Number B-31 in Series, "New Federalism: National Survey of America's Families"
The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.
Between 1995 and 1998, the number of unmarried couples with children increased by 15 percent, from 1.3 million to 1.5 million.1 Although this rapid rise in cohabitation has been well documented, little is known about how children fare in cohabiting families compared with children in traditional two-parent, stepparent, and single-parent households (Bumpass and Raley 1995, Manning and Lichter 1996, Smock and Manning 1997).2 Previous research on the effects of living arrangements on child outcomes has shown that, in general, children living with their married biological parents exhibit the lowest rates of behavioral problems and perform better at school, while children in single-parent families tend to fare worse across these outcomes.
In this brief, we use data from the 1997 National Survey of America’s Families (NSAF)3 to examine whether children living in cohabiting families with their mothers and their mothers’ boyfriends (who are not related to the children) are any better or worse off than children living with just a single mother. We also compare outcomes for children living in families in which their mothers have married their boyfriends (forming "blended" families) with outcomes for children living with a single mother as well as with outcomes for children living with married biological parents.4 Specifically, we focus on behavioral problems among teenagers (12- to 17-year-olds). We begin by documenting the living arrangements of teenagers and then use a multivariate, regression-based model to examine how living arrangements affect teenagers’ behavioral outcomes.
Living Arrangements of Teenagers
Figure 1 shows the living arrangements of teenagers by race/ethnic group.5 Most white teenagers (62 percent) live with married biological parents; in contrast, 49 percent of Hispanic teenagers and only 25 percent of black teenagers live in such families. Whites are also the most likely to live in blended families: 15 percent of them do, compared with 9 and 11 percent of blacks and Hispanics, respectively. For black teenagers, the most common living arrangement is with a single parent; a significant share of Hispanic teenagers live in single-parent families (35 percent), while about one in five white teenagers live in such families.
The percentage of teenagers living in cohabiting families is relatively low (3 to 4 percent) and does not vary dramatically across race/ethnic groups; however, there are some important differences in the types of cohabiting families. About one-quarter of black teenagers and nearly one-third of Hispanic teenagers in cohabiting families live with both biological parents—that is, they live with both their mother and father, but their parents are not married. In contrast, virtually all white teenagers in cohabiting families live with one biological parent and that parent’s boyfriend or girlfriend.
How Teenagers Fare in Blended and Cohabiting Families
Next we examine whether differences in living arrangements affect teenagers’ behavioral and emotional well-being. We specifically focus on three outcomes: (1) whether the teenager has emotional or behavioral problems;6 (2) whether the teenager exhibits a low level of school engagement;7 and (3) whether the teenager has been suspended or expelled from school in the past year.8, 9 We use a multivariate approach to isolate the impact of living arrangements on these outcomes by controlling for other important differences across teenagers, such as parental education and family income levels.10
In general, families with a husband or boyfriend present have higher incomes than families with a single mother. By taking income differences into account, we can detect any advantages (or disadvantages) conferred on teenagers through having a man present beyond the additional income he may provide. However, the additional income itself may provide an important benefit: If a boyfriend’s or step-father’s income lifts a teenager out of poverty, the teenager’s behavioral outcomes are likely to improve.
For ease of presentation and interpretation, we use our multivariate results to generate predicted probabilities for each outcome for each of four separate living arrangements: living with married biological parents, living in a blended family, living with cohabitors (a single mother and her boyfriend who is not related to the teenager),11 and living with a single mother.12 We conduct separate analyses for three race/ethnic groups (whites, blacks, and Hispanics) and present the predicted probabilities for each outcome separately (see table 1).
|TABLE 1. The Relationship between Living Arrangements and Three Behavioral Problem Indicators among 12- to 17-Year-Olds
||Predicted Percentage of Teenagers with Emotional and Behavioral Problems
||Predicted Percentage of Teenagers with Low Levels of School Engagement
||Predicted Percentage of Teenagers Suspended or Expelled from School in the Past Year
|White Teenagers Living with
|Married biological parents
|Blendedmother and step- or adoptive father
|Cohabitorsmother and boyfriend
|Hispanic Teenagers Living with
|Married biological parents
|Blendedmother and step- or adoptive father
|Cohabitorsmother and boyfriend
|Black Teenagers Living with
|Married biological parents
|Blendedmother and step- or adoptive father
|Cohabitorsmother and boyfriend
|Source: Urban Insitute calculations from logit estimates of determinants of emotional and behavioral problems, low school engagement, and suspensions and expulsions on data from the 1997 National Survey of America's Families.
