Fragile Families, Welfare Reform, and Marriage

Irwin Garfinkel,
headshot of Irwin Garfinkel
Irwin Garfinkel Professor Emeritus - Columbia School of Social Work
Ronald B. Mincy, and Sara McLanahan
Sara McLanahan William S. Tod Professor of Sociology and Public Affairs - Princeton University

December 1, 2001

Promoting marriage and two-parent families is of great importance to policymakers. The benefits of marriage for adults (better health, greater longevity, and higher earnings) have been well documented, and the benefits of growing up with two biological parents (more education, greater marital stability, and better mental health) are widely acknowledged. Moreover, Congress and the Bush administration seem determined to make marriage a major issue in the welfare reform reauthorization debate.

Welfare rolls have dropped dramatically since 1996, and large proportions of welfare recipients have moved from dependency to work. At the same time, the proportion of births to unmarried mothers, after several decades of relentless increases, has remained constant at around 33 percent. In response, some policymakers have argued that more dollars from the Temporary Assistance for Needy Families (TANF) should be spent on programs for poor parents who marry. Others have argued that, rather than promoting marriage, TANF money should go towards making poor parents, especially fathers, more “marriageable” or better able to support themselves and their families.

This policy brief will assess “marriage” and “marriageability” strategies and discuss policies to promote both. Strategies for increasing father involvement and improving communication among parents who live apart will also be considered. Most of the analysis is based on data from the Fragile Families and Child Wellbeing Study being conducted at the Center for Research on Child Wellbeing at Princeton University and the Social Indicators Survey Center at Columbia University. The study is following a birth cohort of approximately 5,000 children born to unwed parents in large cities (populations of 200,000 or more) at the turn of the 21st century. Based on interviews with mothers and fathers, the four-year study provides extensive information about parents’ relationships, views on marriage, intentions to marry, and expectations about the role of unmarried fathers.

Unwed Parents: What We Know

One of the most striking findings from the Fragile Families Study thus far is the high rate of cohabitation among unmarried parents. At the time of birth, half of unmarried mothers are living with the fathers of their children. Another third are romantically involved with the fathers, but living apart in what are called “visiting relationships.” Eight percent of parents are “just friends” and 9 percent have “little or no contact.”

The majority of unwed parents are optimistic about their future together. Nearly three quarters of the mothers believe their chances of marrying the father of their child are “50-50” or better. Almost two thirds “agree” or “strongly agree” with the statement, “it is better for children if their parents are married.” There is also strong consensus among unmarried parents about what qualities are necessary for successful marriage. Roughly 90 percent of mothers rate “husband having a steady job” and “emotional maturity” as very important qualities for a successful marriage. In addition, 69 percent of mothers rate “wife having a steady job” as very important.

Most fathers are highly involved during the pregnancy and around the time of birth. According to the mothers surveyed, four out of five fathers provided some financial support during the pregnancy, 84 percent will have their name on the birth certificate, and 79 percent of the children will take the father’s surname. Most fathers say they want to help raise their child, and the overwhelming majority of mothers say they want the fathers to be involved.

At the time their child is born, the vast majority of unmarried parents are committed to each other and to their child. Most mothers and fathers have high hopes about their future together and most view marriage as a positive institution that benefits children. Clearly, these parents are likely to respond positively to programs and policies that promote marriage, which is good news for policymakers who favor this strategy.

Unfortunately, many unmarried parents are poorly equipped to support themselves and their children. Table 1 shows that although nearly all the fathers in the Fragile Families Study worked in the past year, almost three out of ten were out of work in the week before their baby was born. In addition, the human capital of both parents is low: 37 percent of mothers and 34 percent of fathers lack a high school degree, and less than a third of parents have any education beyond high school. These findings are consistent with those in the 1995 Report to Congress on Non-Marital Childbearing by the U.S. Department of Health and Human Services, and with other research on unwed parents and non-custodial fathers. Elaine Sorensen of the Urban Institute in Washington, D.C., for example, finds that poverty rates among non-custodial fathers may be as high as 25 percent, and Irwin Garfinkel and his colleagues at Columbia University report that 20 percent of non-custodial fathers earn less than $6,000 annually. Studies of teen parents and mothers on welfare paint an even bleaker picture of the capabilities of unwed parents.

The Fragile Families Study also shows that although a majority of unmarried parents are in fairly good health, some engage in “unhealthy” behaviors. Three percent of mothers and 5 percent of fathers report that a drug or alcohol problem interfered with their work or personal relationships in the past year. Drug and alcohol problems are likely to be underreported, so we should assume that the true prevalence of substance abuse is higher. Approximately 38 percent of unwed fathers have been incarcerated, which suggests that a substantial proportion of these men have had lifestyles that are potentially harmful to their children. As shown by Cynthia Miller and Virginia Knox in their recent review, evaluations of the Parents’ Fair Share program, a large scale demonstration program that provided services to unwed fathers, found that non-custodial fathers who were delinquent in their child support payments faced severe employment barriers, including criminal records and poor health.

