CHANGES IN NONMARITAL COHABITATION AND THE FAMILY STRUCTURE EXPERIENCES OF CHILDREN ACROSS 17 COUNTRIES

Abstract

Only three decades ago, many demographers believed that the nuclear family – married adults and their biological children – was the modal family structure toward which all societies would rapidly converge (e.g. Goode, 1970). Indeed, during the two decades following World War II, marriage and childbearing in most Western nations tended to: (1) occur early in adulthood; (2) follow a predictable sequence; and (3) be tightly coupled. That is, young couples first married, and then quickly began having children. Over the past few decades in many Western countries, however, marriage and fertility have been increasingly delayed to later adulthood and decoupled from one another, such that the sequence and timing of partnership formation and childbearing have changed dramatically. As a result, most Western nations have experienced increasing rates of out-of-wedlock and out-of-partnership fertility and nonmarital cohabitation1 (as well as divorce) (Goldscheider et al., 2001; Haskey, 2001; Hoem & Hoem, 1988; Kiernan, 2001; Martin & Bumpass, 1989; Murphy, 2000; Noack, 2001; Ostner, 2001; Prinz, 1995; Toulemon, 1997). The pace of change has been so swift that in the preface to the second edition of Marriage, Divorce, Remarriage, Cherlin (1992, p. vii) remarks that only ten years after the publication of the first edition a more appropriate title to the book might have been Cohabitation, Marriage, Divorce, More Cohabitation, and Probably Remarriage. It is certain that recent cohorts of children in the United States and many other Western nations are far more likely to be born out of wedlock or experience the separation of their parents than were the children of previous cohorts. These family changes have sparked popular and scholarly concern over the effects of children's presumed increased exposure to single-parent families (Amato & Booth, 1997; Blankenhorn et al., 1990; Haveman & Wolfe, 1994; Popenoe, 1988, 1993, 1996; Uhlenberg & Eggebeen, 1986). However, what remains unclear is the extent to which increases in nonmarital cohabitation between children's biological parents has accounted for (or at least absorbed) the decline in the centrality of marriage as the primary childrearing family structure. In many Western nations there have been large increases in the prevalence of cohabitation in recent years (Casper & Cohen, 2000; Goldscheider et al., 2001; Smock, 2000), and less dramatic increases in the proportion of childhood spent in cohabiting family structures (Bumpass & Lu, 2000; Heuveline, Timberlake & Furstenberg, 2003). This apparent disjuncture is likely due to a combination of at least two factors: first, because cohabiting unions tend to be less stable than marriages (Axinn & Thornton, 1992; Bennett et al., 1988; Hall & Zhao, 1995; Heuveline et al., 2003), then even large increases in the incidence of cohabitation with children may not have substantially offset the decline in the amount of time children can expect to spend with married parents. Second, and more important for the purposes of this article, much of the increase in the prevalence of cohabitation may have occurred when children are not in the household. In this article we seek to understand the effects of changes in adult nonmarital cohabitation on the resulting family structure experiences of children. Our primary research question can be stated generally as: to what extent have changes in the incidence and duration of adult cohabitation been translated into changes in children's exposure to cohabiting-parent, married-parent, and other family structures? More specifically, we ask: for women and children, how much have the expected probability of experiencing cohabitation and the expected duration in a cohabiting family structure changed over time? How much of the increase over time in women's expected duration in cohabitation occurred while children were co-resident? Finally, how much has the increasing exposure of children to parental cohabitation offset declines in the amount of time children are expected to spend with married biological parents? To address these questions, we analyze partnership and fertility data from 17 industrialized nations. We construct single-decrement and increment-decrement period life tables covering, in most cases, the early 1980s to the early 1990s. These tables yield rich cross-national descriptions of expected levels of exposure to cohabitation among women and children. Social scientific research on cohabitation has been limited primarily to the United States, Canada, and a few European nations (Smock, 2000). We extend this line of inquiry to a larger sample of countries, with the goal of advancing our understanding of the extent to which partnership and fertility patterns are converging (or not) across Western nations. Our results provide, to the best of our knowledge, previously unavailable evidence regarding trends in adult cohabitation and children's family structure experiences across a number of Western nations. Although we do not address questions of the effects of cohabitation on child well-being (Manning & Lichter, 1996), the present analysis is a crucial step in tackling such questions, at least at the national level. We believe that the study of these effects has been hindered by imprecise measurement of childhood family structures, for example, lumping married parents and step-parents together by focusing on the marital status of the head of the household, or assuming the same effect on all children in a given family structure regardless of the time spent in such a family. The present analyses therefore represent a necessary step toward improved subsequent analyses of the societal-level effects of family structures on children.