The Retreat from Marriage: Its Effect on Unmarried Children's Exchange with Parents

Abstract

In the last 20 years, an increased age at marriage and an increased risk of divorce raised the proportion of adults between the ages of 25 and 54 who are unmarried from 15% to 33% (U. S. Bureau of the Census, 1972, 1992). Today, the average parent with two or three adult children is likely to have at least one unmarried child and very likely to experience a child's divorce. The effect of these changes on intergenerational systems of support is virtually uncharted. Although it has been suggested that unmarried children (especially daughters) carry a disproportionate share of the burden of caring for elderly parents (Lang & Brody, 1983), others have speculated that unmarried children are a special burden for their parents (Pillemer & Suitor, 1991; Speare & Avery, 1993). This report uses the National Survey of Families and Households (NSFH) to examine whether divorced and never-married children differ from married children in support given to and received from parents. We separate divorced from never-married children and coresidence from other kinds of support. Our focus is not on care of the elderly, but rather on the routine forms of social support that characterize family relationships during active adulthood. PRIOR RESEARCH CHILDREN'S MARITAL STATUS AND CORESIDENCE The unmarried are very much more likely than the married to reside in their parents' homes during adulthood, and most research shows that, even into parents' old age, children live with parents rather than vice versa (Aquilino, 1990). Among the nearly half of parents over 65 who have an unmarried child, 28% have a coresident child (Aquilino, 1990). For the most part, research shows that these children are a net burden, contributing little economically or in terms of household labor (Grigsby, 1989). CHILDREN'S MARITAL STATUS AND PARENTAL EXCHANGE The consensus about children's marital status and intergenerational exchange breaks down when we turn to other types of parent-child exchange. Findings from previous studies are so mixed that it is clear that measurement and design issues have a critical impact on whether one finds that unmarried children exchange more, exchange less, or are not different from married children. For example, Cooney and Uhlenberg (1992) used NSFH data to conclude that married children receive and perceive more help from their parents than do unmarried children, while Marks and McLanahan (1993) and Hogan, Eggebeen, and Clogg (1993) used the same data set to conclude that unmarried children receive more help than the married. The latter findings appear to depend on excluding childless respondents and on coding coresident children as givers and receivers of help. Findings are also mixed when we ask whether unmarried children give less to their parents. Although Lang and Brody (1983) reported that unmarried daughters give 3 times more assistance to their elderly mothers than married daughters, most studies suggest that the unmarried give less to their parents (Hogan, Eggebeen, & Clogg, 1993; Johnson, 1988; Rossi & Rossi, 1990). THE PROBLEM The analysis to follow compares never-married, divorced, and married respondents to examine whether the retreat from marriage may affect the exchange of parents with their adult children. Although using the same data set as many previous studies, our design and sample are more comprehensive: We examine coresidence (of parent with child and child with parent) separately from nonresidential exchange, examine both giving and receiving, include adult children who are childless as well as those who are parents, and distinguish between never-married and divorced children. In addition to establishing marital status differentials, we examine five moderating factors: gender of child, recency of child's divorce, whether child has own children, child's age, and parents' marital status. Control variables are distance, race, education, number of siblings and presence of sisters, and parents' age, health, and marital status--all of which can be considered causally prior to the child's marital status. …