Strengthening Families: Ideas for TANF Reauthorization
January 22, 2002
The debate over TANF's role in strengthening families and promoting marriage may be among the liveliest of the reauthorization debates. It coincides with the emergence of new evidence showing that "family formation" trends have taken a positive turn. In the 1990s, teen pregnancy rates fell, non-marital birth rates stabilized and the percentage of children living with two parents increased. This article reviews the current political debate, summarizes recent trends, and provides recommendations on how the TANF block grant and related programs could be improved to strengthen families and increase the number of children in stable two-parent families without disadvantaging or stigmatizing single-parent families.
The Current Debate
Arguing that states have done too little to advance TANF's family formation goals, some conservatives have called for placing marriage at the top of the TANF reauthorization agenda. Representative Wally Herger, the chairman of the House Ways and Means Committee's Human Resources Subcommittee, has said: "During the first phrase of welfare reform, we made sure we were putting people to work. I believe that now is the time to stress the importance of marriage."  Progressive voices, as well as organizations representing the interests of states, have been much more cautious. While there is general agreement among these groups on the importance of marriage, there is greater skepticism about the extent to which government should be involved in the business of directly promoting marriage, a concern shared by some conservatives. Even among progressives, however, there is substantial agreement on a set of policy measures that would advance family formation goals, such as those aimed at reducing teenage pregnancy, ensuring that child support payments go directly to low-income children, and enhancing the employment prospects of young parents and potential parents.
Robert Rector and Patrick Fagan of the Heritage Foundation have made the most forceful arguments for placing marriage at the top of the reauthorization agenda. Rector has claimed that states have done "virtually nothing" in this area and proposes earmarking at least 10 percent of all federal TANF funds (about $1.6 billion per year) for specific "pro-marriage activities" such as school-based marriage education programs, public advertising campaigns, marriage mentoring programs, pro-marriage counseling during pregnancy, and "community-wide marriage policies."  In addition, Rector and Fagan have proposed setting up a new federal Office of Marriage Initiatives that would be funded in part by transferring some funds from child support and family planning programs. They also would eliminate or reduce what they view as widespread marriage penalties in existing means-tested programs.
Before joining the Bush Administration, Wade Horn proposed that Congress require states to indicate in their state plans how they would use TANF funds to encourage marriage and specify that "the intent of the 1996 law was to promote marriage, not cohabitation or visits by nonresident parents." He also proposed eliminating or reducing what he views as financial disincentives for marriage, including the more restrictive TANF eligibility rules for two-parent families that exist in some states.
In contrast to proposals that focus primarily on marriage penalties and specific "marriage promotion" activities, others have argued for a strategy that focuses more on reducing teenage pregnancy, supporting "fragile families," whether married or unmarried, and increasing the career prospects of low-income fathers. Both Daniel Lichter, a sociologist at Ohio State University, and Belle Sawhill of the Brookings Institution have argued that preventing out-of-wedlock and teen childbearing offers the most effective route to meeting TANF's family formation goals.  In a recent policy brief, Sara McLanahan, Irwin Garfinkel, and Ronald Mincy propose a comprehensive set of strategies, including services to strengthen fragile families, disregarding a significant portion of the earnings of resident fathers when determining a family's TANF eligibility and benefits, and expressing child support obligations as a flat percentage of a noncustodial parent's income.
This debate comes at the same time as important "family formation" trends have taken a positive turn. For most of the last half of the twentieth century, nonmarital birth rates trended upward at a steady pace. This increase, combined with declines in marriage and marital stability, had a dramatic impact on family structure. Between 1960 and 1995, the percentage of children living with only one parent more than doubled, from about 12 percent to 27 percent. More recently, however, many of these trends have either stabilized or reversed direction.
- The teen birth rate — the number of births for every 1,000 young women age 15-19 — has fallen significantly since the early 1990s.  After peaking at 62 in 1991 — the highest rate since the early 1970s — the rate declined each subsequent year. In 2000, the rate was 48.7, the lowest level ever reported for the United States. The black teen birth rate fell by 20 percent, from 110.8 in the first half of the 1990s to 88.4 in the second.
- The nonmarital birth rate — the number of births for every 1,000 unmarried women of childbearing age — also declined, albeit at a less dramatic pace than the teen birth rate. The nonmarital birth rate increased from 43.8 in 1990 to 46.9 in 1994, but then declined in each subsequent year, except 1998. By 1999, the rate had returned to its 1990 level.
