Table of Contents | Previous | Next |
Chapter 1: A Statistical Portrait of Stepfamilies in the U.S.
A key element of the Administration for Children and Families’ (ACF’s) strategy for implementing the Healthy Marriage Initiative (HMI) has been to identify subpopulations that would benefit from specialized program approaches.1 This report presents a conceptual framework for programs to promote stable and healthy marriage among one important subgroup, stepfamilies. Specifically, our charge from ACF was to focus on low-income married couples where one or both spouses have children by other partners. The conceptual framework is based on a review of the research literature on stepfamilies and on an informal study of marriage education programs currently serving stepfamilies.
There are several reasons to devote special attention to stepfamilies. First, as documented in the next section, stepfamilies are common as a result of divorces, remarriages, and first marriages following out-of-wedlock births. Second, as discussed in the next chapter, stepfamilies face a variety of unique challenges that generally go unaddressed in depth in most existing marriage education curricula but which may put them at higher risk for dissolution than non-stepfamilies. These challenges arise in part from complex relationships with stepchildren, former partners, and half- and stepsiblings. Third, although children can do well in a variety of family forms, it appears that living in a stepfamily is associated with greater risk for a variety of negative outcomes for children when compared to living in a nuclear family (in this paper we define “nuclear” family as one that includes only a married man and a woman and their children in common). On average, children in stepfamilies do worse on measures of social and emotional well-being when compared to children living in nuclear families (e.g., Cherlin & Furstenburg, 1994; McLanahan & Sandefur, 1994; see Coleman et al., 2000 and Ganong & Coleman, 2004 for reviews).2
Our charge from ACF was to focus on services that would benefit especially low-income stepfamilies, in keeping with the special concerns with economically disadvantaged groups in the federal HMI. Research indicates that low-income married couples are comparatively more vulnerable to marital dissolution and that their children are at risk for negative outcomes. The combination of economic strain and stepfamily structure establish low-income stepcouples as an important target population for marriage education.
The goals of this project were to review research literature on the demography, needs, and programs for stepfamilies; to review selected existing programs to learn more about the needs of stepfamilies and identify issues related to program development; to develop a conceptual framework for programs based on the literature review and the program study; and to suggest directions for future research. Accordingly, in this first chapter, we summarize the basic demography of stepfamilies. In the second chapter, we present a conceptual framework to guide marriage education programs for low-income stepfamilies. In the third chapter, we summarize our observations of selected marriage education services for stepfamilies. In the last chapter, we present general themes related to stepfamilies and programs that serve them and offer our recommendations for research on stepfamilies and stepfamily marriage education programs.
Although stepfamilies are not generally defined by marriage,3 in this report, we use the term “stepfamily” to refer to stepfamilies containing only married couples. The term “stepfamily” came into use in an earlier era of family history, to describe how new spouses “stepped into” the role of parenting children, typically through marriage to a widow or widower.4
By way of background, this chapter provides a statistical portrait of stepfamilies. We first touch briefly on the changing circumstances underlying the creation of stepfamilies today. We then summarize the evidence on stepfamily prevalence.
The “New” Stepfamily
In earlier U.S. history (i.e., 1700s to late 1800s), stepfamilies were created primarily when parents remarried after the death of a spouse. Stepfamilies were fairly common, due to lower life expectancy for men and women and high rates of maternal mortality during childbirth (Coontz, 2002).
In the latter half of the 20th century, dramatic increases in life expectancy and divorce rates gave rise to a new type of stepfamily. The new stepfamily is characterized by substantially more complex relationships, as both children and adults must adjust not only to new relationships within the household (e.g., stepparent/stepchild, stepsibling) but also to new relationships with family members outside the household (e.g., former spouses/partners and children residing with former spouses/partners) (Bramlett & Mosher, 2002; Bumpass & Lu, 2000).
In addition, substantial increases in rates of non-marital childbearing in recent decades imply that increasing numbers of first marriages between adults with children from previous relationships are forming stepfamilies (Bumpass, Raley, & Sweet, 1995; Ganong & Coleman, 2004). Approximately 50% of cohabiting couples are stepcouples as well (Seltzer, 2000). In this report, we focus on married stepcouples.
Divorce and Remarriage
Although a slight downward trend has been observed in the divorce rate in recent years,5 indications are that approximately half of adults who marry will divorce, the majority in their first 10 years of marriage (Bramlett & Mosher, 2001; 2002). National survey results indicate that economic disadvantage—whether measured by family income, education, or neighborhood poverty—is associated with a higher risk of first marriage disruption. Disruption rates are also higher for African American than for White or Hispanic women. Current estimates for first marriages suggest that 48% of Whites’ marriages, 52% of Hispanics’ marriages, and 63% of African Americans’ marriages end in divorce within the first 20 years of marriage (Bramlett & Mosher, 2001). First marriage disruption rates are highest for women who marry young, have a child before marrying, are not working at the time of marriage, are less religious, and did not live with both of their parents throughout the period they were growing up (Bramlett & Mosher, 2002).
