What Do "I Do"s Do? Potential Benefits of Marriage for Cohabiting Couples with Children

Gregory Acs, Sandi Nelson

No. B-59 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.

Note: This report is available in its entirety in the Portable Document Format (PDF).

---

Beginning with the 1996 federal welfare reform bill, which included promoting marriage as a goal, and continuing with President G.W. Bush's declaration that, "my administration will give unprecedented support to strengthening marriages," using government policies to actively promote marriage as a way to reduce poverty and improve the well-being of children and families has gained currency (Ooms 2002). Among the most logical targets for marriage promotion activities are cohabiting couples, especially those with children. Unlike unattached single parents, cohabitors already demonstrate a certain level of commitment to one another, and they and their children may benefit from the material and psychosocial advantages that married couple families accrue.

Previous research demonstrates that children living with cohabitors fare worse than children living with married parents, and in some cases no better than children living with single parents, on a host of outcome measures (Acs and Nelson 2002, 2003; Nelson, Clark, and Acs 2001; Manning and Brown 2003; Manning and Lichter 1996). Specifically, compared with children living with their married biological or adoptive parents, children living with cohabitors are more likely to experience material hardships such as poverty, food insecurity, and housing insecurity; more likely to exhibit behavioral problems or problems in school; and less likely to have positive interactions with their parents (such as being read to frequently).

No one would suggest that the differences in well-being between married and cohabiting families will simply vanish if cohabiting couples were to marry. Clearly some differences in well-being reflect differences in the characteristics (such as age or education) of those who choose to cohabit rather than marry. But advocates of marriage promotion argue that marriage has intrinsic benefits that would accrue to couples that marry. This brief assesses the potential size of these intrinsic benefits. To do this, we examine how much of the difference in well-being between children in cohabiting and married couple families can be explained by differences in the characteristics of these families. Any remaining differences may be the result of the intrinsic benefits of marriage. As such, we establish an upper bound of sorts for the potential benefits of promoting marriage among cohabiting families.

We find that differences in characteristics explain 50 to 80 percent of the differences in child well-being between cohabiting and married families. Thus, if programs could successfully promote marriage among cohabiting couples with children, we would expect modest but non-negligible improvements in child well-being.

Why Does Marriage Matter?

In the short run, getting married won't change the characteristics of cohabiting couples, and there are several well-documented differences between cohabiting and married couples with children. Only about half of children in cohabiting families live with both their biological or adoptive parents; the other half live with one parent and that parent's current partner who is unrelated to the child. In contrast, about eight of nine children in married families live with their two parents and only one of nine live in stepparent families (Acs and Nelson 2003). Research shows that even in married families, children fare worse in stepparent families (McLanahan and Sandefur 1994). The disproportionate share of children in cohabiting families living with one adult who is not a biological parent likely lowers their average outcomes.

There are other important differences as well. Manning and Brown (2003) show that compared with children in married couple families, those living in cohabiting families are more likely to be nonwhite and have less educated parents. Further, Fields and Casper (2001) find that couples in which women are more educated and earn appreciably more than their male partners are more common among unmarried couples than married ones. Unmarried couples are also more likely to be interracial than their married counterparts are.

Beyond any observable differences, well-being may be higher in married couple families than in cohabiting families because there are intrinsic benefits to marriage. Waite and Gallagher (2000) note that married couples enjoy an advantage in emotional health over cohabiting couples. Not only may this emotional health contribute to better parent-child interactions in married couple families, but it may also allow married men and women to be more productive at work, increasing their material well-being. Lerman (2002) suggests that marriages are more stable than cohabiting unions, making it easier for married couples to make long-term investments in their skills and assets. Further, extended family networks may be more inclined to offer material and in-kind assistance to married couples than cohabitors. Capturing these potential intrinsic benefits to marriage is the goal of marriage promotion.