Many policymakers and advocates worry that promoting marriage will increase domestic violence. The Fragile Families Study data suggest that violence is rare among new unwed parents. Only 4 percent of mothers and 14 percent of fathers report being hit or slapped by their partner during the past year. While these figures are reassuring, they should be viewed with caution. Mothers are likely to underreport the incidence of violence, especially if they are still romantically involved with the fathers. Indeed, reported rates of violence are much higher among mothers who are no longer in contact with the father of their child.

Finally, 11 percent of mothers and 7 percent of fathers report that the other parent is “never fair or willing to compromise.” At worst, such behavior may be a precursor to physical violence. At best, it signals a relationship in trouble.

Estimates based on preliminary data from twenty cities in the Fragile Families Study indicate that less than 20 percent of the new parents had married by the time their child was 12-18 months old. Table 2 shows that of the parents who were living together at birth, 12 percent had married and an additional 63 percent were still cohabiting. Thus, 75 percent of the children who were living with both biological parents at birth were still living with both parents nearly a year and a half later. Research based on the National Longitudinal Survey of Youth suggests that about half of unmarried parents who are cohabiting at birth are still living together after six years.

The findings discussed above underscore the precarious socioeconomic circumstances of unwed parents and the barriers to marriage that many of them face. They also support the argument that a substantial proportion of unwed parents are not ready for marriage because of low employment skills, risky behavior such as drug use, and poor relationship skills (defined as the ability to maintain a nonviolent, mutually supportive relationship).

Policy Implications

Marriage and cohabitation among fragile families can be encouraged by increasing the capabilities of parents, reducing marriage and cohabitation penalties in current spending programs and tax policies, and making child support enforcement more suitable to the circumstances of unmarried parents who live together. No matter how successful such policies are, however, a substantial proportion of unwed parents will live apart. The dilemma for policymakers is how to address the needs of these parents and their children without undermining marriage. Based on the findings in the Fragile Families Study and broader research on the effects of public policy on families, a reform agenda for promoting marriage, marriageability, father involvement, and the security of single-parent families should be considered as part of the welfare reform reauthorization debate.


Services to strengthen fragile families should begin before or at birth when the overwhelming majority of unwed parents are still romantically involved and should offer services to mothers and fathers. The “magic moment” of birth may be particularly important for motivating fathers. The Parents’ Fair Share program, which had limited success with fathers, provided help too late-long after the romantic relationship between the mother and father had ended.

A promising model to build on is the home visiting nurse program pioneered in Elmira, New York, and Memphis, Tennessee, by David Olds and his colleagues at the University of Colorado in Denver. In the Olds program, now being implemented statewide in Oklahoma and at over 200 sites in 23 other states, nurses visit first-time mothers during pregnancy and for two years following birth. This program could be extended to include fathers as well as mothers and assessment of both parents on education, employment status, health and mental health issues (including substance abuse problems), and relationship problems (including domestic violence). When a problem is indicated, one or both parents would be offered appropriate help. In addition, the visiting nurses would inform parents of the other supports and services for which they are eligible, including health care, welfare, child support enforcement, the Earned Income Tax Credit, and child care programs.

TANF Reforms

Services alone are unlikely to substantially strengthen fragile families. To the extent that welfare policies or practices favor one-parent families over two-parent families, they discourage marriage and cohabitation and push biological fathers out of the picture. Although many state TANF programs appear to have reduced or eliminated restrictions for two-parent families, others still retain such restrictions. Similarly, many states and localities give preference to one-parent families in allocating scarce child care and housing subsidies. If the goal is to promote marriage and family stability, states should eliminate this kind of unfavorable treatment of two-parent families.

The absence of categorical restrictions, however, is still not sufficient to make welfare policy neutral with respect to family formation. Because welfare is income tested, it creates an incentive for fathers with earnings and mothers without earnings to live apart (or feign living apart). To reduce this disincentive, only a portion of a resident father’s earnings-say 50 percent-should be counted when determining a family’s eligibility and benefits for TANF. Although doing so will increase welfare costs and caseloads in the short run, the time limits and work requirements of the new TANF program would limit these extra costs, and the long-term gains in strengthening two-parent families will be considerable for families as well as society.

Child Support Reforms

Strict child support enforcement also reduces the disincentives to marriage and cohabitation in welfare policy by increasing the costs (for fathers) of living separately. However, if child support obligations are imposed on cohabiting fathers or are grossly inconsistent with their ability to pay, they may drive fathers away and discourage their involvement. If the parents reside together, they should be treated as a family by TANF, only a portion of the income of each parent should be counted in determining eligibility and benefits, and services should be provided to fathers as well as mothers.

Services for fathers, like those for mothers in TANF, should be geared primarily towards obtaining employment. In cases in which either the mother or father demonstrates the potential to benefit from further education and training, however, TANF should provide such support. Services for both mothers and fathers should also be directed at educating the parents about their mutual rights and responsibilities, including establishing the paternity of the father.