- In the late 1990s, the percentage of children, particularly low-income children, living in two-parent families increased while the percentage living in single-parent families fell.  Between 1995 and 2000, the number of children living with single mothers fell by slightly over one million, and the percentage of children living in single-mother families declined by 1.5 percentage points. The drop was larger among lower-income children: the percentage of lower-income children (roughly the poorest two-fifths of children) living with single mothers declined by 3.9 percentage points between 1995 and 2000, to 32.7 percent. Furthermore, the proportion of lower-income children living with married parents rose by 2.2 percentage points over the same period. While these changes in children's living arrangements seem generally positive, it is not yet possible to conclude that these changes have had additional positive (or negative) effects on child well-being or on the quality of family relationships. 
Policy changes may have played some role in these changes, but it seems unlikely that welfare policy changes were the driving factor. The downtrend in teen pregnancy and the flattening-out of nonmarital birth rates started before enactment of the 1996 welfare law. While the reversal of the decline in the share of children living with married parents probably began in 1995 or 1996, thus more closely coinciding with the passage of the welfare law, there is little evidence that welfare changes were primarily responsible for this trend. For example, a recent analysis finds that states implementing so-called "family cap" policies, which deny incremental benefits to additional children born to TANF recipients, or more stringent sanction and time-limit policies did not experience larger increases in the share of young children living with married mothers between 1995 and 2001 than states without these policies. Given the timing of the changes and the limited evidence of a strong welfare link, it seems likely that a range of factors, including growing incomes, changing social and cultural norms, the strengthening of the child support enforcement system, the expansion of the EITC and other work supports, as well as other welfare reform efforts, have all affected family formation decisions.
Marriage and Social Policy
Growing up with two parents clearly offers economic benefits for children. Children in families headed by married or cohabiting couples — which have two potential earners — have much lower poverty rates than children living with single parents. Married couples also tend to be better off economically than cohabiting couples. Married families are less likely to be poor than cohabiting couples even after controlling for certain other factors likely to affect poverty status, including race, education, age of parents, and number and age of children.  And poor married families appear to be less likely to miss meals or to fail to pay housing-related costs than other poor families, including poor cohabiting couples.  While the correlation between marriage and these positive economic outcomes is suggestive of some independent effect of marriage on income and hardship — that is, beyond that gained by simply adding an additional earner to a single-parent household — it also may be that married couples have other unmeasured characteristics that lead to increased incomes and reduced material hardship.
The bulk of the evidence also suggests that growing up with both biological parents has positive effects on child well-being independent of income. Although most children raised by single mothers do quite well, there are strong indications from research findings that, all else being equal, the absence of a biological father increases the risk of a number of negative outcomes for children, including lower educational attainment, increased likelihood of teenage childbearing, and diminished early labor force attachment.  It should be noted, however, that this research does not show that marriage per se has an independent effect on child well-being. In fact, children in stepparent families do no better on various measures of child well-being than children in single-parent families.  And we know relatively little at this point about whether children who grow up with two unmarried biological parents fare any differently than children who grow up with two married biological parents. 
While marriage potentially offers both economic and developmental benefits for children, there are reasons to be cautious about the manner in which government policy attempts to affect marriage decisions. If the emphasis is on "quantity" rather than "quality," marriage promotion activities could end up encouraging unstable or high-conflict unions. It may be especially difficult to ensure stable unions when the partners are young or for mothers who have already had a child prior to marriage — risk factors which are quite common among the disadvantaged population served by TANF. Several studies have shown that women who marry at younger ages, especially women who marry in their teens, have less stable marriages than women who marry at an older age. For disadvantaged women, a marriage that ends in divorce may leave them worse off than if they had never married in the first place. A recent study using data from the 1995 National Survey of Family Growth finds that about one-third of women who married after having their first child born out-of-wedlock were divorced at the time of the survey. Notably, the study finds that these women were worse off economically compared to similar unwed women who did not marry. 
Are Marriage Penalties a Significant Problem?
Policy changes that would reduce "marriage penalties" figure prominently in some proposals to strengthen family formation. While frequent claims are made about the size and significance of marriage penalties in tax and social welfare programs, there is little evidence to support claims that marriage penalties as conventionally understood have had a negative effect on marriage rates. For example, most recent research on the effects of the EITC on marriage and family formation finds no or only very limited evidence that EITC marriage penalties have had an effect on marriage rates.1
The tax system includes marriage bonuses for some families and marriage penalties for others, depending on the employment status and earning levels of the parents. In contrast, when the added costs of maintaining two separate households and other factors are considered, the social welfare system does not present significant marriage penalties for couples with a biological child in common.2
- Safety net programs such as TANF cash assistance programs, the food stamp program, housing, and Medicaid make no distinction between married parents who live together and unmarried parents who live together with their children. The income of both parents is counted when determining eligibility for and level of benefits in these programs, regardless of the marital status of parents who live in the same household with their children. In states that impose stricter eligibility requirements on two-parent families, these rules are applied to both cohabiting and married-couple two-parent families. Thus, there is no eligibility"penalty" if a cohabiting couple decides to marry.