Data and Definitions

The 2002 round of the National Survey of America's Families (NSAF) is an ideal data source for this analysis.¹ In addition to providing considerable detail on children's living arrangements, the NSAF contains a rich set of child and adult well-being measures. We focus on four different living arrangements for children: (1) children living with two biological or two adoptive parents who are married (married parent families); (2) children living with a biological parent who is married to either a stepparent or an adoptive parent (married stepparent families); (3) children living with two biological or two adoptive parents who are not married to each other (cohabiting parent families); and (4) children living with a biological or adoptive parent and that parent's boy/girlfriend, who is not a parent of the child (cohabiting stepparent families). Because research indicates that married stepparent families are very different from married parent families, we compare cohabiting parent families with married parent families and cohabiting stepparent families with married stepparent families. Indeed, these comparisons are logical because if cohabiting parents were to marry they would form a married parent family, while cohabiting stepparents would form a married stepparent family.

How Do Cohabiting and Married Families Differ?

Comparing Cohabiting and Married Parents. The characteristics of cohabiting and married parent families differ markedly (table 1). Cohabiting fathers are less likely to work than married fathers (82.2 percent versus 92.8 percent), and they are less likely to have worked full-time in the previous year (69.9 percent versus 85.7 percent). They are also far more likely to be high school dropouts (34.7 percent versus 12.0 percent) and to be under age 25 (16.8 percent versus 2.1 percent). Mothers in cohabiting parent families also differ significantly from their counterparts in married parent families. Cohabiting mothers are considerably more likely to be high school dropouts than married mothers (34.9 percent versus 10.5 percent) and more likely to be under age 25 (24.7 percent versus 4.0 percent). Although cohabiting mothers are less likely to be currently employed than married mothers (53.3 percent versus 60.1 percent), the two groups are equally likely to have worked and to have worked full-time during the prior year.

TABLE 1. Characteristics of Married and Cohabiting Couples with Children (percent, except where noted)
	Biological Parents		Stepparents
	Married	Cohabiting	Married	Cohabiting
Father
Currently employed	92.8	82.2*	88.8	82.4‡
Work status last year
Full time/full year	85.7	69.9*	79.9	73.8‡
Part time or part year	10.9	21.8*	13.5	19.2‡
No work last year	3.4	8.2*	6.6	7.0
Education
No high school diploma or GED	12.0	34.7*	11.9	25.4†‡
High school diploma or GED	31.8	46.9*	47.8	49.8
More than high school diploma or GED	56.2	18.4*	40.3	24.8†‡
Age
Less than 25	2.1	16.8*	3.3	9.1†‡
25 to 44	72.8	73.3	76.5	76.6
Over 44	25.1	9.9*	20.2	14.3†‡
Mean age (years)	39.4	32.8	38.2	35.3
Mother
Currently employed	60.1	53.3*	65.5	66.9†
Work status last year
Full time/full year	32.9	32.5	47.9	48.3†
Part time or part year	33.8	33.0	29.0	34.4
No work last year	33.4	34.5	23.1	17.3†‡
Education
No high school diploma or GED	10.5	34.9*	9.8	24.2†‡
High school diploma or GED	29.2	42.0*	45.2	44.0
More than high school diploma or GED	60.3	23.2*	45.1	31.8†‡
Age
Less than 25	4.0	24.7*	4.3	10.8†‡
25 to 44	80.6	71.8*	86.3	81.4†‡
Over 44	15.4	3.6*	9.5	7.8†
Mean age (years)	37.0	29.9	35.5	33.7
Race/ethnicity of parents
Both white, non-Hispanic	68.6	32.5*	63.7	54.4†‡
Both black, non-Hispanic	6.1	14.6*	12.0	14.4
Both Hispanic	12.5	36.1*	7.8	14.1†‡
Both Asian/Pacific Islander	4.4	1.8*	1.3	0.6
Both American Indian	0.2	0.8	0.2	0.3
Mixed race/ethnicity	8.3	14.3*	15.0	16.2
Unweighted sample size	19,970	1,139	2,739	1,033
Source: 2002 National Survey of America's Families (NSAF); data weighted using 2002 NSAF child weights. * Difference between cohabiting biological parents and married biological parents is statistically significant at the 90% confidence level. † Difference between cohabiting biological parents and cohabiting stepparents is statistically significant at the 90% confidence level. ‡ Difference between cohabiting stepparents and married stepparents is statistically significant at the 90% confidence level.