Differences are significant at the 90 percent confidence level.
a. Statistically significant difference from teenagers living with married biological parents.
b. Statistically significant difference from teenagers living with a single mother.
Note: Predicted probabilites for teenagers with unmarried biological parents are not shown because of insufficient sample sizes.
Teenagers in Cohabiting Families
White and Hispanic teenagers living in cohabiting families fare worse, on average, than those living with single mothers. White teenagers living with cohabitors are significantly more likely to exhibit low school engagement than those living with a single mother (39.3 and 27.9 percent, respectively). Similarly, white teenagers living with cohabitors are more likely to be suspended or expelled from school than those living with a single mother (23.0 and 1.3 percent, respectively). In addition, white teenagers living with cohabitors are just as likely as those living with a single mother to have emotional and behavioral problems.
Similarly, for Hispanic teenagers, living in a cohabiting family significantly increases the likelihood of exhibiting problem behaviors relative to living with a single mother across all three outcome measures. Hispanic teenagers living with cohabitors have a 37.7 percent chance of experiencing emotional and behavioral problems, while those living with a single mother have only a 10.8 percent chance. The differences for school behavior are just as dramatic: Hispanic teenagers living with a single mother have a 38.3 percent likelihood of having low engagement with school and a 15.2 percent chance of being suspended or expelled; those living with cohabitors have a 66.2 and 42.0 percent likelihood, respectively.
The impact of cohabitation on black teenagers is less dramatic. Black teenagers living in cohabiting families have the same levels of emotional and behavioral problems and school engagement as black teenagers living with single mothers. However, they are more likely to have been suspended or expelled than those living with single mothers (56.4 and 30.6 percent, respectively). Overall, black teenagers with a cohabiting mother are no better off than those with a single mother; on some measures they are actually worse off.
Teenagers in Blended Families
Although cohabitation rates are increasing, far more teenagers live in blended families, and some may argue that outcomes for teenagers would improve if single mothers and cohabitors were to marry, forming blended families. Table 1 shows that living in a blended family is neither better nor worse than living in a single-mother family for white teenagers in terms of the behavioral outcomes we consider. Indeed, for teenagers in both blended and single-mother families, the chance of exhibiting behavioral or emotional problems and the chance of being expelled or suspended is about 1 in 10; the chance of having low school engagement is about 1 in 4. Thus, although outcomes for white teenagers in blended families are no better than those for white teenagers in single-mother families when other factors are held constant, blended arrangements do not have the same negative outcomes as those associated with cohabitation.
In these respects, Hispanic teenagers are similar to white teenagers. Hispanic teenagers living in blended families are equally likely to have emotional and behavioral problems and to exhibit low levels of school engagement as are teenagers living with single mothers, and they are more likely to be expelled or suspended from school (23.1 and 15.2 percent, respectively). Overall, living in a blended family rather than with a single mother is not associated with better outcomes for Hispanics.
In contrast, black teenagers living in blended families generally fare better than their counterparts living with single mothers: They are less likely to have emotional and behavioral problems (6.1 and 11.7 percent, respectively) and to have been suspended or expelled (18.9 and 30.6 percent, respectively). Black teenagers in blended and single-mother families have similar levels of low school engagement (36.1 and 32.5 percent, respectively). Overall, a black teenager, unlike a white or Hispanic teenager, is better off in a blended family than in a single-mother family.
Teenagers Living with Married Biological Parents
Not surprisingly, teenagers living with married biological parents are far less likely to exhibit behavioral problems than are those living with cohabitors and those living in blended families. Indeed, only 3.6 percent of white teenagers living with married biological parents have emotional and behavioral problems, 18.9 percent have low levels of school engagement, and 7.1 percent have been suspended or expelled. The predicted incidences of these problems for Hispanic teenagers in married biological parent families are similar to those for whites.