All unwed fathers, including those who live with the mother and child, should be required to establish paternity. Those who live with the mother, however, should not be required to pay child support. Both parents should be fully informed of the nonresident parent’s potential child support obligation in the event of a separation.

If the parents live apart, fathers should be required to pay child support, and enforcement should be strict. But the amount of the obligation should be proportional to fathers’ ability to pay. Poor fathers are routinely required to pay much higher proportions of their income than middle- and upper-income fathers, and many are required to pay unreasonable amounts of arrearages (past-due child support). These unrealistic arrearages arise because child support agencies and courts base these payments not on fathers’ actual earnings, but on their “presumptive” earnings; e.g., the minimum wage multiplied by full-time, full-year work or, if he is unemployed, how much he earned at some point in the past. Fathers are even required to pay back the mother’s welfare costs and, in some states, her Medicaid costs. Many fathers who become unemployed or incarcerated build up huge arrearages during these periods of unemployment. Such onerous child support obligations are rarely paid in full, and they can prompt fathers to avoid legitimate work where their wages are easily intercepted. Ultimately, they breed resentment among fathers and mothers toward the system and perhaps toward each other. Imprisonment for non-payment of support exacerbates this negative dynamic. Given what we know about the low earnings capacity of most unwed fathers, these practices are not likely to be effective and may even have unintended negative consequences.

If child support obligations were expressed as a flat percentage of the father’s income in every state, many of these problems would be reduced. Obligations would automatically decline when the father is unemployed or in jail and would automatically go up when his earnings rise.

Judi Bartfeld and Irwin Garfinkel of Columbia University find that support orders expressed as a percentage of income lead to substantially higher, not lower payments. States should reconsider their guidelines so that the child support obligations imposed on low-income, nonresident fathers can be no higher in percentage terms than those imposed on middle-income, nonresident fathers. Finally, through TANF or other workforce programs, unemployed fathers should be offered a job at the minimum wage and be required to pay a portion of their earnings in child support.

Most states now reduce TANF benefits by one dollar for each dollar of child support paid. This policy reduces the incentive for mothers to cooperate with the child support program and for fathers to pay child support. Counting only a portion of support in determining eligibility and benefits would increase cooperation and payments as well as the child’s standard of living. Congress should require or encourage states to ignore a substantial portion-say 50 percent-of child support payments in determining TANF eligibility and payments.

Finally, creating a publicly-financed child support benefit that is conditional on the prior establishment of a child support order and tied to fathers’ payments would have positive effects on both mothers and fathers. Although a public child support benefit would also increase government expenditures and parents’ incentives for living apart, it would nonetheless reduce the poverty, insecurity, and welfare dependence of single mothers and their children.

Policy Experiments

Congress is not likely to enact all of the changes discussed in this brief, but could provide federal funding for state-run experiments to help ascertain the full costs and benefits of these and similar reforms. Such experiments will have to be carefully designed and monitored, and they will be costly if they are to yield useful information. Participants would be randomly assigned to a control group or to experimental groups that receive one or more of the new policies outlined above. Multiple treatments would be necessary to disentangle the effects of different components of the reform package, and the experiment should last at least five years or it will fail to mimic permanent changes in policy. Both mothers and fathers would have to be interviewed periodically in order to assess the effects.

Based on the costs of previous large-scale social science experiments, we estimate that the costs of the experiment would be at least several hundred million dollars. While such an experiment is costly, it is cheap compared to the full fiscal costs of implementing all of the recommendations and will allow policymakers and social scientists to determine whether the benefits it produces warrant an expanded slate of reforms.

Additional Reading

Bartfeld, Judi, and Irwin Garfinkel. 1996. “The Impact of Percentage-Expressed Child-Support Orders on Payments.” Journal of Human Resources, 31(4): 794-815.

Garfinkel, Irwin, and others. 1998. Fathers Under Fire: The Revolution in Child Support Enforcement. New York: Russell Sage.

Lerman, Robert, and Elaine Sorensen. 2000. “Father Involvement with their Nonmarital Children: Patterns, Determinants, and Effects on their Earnings.” Marriage and Family Review, 29(2): 137-158.

McLanahan, Sara, and Gary Sandefur. 1994. “Growing Up with a Single Parent: What Hurts, What Helps.” Cambridge: Harvard.

Miller, Cynthia, and Virginia Knox. 2001. “The Challenge of Helping Low-Income Fathers Support their Children: Final Lessons from Parents’ Fair Share.” New York: Manpower Demonstration Research Corporation.

Olds, David L., and others. 1997. “Long-term Effects of Home Visitation on Maternal Life Course and Child Abuse and Neglect.” Journal of the American Medical Association, 278(8): 637-643.

Sorensen, Elaine, and Chava Zibman. 2001. “Poor Dads Who Don’t Pay Child Support: Deadbeats or Disadvantaged?” Washington, D.C.: Urban Institute.

U.S. Department of Health and Human Services. 1995. “Report to Congress on Out-of-Wedlock Childbearing.” Hyattsville: National Center for Health Statistics.

Waite, Linda J. 1995. “Does Marriage Matter?” Demography 32(4): 483-507.