There are also notable differences in race and ethnicity. Table 1 shows that only 8.3 percent of children living with their married parents have parents from different racial or ethnic backgrounds; in contrast, 14.3 percent of children living with cohabiting parents have parents with mixed backgrounds. It is also interesting that children living with their married parents are more than twice as likely to have two white, non-Hispanic parents than those living with cohabiting parents (68.6 versus 32.5 percent).

Comparing Cohabiting and Married Stepparents. Table 1 shows that "fathers" in cohabiting stepparent families are less likely to work, have lower levels of education, and are younger than their counterparts in married stepfamilies. "Mothers" in cohabiting stepparent families are just as likely to work as those in married stepfamilies, but they are on average less educated and younger. There is little difference in the shares of children living with stepparents of different racial/ethnic backgrounds (15.0 percent versus 16.2 percent). Children with married stepparents are slightly more likely to have white, non-Hispanic parents than those living with cohabiting stepparents (63.7 percent versus 54.4 percent). This difference is appreciably smaller than the one we find for children living with their biological or adoptive parents.

Assessing the Effects on Children's Well-Being

Differences in the characteristics of married and cohabiting couples likely account for some of the differences in the well-being of children in these types of families. The well-being measures considered here include the share of children who live in poor families, the share who live in low-income families (those with incomes below 200 percent of the federal poverty level, or FPL), the share who live in families that experienced food insecurity, the share who live in families in which the responding parent is highly aggravated, and the share who live in families in which the responding parent is in poor mental health.²

To assess how much of the differences in well-being can be attributed to differences in characteristics, we completed a series of simulations showing how children in cohabiting families would fare if their parents were to marry. The simulations are based on a series of regressions that capture the relationships between parents' characteristics and the well-being outcomes of children in married couple families.³ The regressions include measures of mother's and father's work effort, education, race/ethnicity, the number of children, the presence of young children, and whether the couple has been together for more than a year. The simulations then use the estimates from these regressions to predict outcomes for children in cohabiting couple families if the characteristics of cohabiting couple families do not change upon marriage. Separate simulations are run for married parent and married stepparent families for each outcome. The results are then combined so we can discuss the average potential benefits of marriage for all children in cohabiting families.

If the only differences between children living with married and cohabiting couples are captured by the measured characteristics, then the simulated outcomes will equal the actual outcomes for children in cohabiting families. However, if there are unobserved or unmeasured differences between the two groups, the simulated outcomes will differ from the actual outcomes. The difference between the actual and simulated outcomes for cohabitors is the result of both any intrinsic benefits of marriage cohabiting couples would gain if they were to marry and any unobserved differences between the groups not adequately captured in the regression model. Thus, this comparison defines an upper bound on the potential benefits of marriage promotion.

Table 2 shows that if all cohabitors were to marry, then the poverty rate for children in these families would fall from 19.5 to 16.1 percent. This is still well above the 7.7 percent poverty rate for children with married parents. Similarly, if all cohabitors were to marry, the share of children in low-income cohabiting families would fall from 50.8 to 43.4 percent but remain considerably above the 25.7 percent level for children with married parents. The share of children in cohabiting families experiencing food-related hardships would decline from 46.3 to 39.4 percent if their parents were to marry; again, this is well above the 21.3 percent food insecurity rate experienced by children living with married parents. If cohabiting couples were to marry, the probability that their children would have a highly aggravated parent would fall from 11.3 to 10.6 percent and the probability of having a parent in poor mental health would fall from 24.9 to 19.2 percent. These are higher than the 7.9 percent of children with married parents who have a highly aggravated parent and the 12.0 percent who have a parent in poor mental health.