Most divorced individuals go on to remarry (approximately 75% [Furstenberg & Cherlin, 1991]), and the majority of these individuals bring children from the previous relationship (Coleman et al., 2000). Typically, these new unions form quickly: about half of all divorced individuals remarry within five years and three quarters remarry within 10 years (Bramlett, & Mosher, 2002; Furstenberg & Cherlin, 1991; Kreider & Fields, 2001). Having low income and living in a poor neighborhood are associated with a lower chance of remarriage, whereas education has little association with remarriage rates. Younger adults are more likely to remarry than older ones. Whites and Hispanics are more likely to remarry than African Americans.6 Men were found more likely to remarry and to do so more quickly than women (Kreider, 2005). Having children from a previous marriage lowers the probability of remarriage, but more so for women than for men (Bramlett & Mosher, 2002; Buckle, Gallup, & Rodd, 1996).
Remarriages dissolve at slightly higher and faster rates than first marriages (e.g., Bumpass et al., 1990; Bramlett & Mosher, 2002; Krieder, 2005). The cumulative probability of first marriage dissolution after 10 years of marriage is 33%, and the probability of second marriage dissolution after 10 years is 39% (Bramlett & Mosher, 2001).7 Divorce among remarrieds also tends to occur more quickly compared with first marriages. Fifteen percent of remarriages have ended after three years, whereas nearly 25% have ended after five years compared to 12% and 20%, respectively, for first marriages (Bramlett & Mosher, 2002). Data do not exist for the probability of divorce among first marriages that form stepfamilies.
Although findings from non-representative sample studies have been mixed (Ganong & Coleman, 2004), representative demographic studies indicate that remarriages where children from a prior relationship are present are at even greater risk of dissolution than those without children. After 10 years of remarriage, the probability of disruption is 32% for women with no children at remarriage. For women with children at remarriage, the probability is 40 to 44% (Bramlett & Mosher, 2002).
Another factor associated with the likelihood of re-divorce is race/ethnicity. The remarriages of African American women (as are the first marriages) are more likely to end in separation or divorce, compared to White or Hispanic women (Bramlett & Mosher, 2002). After 10 years of marriage, the probability ending a remarriage is 29% for Hispanic women, 39% for White women, and 48% for African American women (Bramlett & Mosher, 2002).
Trends in Stepfamily Prevalence
Although indications are that living in a stepfamily is an increasingly common experience, efforts to track trends at the population level have not improved over the years. Available statistics do not allow us to establish specific rates of increase or decrease in stepfamily prevalence. To support estimates of stepfamily prevalence, surveys must ascertain at a minimum whether any children in the household are the biological children of only one spouse. The most thorough approach is to determine every existing child’s relationship to both spouses, whether a household resident or not. An alternative approach is to include only a summary question asking whether any existing children (i.e., resident or nonresident) are the biological children of only one spouse. It is also important to utilize similar survey questions across time in order to identify trends.
Unfortunately, very few national surveys include questions to accurately identify stepfamilies8 and the richest of them are somewhat older surveys (e.g., National Survey of Families and Households [NSFH], a longitudinal survey with waves in 1987-88, 1992-94, and 2001-2002) or cover specialized samples (e.g., Fragile Families and Child Wellbeing Survey, covering new parents in large cities, with an emphasis on unwed parents).
The paucity of recent population estimates makes it difficult to establish trends in stepfamily prevalence with precision. Earlier research projected increases in living in stepfamilies for children born in the 1980s, compared with those born in the 1970s (Glick, 1989), estimating that 33% of children born in the 1980s would spend some time in their growing years in a stepfamily, up from 14% born in the 1970s. New analyses conducted with the 1987-1988 NSFH data indicate that 31% of married adults of child-bearing age grew up in a stepfamily (see Appendix A).
More recent evidence on cohort trends in divorce and remarriage suggests that stepfamily prevalence is likely to have increased somewhat in the last decade or so, albeit at a slower rate. For example, Amato, Johnson, Booth, and Rogers (2003) find that the fraction of U.S. married couples in second or later marriages increased from 20 to 29% between 1980 and 2000. Other evidence suggests that although first marriage disruption rates have declined across recent cohorts, dissolution rates for second marriages have trended slightly upwards (Kreider, 2005). These slight increases in remarriage and re-divorce experience, however, still do not provide specific information on trends for stepfamily prevalence.
Information is even more limited for the low-income subgroups of interest to healthy marriage initiatives. There are, however, a few hints that the prevalence of stepfamilies may have continued to increase among more disadvantaged groups. For example, Raley and Bumpass (2003) found that divorce rates increased among married women without a high school diploma while remaining constant among better-educated women first marrying between 1980 and 1994.
