the interconnections among these dimensions, specifically how they mediate and reciprocally influence each other. BACKGROUND The literature on intergenerational family relations is generally sanguine about the state of intergenerational relationships in the contemporary family. Studies find that despite fears of mass alienation and abandonment, the vast majority of parents visit or speak once a week or more with at least one child (Shanas, 1979) and most live within an hour of one child (Lawton, Silverstein, & Bengtson, in press). Also, aging parents can and do rely on their children to provide caregiving and other forms of assistance (Brody & Schoonover, 1986). In spite of the generally optimistic tone of these conclusions, there are substantial sources of variation in the quality of intergenerational relations. For instance, while the level of affection between parents and children is generally high, it is weakened in the case of parental divorce, especially between divorced fathers and their children (Cooney & Uhlenberg, 1990; Lawton, 1990). …