- In fact, there are generally strong incentives for a couple with a child in common to live together, rather than apart. In a state that does not impose additional eligibility restrictions on two-parent families, if the parents decide not to live together, neither parent is likely to be "better off" financially. Two separate households must now be maintained. The custodial parent and child — who must pay for housing costs with just one income — may be more likely to qualify for means-tested benefits due to the loss of income that often occurs when a parent leaves the home. These benefits, however, generally will not leave the family with higher disposable income than when both parents were together. Finally, the noncustodial parent — in addition to paying his or her housing costs — must pay child support to the custodial parent. In short, neither parent generally has a financial incentive to separate.
Marriage penalties can exist in other situations — notably when a custodial parent marries someone other than the child's biological parent. In this case, the social welfare system may "deem" some income from the step-parent to the custodial parent and child when determining some types of benefits. Such deeming would not occur if the adults cohabited rather than married. (Food stamp benefits, however, are not affected by the decision to cohabit rather than marry since these benefits are based on total household income.) Because marriage penalties do not present a serious problem in the case of parents with a child "in common," such arguments should not be used to justify providing enhanced benefits or preferential treatment to two-parent families or to reduce assistance to single-parent families.
1. See, e.g., David Ellwood, The Impact of the Earned Income Tax Credit and Social Policy Reforms on Work, Marriage, and Living Arrangements, 4 National Tax Journal 1063, December 2000.
2. For a fuller discussion of these issues, see Wendell Primus and Jennifer Beeson, Safety Net Programs, Marriage and Cohabitation, paper presented at "Just Living Together: Implications for Children, Families, and Social Policy,"October 30-31, 2000 at Pennsylvania State University.
Domestic violence also is a serious concern in many relationships. Several studies have found that between 15 to 30 percent of welfare recipients have been recent victims of domestic abuse. While marriage-promotion advocates often point out that domestic violence rates are higher among cohabiting couples than among married couples, it remains unclear whether this is a causal effect of marriage (something about the institution of marriage itself has a dampening effect on domestic violence) or due to selection effects (less violence-prone couples may be more likely to marry and remain married). 
Even if we were able to address the quality of relationships, little is known about what kinds of policies and programs could produce increases in marriage rates. It remains to be seen whether the activities for which some have proposed earmarking TANF funds — school-based marriage-education programs, public advertising campaigns, marriage mentoring programs, pro-marriage counseling during pregnancy, and community-wide marriage policies — are actually effective in increasing marriage rates and marital stability. Dramatic statistics are often cited by advocates of marriage promotion strategies, but they do not appear to be the product of rigorous evaluations. (Moreover, the most-commonly cited statistics relate to decreases in divorce rates among the general population in particular localities rather than increases in marriage among disadvantaged individuals.) Only one of the strategies mentioned above, premarital education, appears to have been carefully studied and its effectiveness remains far from clear. 
The strongest evidence that social programs can have a positive impact on marriage rates of disadvantaged individuals comes not from a program that overtly sought to influence marriage decisions, but rather from what is perhaps the most progressive welfare reform demonstration program conducted in the United States in the last decade, the Minnesota Family Investment Program (MFIP). MFIP provided generous financial incentives and grant increases for both single and two-parent families, regardless of their marital status, and also eliminated restrictive rules that limited participation by two-parent families. MFIP reduced poverty rates and increased marriage rates for both single-parent and two-parent families.  Married two-parent families were more likely to remain married — 67 percent of MFIP two-parent families were married and living together after three years compared to 48 percent of AFDC control group two-parent families. Single parents were somewhat more likely to marry — 10.6 percent of single parents who received MFIP were married and living with a spouse after three years compared to 7 percent of single parents in a control group. While there is no consensus on why MFIP increased marriage rates, the researchers who conducted the study suggest that reductions in financial strain for two-parent families in MFIP reduced sources of marital stress and instability.
Supporting Marriage and Strengthening Families: Ideas for TANF Reauthorization
Government should create a better environment for marriage and families by reducing nonmartial births, providing work supports and promoting career advancement for both mothers and fathers, and improving economic security for all families. Essential elements of an agenda to improve child well-being and strengthen families include:
- a safety net that does not discriminate against two-parent families and ensures that more two-parent families who are eligible for benefits receive them,
- a strong child support enforcement system that increases the financial well-being of children,
- programs to help low-income fathers meet their financial and parenting responsibilities,
- initiatives that further decrease teen pregnancies, and
- a research agenda that would develop a knowledge base on which to build successful programs in these areas.