TABLE 2. Changes in the Well-Being of Children in Cohabiting Families if Couple Were to Marry (percent)
	Baseline		Simulated rates if cohabitors marry
	Cohabiting	Married
Poverty	19.5	7.7	16.1
Low-income status	50.8	25.7	43.4
Food insecurity	46.3	21.3	39.4
Highly aggravated parent	11.3	7.9	10.6
Parent in poor mental health	24.9	12.0	19.2
Source: Authors' calculations from the 2002 National Survey of America's Families (NSAF); data weighted using 2002 NSAF child weights.

Figure 1 illustrates the extent that differences in characteristics account for differences in well-being. Over 70 percent of the differences in poverty, low-income status, food insecurity, and parental aggravation as well as over 50 percent of the difference in poor parental mental health between children with cohabiting and married parents can be explained by differences in the characteristics of couples that choose to marry. The remaining differences can be attributed to unmeasured differences and the intrinsic benefits of marriage.

Recall that these simulations assume that there are no changes in cohabitors' characteristics after marriage. Some would argue that cohabiting fathers may increase their work effort after marriage and that this would be a benefit of marriage. In another simulation, we increase the average work of cohabiting fathers to the level observed among married fathers and find that the simulated poverty rate would fall from 16.1 to 14.1 percent, the low-income rate from 43.4 to 41.0 percent, and the food insecurity rate from 39.4 to 37.6 percent. Even after marriage and increases in male work effort, children living with cohabiting couples would still be more than twice as likely to experience poverty, low-income status, and food hardships than children living with married couples today.

Discussion

Study after study consistently documents that, on average, children living with their married biological or adoptive parents fare better on a host of indicators than children in any other living arrangement. As such, the idea of promoting marriage to improve child well-being has currency. Cohabiting couples with children are likely targets for marriage promotion activities because they are already in a relationship. However, it is unrealistic to expect that cohabiting couples will achieve the outcomes observed for currently married couples simply by getting married.

The lion's share of the differences in poverty, low-income status, food insecurity, and high parent aggravation between children living with married and cohabiting couples can be attributed to measurable differences in the characteristics of cohabiting and married couple families. These differences in characteristics also account for more than half of the difference between the two groups in poor parental mental health. By and large, differences will persist, at least in the short run, even if cohabitors marry. As such, the benefits of marriage promotion are clearly limited. Nevertheless, even after differences in characteristics are taken into account, programs that can successfully encourage marriage among cohabiting couples with children could improve child and adult well-being.

The regression-based simulations rely on several strong assumptions about the relative importance of unmeasured differences between cohabiting and married families and the relative stability of the characteristics of cohabitors in the wake of marriage. Overall, these simulations likely overstate the potential benefits of marriage promotion, but they do provide a more reasonable upper bound for these effects than the outcomes for those currently living in married families.

With these upper-bound estimates, we can move beyond the reasonable theoretical arguments for marriage promotion and begin a more informed inquiry into whether marriage rates can be increased effectively, whether the potential benefits of marriage promotion exceed the costs, and how effectively funds devoted to marriage promotion address child well-being, compared with funding other activities such as investments in child development or programs to reduce nonmarital childbearing.

---

Notes

1. When weighted, the NSAF is nationally representative of the civilian, noninstitutionalized population under age 65, with data on almost 40,000 households in 2002. For more information on the NSAF, see Dean Brick et al. (1999).

2. Questions about child well-being are directed to the Most Knowledgeable Adult (MKA) about the children in the family. We restrict our analysis to families in which the MKA is the child's parent, which is usually the case. Questions about parental aggravation and mental health are asked of the MKA—usually the mother. Our unit of analysis is always the child.

Unlike the official poverty rate, the poverty and low-income rates for children presented here take the income and needs of all family members into account, including all persons related by blood, marriage, adoption, or through a cohabiting relationship. Low-income is defined as family income below 200 percent of FPL adjusted for family size and poverty is defined as family income below 100 percent of FPL. Note also that low-income and poverty status is based on income from the previous calendar year.