It seems possible also that the numbers of first marriages involving children from previous (unmarried) partners increased more rapidly among more disadvantaged groups and minority groups. From 1960 to 1990, the fraction of women having a first child prior to marriage increased dramatically among women with the lowest education levels and among African American women (Ellwood & Jencks, 2001, Tables 5-10). Furthermore, a high fraction of these pre-marital births occurred more than three years prior to marriage, implying that many may be children by former partners.
The Prevalence of Stepfamilies
In this section we note some of the older estimates of stepfamily experience, including new estimates of stepfamily prevalence from the 1987-88 NSFH. A more detailed discussion of these prevalence estimates is found in Appendix A. We then present information from a recent sample of married couples.
Using data from the 1980s and the 1990 Census, Norton and Miller (1992) estimated that 5 million households in the U.S. (approximately 1 in 10) were stepfamily households. Data gathered in the late 1980s showed that 46% of all marriages in a year were remarriages for one or both partners (National Center for Health Statistics, 1993; U.S. Census Bureau, 2000, Table 145; Wilson & Clark, 1992).
Drawing on the earlier and more representative NSFH data, among all U.S. married couples of childbearing age (with a wife under age 35) in 1987-88, just over 1 in 10 of these families (11%) were residential stepfamilies. Excluding childless couples, the percentage was slightly higher (12%). In the large majority of stepfamilies (8 of the 11%), only the mother had children from another partner. A similar fraction (9%) of the population had non-resident minor children by former partners. In most (7%) of these instances, only the husband had non-resident children.
More comprehensive measures of stepfamily experience in 1987-1988 from the NSFH indicate that 17% of all married couples of childbearing age had at least one spouse who had had a child with a former partner (resident or nonresident). Again, we note that couples in which the stepchildren were over the age of 18 were not identified.
Recent data from the “Baseline Survey of Attitudes, Beliefs, and Demographics Relating to Marriage and Family Formation” conducted in Florida as part of that state’s healthy marriage initiative (Karney et al., 2003) showed that 18.3% of all Florida households contain a married couple with at least one residential stepchild (Karney et al., 2003, calculated from statistics in Table 5).
In this study, telephone interviews were conducted with a representative sample of 4,508 Florida residents over the age of 18, including oversamples of blacks, Hispanics, low-income households, individuals receiving Temporary Assistance for Needy Families (TANF), and residents of the four major metropolitan areas in Florida (Jacksonville, Orlando, Miami, and Tampa). In addition, samples of 500 individuals each were collected from randomly selected residents of California, Texas, and New York. Data were weighted in order to achieve the study’s goals of making reliable inferences and ensuring a representative sample of the population from which it was drawn.
Among married couple households with children, 40% answered “yes” to either the question, “Do you have a child from a previous relationship?” or “Does your spouse have a child from a previous relationship?” (Karney et al., 2003, calculated from statistics presented in Table 5). Among low-income married couples (below 200% poverty level) with children, this corresponding proportion is 49%, compared to 37% for higher income (greater than 400% of poverty). Among White and Hispanic married couple households with children, 39% and 36%, respectively, include at least one spouse who has one or more (resident or non-resident) children from a previous relationship. Among African-American married couple households with children, 55% report a stepparent-stepchild relationship for at least one partner. These questions appear to capture stepparent-stepchild relationships regardless of residence of the child or age of the child.
We emphasize that the Florida population differs from the U.S. population on ethnic minority proportions, and therefore present this information without inferences for the U.S. population proportions. Variations on the percentage of married couples with children who have at least one stepchild in the three other states sampled were 33% for California, 28% for New York, and 37% for Texas (Karney et al., 2003, Table 28).
Another method of examining prevalence of the stepfamily experience is to focus on the child’s experiences. Fields (2001) used the 1996 SIPP to develop estimates of the proportion of children under 18 currently living with stepparents. His statistics (which include both married and cohabiting couples) show that 7% of all children lived with a stepparent in 1996. There was little variation between Hispanic (5%), African American (6%), and non-Hispanic White (8%) children (Fields, 2001, Table 1). Fields also provides estimates of the fraction of children living in any family in which either a stepparent, stepsibling, or half-sibling was present. This expansive definition of stepfamilies included families with two unmarried parents, as well as single-parent families containing a stepsibling or half-sibling from a former partner. Results showed that 17% of all children, 20% of African American children, and 15% of Hispanic, and 16% of non-Hispanic White children lived in stepfamilies, so defined (Fields, 2001; Table 4).
Conclusions
Although existing data do not allow for the identification of specific rates of increase or decrease of stepfamily proportions in the population, we have indications that the stepfamily experience is quite common. Increased life expectancy, divorce rates, nonmarital birth rates, and multiple- partner fertility combine to create a large number of couples in complex family systems.
| Table of Contents | Previous | Next |