Compared with black teenagers who live in cohabiting families, those living with married biological parents are significantly less likely to experience emotional and behavioral problems (14.9 and 6.1 percent, respectively) and to have been suspended or expelled from school (56.4 and 10.4 percent, respectively); they are also less likely to exhibit low school engagement (37.1 and 25.7 percent, respectively), but the difference is not statistically significant. Black teenagers living with married biological parents also run a significantly lower risk of poor outcomes relative to those who live in blended families on two of the three measures, although they are equally likely to experience emotional and behavioral problems. Black teenagers in both blended and married biological parent families have a 6.1 percent chance of exhibiting emotional and behavioral problems.
The living arrangements of teenagers vary significantly by race and ethnicity, with white teenagers far more likely to live with married biological parents and black teenagers more likely to live in single-parent families. Further, these differences in living arrangements are associated with differences in teenagers’ behavior.
Specifically, our analysis of the relationship between teenagers’ living arrangements and their behavioral outcomes shows that living with a single mother and her boyfriend is no better than living with a single mother. In many cases (particularly for whites and Hispanics), it is significantly worse. The most favorable outcomes we observe are for teenagers living with their biological parents who are married to each other. Beyond that arrangement, for white and Hispanic teenagers, living in a blended or single-mother family is relatively interchangeable, and living with cohabitors is associated with the poorest behavioral outcomes. For black teenagers, however, living in a blended family is associated with better outcomes than living with either a single mother or with cohabitors.
Our analysis is subject to certain limitations. First, we find a correlation between cohabitation and poorer behavioral outcomes for teenagers. This may indicate that the teenager’s problems stem from the presence of a cohabitor; however, there are several other potential explanations. For example, a single mother with a troubled teenager may be more likely to seek out a partner than a single mother without a troubled teenager. Alternatively, unmeasured characteristics of the mother may affect both child outcomes and her marital/cohabitation status. It may also be the case that differences in outcomes for children living with a mother and her boyfriend compared with a mother and a stepfather are related to the quality of the mother–partner relationship.
Second, there may be different outcomes for teenagers living with cohabitors, depending on whether the cohabitors are both biological parents of the teenager or are a mother with her current boyfriend. Unfortunately, our analysis is constrained by the small number of cases of cohabiting biological parents, particularly when we separate the population by the race/ethnicity of the teenager. Consequently, we only present results for cohabiting families in which the male is the mother’s boyfriend, not the biological father of the teenager.
Third, our analysis is limited to only three outcomes: two addressing school problems and one dealing with emotional and behavioral problems.
Finally, this analysis examines the relationship between behavioral problems and living arrangements for 12- to 17-year-olds only. Younger children may benefit from having their biological father present even if he is not married to the child’s mother. Future analyses should address the relationship between behavioral problems and living in both types of cohabiting couple families for children of all ages.
The number of children living with cohabiting couples is increasing despite legislators’ efforts to encourage the formation and maintenance of married two-parent families through the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996. Our research suggests that cohabiting couple families are not simply an extension of traditional married biological or blended parent families. Instead, children in these types of families represent a unique group that is particularly at risk for behavioral problems, often more so than their counterparts in single-mother families. Policymakers and researchers should be mindful of these differences when formulating and examining policies that may influence family living arrangements.
1. Unmarried couples with at least one child under age 15. U.S. Bureau of the Census. 1999. "Table No. 68: Unmarried Couples, by Selected Characteristics: 1980 to 1998." In Statistical Abstract of the United States 1999. Washington, D.C.: U.S. Government Printing Office.
2. Cohabiting couple families comprise two distinct types of living arrangements: children living with unmarried biological parents and children living with one biological parent and that parent’s partner.
3. The 1997 NSAF provides information on a nationally representative sample of the civilian noninstitutionalized population under age 65 and their families. The survey contains information on more than 44,000 households. Information on the children in the NSAF was obtained from the parent or guardian in the household most knowledgeable about the child’s education and health care. For 72 percent of the children, this respondent—labeled the most knowledgeable adult (MKA)—was the mother. For more information on NSAF survey methods, see Dean Brick et al. (1999).