Food insecurity is defined as whether the respondent or anyone in the family experienced at least one of three food-related concerns in the past 12 months. For a complete description of this indicator, see Zedlewski and Brauner (1999).

An index of parent aggravation was created by summing the responses to four items asked of the MKA: how often in the past month the MKA felt the child was much harder to care for than most, felt the child did things that really bothered the respondent a lot, felt he or she was giving up more of his or her life to meet the child's needs than he or she ever expected, and felt angry with the child. For a complete description of this indicator, see Ehrle and Moore (1999).

The respondent mental health scale is derived by summing the responses to five items that ask how often in the past month the MKA had been a very nervous person, felt calm or peaceful, felt downhearted and blue, had been a happy person, and felt so down in the dumps that nothing could cheer him or her up. For a complete description of this indicator, see Ehrle and Moore (1999).

3. A complete discussion of the simulation strategy as well as full regression results appear in Acs and Nelson (2004). Note that the overall explanatory power of the regressions on parental aggravation is small, with R2s ranging from 0.016 to 0.033.

References

Acs, Gregory, and Sandi Nelson. 2002. "The Kids Are Alright? Children's Well-Being and the Rise in Cohabitation." Washington, DC: The Urban Institute. Assessing the New Federalism Policy Brief B-48.

———. 2003. "Changes in Family Structure and Child Well-Being: Evidence from the 2002 National Survey of America's Families." Paper presented at the National Poverty Center Conference on Marriage and Family Formation among Low-Income Couples, Washington, D.C., Sept. 4-5.

———. 2004. Should We Get Married in the Morning? A Profile of Cohabiting Couples with Children. Washington, DC: The Urban Institute. Assessing the New Federalism Discussion Paper 04-01.

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, DC: The Urban Institute. National Survey of America's Families Methodology Report No. 1.

Ehrle, Jennifer, and Kristin Anderson Moore. 1999. 1997 NSAF Benchmarking Measures of Child and Family Well-Being. Washington, DC: The Urban Institute. National Survey of America's Families Methodology Report No. 6.

Fields, Jason, and Lynne Casper. 2001. "America's Families and Living Arrangements." Current Population Reports P20-537. Suitland, MD: U.S. Census Bureau.

Lerman, Robert. 2002. "Marriage and the Economic Well-Being of Families with Children: A Review of the Literature." Report prepared for the U.S. Department of Health and Human Services (DHHS), Office of the Assistant Secretary for Planning and Evaluation. Washington, DC: DHHS.

Manning, Wendy, and Susan Brown. 2003. "Children's Economic Well-Being in Cohabiting Parent Families: An Update and Extension." Mimeo. Bowling Green, OH: Center for Family and Demographic Research, Bowling Green State University.

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 Sandefur. 1994. Growing Up with a Single Parent: What Hurts, What Helps. Cambridge: Harvard University Press.

Nelson, Sandi, Rebecca Clark, and Gregory Acs. 2001. "Beyond the Two-Parent Family: How Teenagers Fare in Cohabiting Couple and Blended Families." Washington, DC: The Urban Institute. Assessing the New Federalism Policy Brief B-31.

Ooms, Theodora. 2002. "Marriage and Government: Strange Bedfellows?" Marriage and Couples Policy Brief No. 1. Washington, DC: Center for Law and Social Policy.

Waite, Linda, and Maggie Gallagher. 2000. The Case for Marriage. New York: Doubleday.

Zedlewski, Sheila, and Sarah Brauner. 1999. Declines in Food Stamp and Welfare Participation: Is There a Connection? Washington, DC: The Urban Institute. Assessing the New Federalism Discussion Paper 99-13.

---

About the Authors

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.

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.

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 Assessing the New Federalism project is currently supported by The Annie E. Casey Foundation, The Robert Wood Johnson Foundation, the W. K. Kellogg Foundation, The John D. and Catherine T. MacArthur Foundation, and The Ford Foundation.

---

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.

Usage, posting and reprint of materials on the UI web site:

Most publications may be downloaded free of charge from the web site in PDF format. This information may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required.

Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact paffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.