4. McLanahan and Sandefur (1994) examine the outcomes for children in stepfamilies. Whether stepchildren are more similar to their counterparts living with married biological parents or to those living with a single parent varies depending on the race/ethnicity and sex of the child and the specific outcome considered. Note that McLanahan and Sandefur define children in stepfamilies as those who live with only one biological parent who is married. For this analysis, we prefer the term "blended families" instead of stepfamilies, because we include biological parents married to stepparents and adoptive parents in our definition of blended families.
5. White and black teenagers who were identified as having Hispanic ethnicity were coded as Hispanics in this analysis.
6. For the emotional and behavioral problems indicator, the MKA was asked whether the child exhibited signs of external distress (not getting along with other kids, acting too young for his or her age, lying, or cheating) and internal distress (sadness, depression, or feelings of worthlessness) in the past month. The MKA was asked to decide whether the focal child exhibited the behavior often, sometimes, or never. The responses were coded on a three-point scale, with 1 for often, 2 for sometimes, and 3 for never. A composite measure of behavioral and emotional problems was derived from the responses. Responses were totaled to create a scale score ranging from 6 to 18. A child whose MKA responses totaled 12 points or fewer on the behavioral problems scale received a value of 1 for high levels of behavioral problems. Children whose score was greater than 12 points received a value of 0. This scale was originally developed from the Child Behavior Checklist for the National Health Interview Survey (NHIS) as an indicator of children’s mental health. See NSAF Methodology Report 6: 1997 Benchmarking Measures of Child and Family Well-Being.
7. To assess school engagement, the MKA was asked about the extent to which the child did schoolwork only when forced to, did just enough to get by, always did homework, and cared about doing well in school. The response categories were coded 1 for all of the time, 2 for most of the time, 3 for some of the time, and 4 for none of the time. The responses to these four questions were combined to generate a measure of school engagement. Responses were totaled to create a scale score ranging from 4 to 16. A child whose MKA responses totaled 10 points or fewer on the school engagement scale received a value of 1 for low school engagement. Children whose score was greater than 10 points received a value of 0. Jim Connell and Lisa J. Bridges at the Institute for Research and Reform in Education in California created the school engagement scale in 1996, which includes the four indicators used in the NSAF school engagement index. See NSAF Methodology Report 6: 1997 Benchmarking Measures of Child and Family Well-Being.
8. For suspensions and expulsions, the MKA was asked whether the child had been suspended (both in and out of school) or expelled from school in the past 12 months.
9. Moore et al. (2000) examine the incidence of several emotional and behavioral outcomes for children using both the 1997 and 1999 waves of the NSAF. They report that in 1997, 8.8 percent of teenagers had emotional and behavioral problems and 13.9 percent had been expelled or suspended from school; their findings for 1999 are similar.
10. Our universe includes 12- to 17-year-olds who live with their biological mother in one of five types of living arrangements: married biological parents, unmarried biological parents, blended parents (i.e., with a stepfather or adoptive father), mother cohabiting with a male who is unrelated to the child, and mother only. The universe is further restricted to teenagers whose MKA was either the biological mother or the mother’s spouse or partner. The variables included in the regression model include the child’s living arrangement (married biological parents, unmarried biological parents, blended parents, cohabiting mother, or single mother); the child’s gender; the MKA’s education level (less than a high school education, a high school degree or GED, or more than a high school education); and the child’s social family’s poverty level (less than 100%, 100%–200%, or more than 200% of the federal poverty level). Full results of the logit regression analysis are available upon request. Note that the family income includes the income of unmarried partners and relatives of unmarried partners. Because research suggests that unmarried partners do not necessarily share resources with their partner or their partner’s children, using this broad definition of family income may overstate the resources available to the child.
11. Because of insufficient sample sizes, the regression results for teenagers living with unmarried biological parents are not presented. Not only is this type of living arrangement rare for teenagers, it also represents a very select population because it is an extremely stable nonmarital arrangement.
12. Predicted probabilities are computed using the appropriate living arrangement variables and the mean characteristics for the sample for all the other (non-living-arrangement) variables. Unadjusted probabilities of teenagers’ behavioral outcomes by race/ethnicity compare favorably with adjusted probabilities and are available upon request.
Bumpass, Larry, and R. Kelly Raley. 1995. "Redefining Single-Parent Families: Cohabitation and Changing Family Reality." Demography 32 (February): 97–109.
Dean Brick, Pat, Genevieve Kenney, Robin McCullough-Harlin, Shruti Rajan, Fritz Scheuren, Kevin Wang, J. Michael Brick, and Pat Cunningham. 1999. 1997 NSAF Survey Methods and Data Reliability. Washington, D.C.: The Urban Institute. National Survey of America’s Families Methodology Report No. 1.
Ehrle, Jennifer, and Kristen Anderson Moore. 1999. Benchmarking Child and Family Well-Being Measures in the NSAF. Washington, D.C.: The Urban Institute. National Survey of America’s Families Methodology Report No. 6.
Manning, Wendy, and Daniel T. Lichter. 1996. "Parental Cohabitation and Children’s Economic Well-Being." Journal of Marriage and the Family 58 (November): 998–1010.
McLanahan, Sara, and Gary Sadefur. 1994. Growing Up with a Single Parent: What Hurts, What Helps. Cambridge, MA: Harvard University Press.
Moore, Kristen Anderson, Juliet L. Hatcher, Sharon Vandivere, and Brett V. Brown. 2000. Children’s Behavior and Well-Being: Findings from the National Survey of America’s Families. Washington, D.C.: The Urban Institute. Snapshots of America’s Families II.
Smock, Pamela J., and Wendy Manning. 1997. "Cohabiting Partners’ Economic Circumstances and Marriage." Demography 34 (August): 331–341.
U.S. Bureau of the Census. 1999. "Table No. 68: Unmarried Couples, by
Selected Characteristics: 1980-1998." In Statistical Abstract of the United States 1999. Washington, D.C.: U.S. Government Printing Office.
About the Authors
Sandi Nelson is a research associate in the Urban Institute’s Income and Benefits Policy Center, where she currently focuses on changes in the living arrangements of children and families. Her research interests include measuring the well-being of children and families, as well as the effects of welfare policy on family structure.
Rebecca L. Clark cowrote this brief when she was a senior researcher at the Urban Institute, where she did research on issues related to the well-being of children, including children living with relatives other than their parents and on federal expenditures on children. Dr. Clark’s other areas of interest included the impact of immigrants, especially unauthorized aliens, on the United States. She is now a program official at the Demographic and Behavioral Sciences Branch of the National Institute of Child Health and Human Development, where her portfolio includes race and ethnicity, immigration, migration and population distribution, population and the environment, and demographic methods.
Gregory Acs is a senior research associate in the Urban Institute's Income and Benefits Policy Center. His research focuses on social insurance, social welfare, and worker compensation. Currently, he is studying trends in family living arrangements as well as the status of families that have left welfare.
About the Series
This series presents findings from the 1997 and 1999 rounds of the National Survey of America's Families (NSAF). Information on more than 100,000 people was gathered in each round from more than 42,000 households with and without telephones that are representative of the nation as a whole and of 13 selected states (Alabama, California, Colorado, Florida, Massachusetts, Michigan, Minnesota, Mississippi, New Jersey, New York, Texas, Washington and Wisconsin). As in all surveys, the data are subject to sampling variability and other sources of error. Additional information on the NSAF can be obtained at http://newfederalism.urban.org.
The NSAF is part of Assessing the New Federalism,a multiyear project to monitor and assess the devolution of social programs from the federal to the state and local levels. Alan Weil is the project director. The project analyzes changes in income support, social services, and health programs. In collaboration with Child Trends, the project studies child and family well-being.
The project has received funding from The Annie E. Casey Foundation, the W.K. Kellogg Foundation, The Robert Wood Johnson Foundation, The Henry J. Kaiser Family Foundation, The Ford Foundation, The David and Lucile Packard Foundation, The John D. and Catherine T. MacArthur Foundation, the Charles Stewart Mott Foundation, The McKnight Foundation, The Commonwealth Fund, the Stuart Foundation, the Weingart Foundation, The Fund for New Jersey, The Lynde and Harry Bradley Foundation, the Joyce Foundation, and The Rockefeller Foundation.