This agenda should not be pursued at the expense of single-parent families (for example, by cutting benefits for single-parent families or instituting program preferences or quotas for two-parent families) or in lieu of making further improvements in the current welfare system for single-parent families. If anything, the research findings showing that children in single-parent families face greater economic and developmental risks than children in two-parent families should lead to intensified efforts to enhance the economic security of children in single-parent families. Sara McLanahan makes just this point when she argues that "based on what is known to date, reducing the economic insecurity of single mothers is probably the most effective tool for protecting children from the negative consequences of family disruption" and that "[i]f single mothers were more economically secure, they might take more time in selecting a new partner, which, in turn, might make remarriage more beneficial for children."
Providing Supports to Two-Parent Families
As the MFIP findings suggest, programs that provide income support and employment services may help low-income two-parent families stay together by reducing financial sources of martial stress and instability. Unfortunately, few state welfare programs provide financial incentives for work that match those that were provided in the MFIP pilot and some states retain stricter eligibility criteria for two-parent families than for single-parent families. (To their credit, states have made notable advances in both these areas: almost every state now provides more generous work incentives than were available under AFDC and most states eliminated two-parent restrictions.) Moreover, poor two-parent families receive food stamps and TANF cash assistance at less than half the rate of poor single-parent families. While the disparities in TANF participation may be partially attributable to the more rigid two-parent eligibility restrictions in some states, the problem clearly involves more than formal policy barriers. It seems likely that many poor two-parent families do not know that they may be eligible for benefits or are reluctant to participate.
There are several areas in which policies can be improved to provide additional support for two-parent families:
- States should not be permitted to discriminate against two-parent families in establishing eligibility for benefits and services under TANF.
- The separate work participation rate for two-parent families should be eliminated. The current rate — under which a state faces penalties if less than 90 percent of two-parent families are participating in work activities — may create a disincentive for states to serve such families.
- States should be required to forgive child support debt owed to the state if a low-income, separated couple marries or remarries and to hold child support debt owed to the state in abeyance when a separated couple with children reunites but does not marry.
- Low-income parents — in both single- and two-parent families — should have access to public health insurance. This could be done by expanding the size of state allotments in the State Children's Health Insurance Program (SCHIP) and giving states the flexibility to use the new funds to extend Medicaid and SCHIP coverage to the uninsured parents of children eligible for those programs.
Child Support Enforcement
A strong child support enforcement system can help improve child well-being and strengthen families. There is substantial evidence that vigorous enforcement of support obligations promotes marriage by reducing non-marital childbearing and divorce rates.  Over the past two decades, sweeping changes have been made to the nation's child support enforcement system. These reforms appear to have improved the system's overall performance dramatically. Between 1991 and 2000, the number of paternities established nationally increased by 229 percent and the amount of child support collected in the federal-state child support system increased by 159 percent.
While these statistics are heartening, further improvements are necessary to increase the extent to which children benefit from the payment of child support. The most fundamental unfinished reform involves clarifying the program's muddled mission in which the recovery of welfare costs sometimes limits the extent to which children benefit from the payment of child support. 
- Families who leave TANF and are owed past-due child support should have first claim on all child support payments made by non-resident parents. While Congress made substantial progress on this front in the 1996 welfare law, child support collected through the interception of federal tax refunds by the IRS — the single largest source of collections of past-due child support — is still retained by the federal government and the states to pay off any unreimbursed costs of providing assistance to the family.
- Child support paid by non-resident parents of children receiving TANF should go directly to the child rather than being retained by the state. In addition, states should be encouraged to disregard at least a portion of the payment in calculating the family's welfare grant. Where child support is disregarded, states should not have to remit any share of the support to the federal government, as is currently required. As early results from demonstration projects in Vermont and Wisconsin show, child support pass-though and disregard policies can have a positive effect on both the number of fathers who pay child support and the average amount of support paid by fathers. 
Assistance for Low-Income Fathers
Even with an improved child support system, many low-income fathers do not pay child support regularly. While some fathers are simply unwilling to provide support to their children, many fathers want to be involved in their children's lives but face considerable barriers. Fathers of poor children are often poor themselves and have limited ability to pay significant amounts of child support. The most disadvantaged low-income fathers are similar in many respects to disadvantaged low-income single mothers: they often are young, lack high school diplomas and work skills, and have limited work experience. Many also struggle to cope with substance abuse, legal problems, job discrimination, and lack of affordable transportation and housing. While a growing number of states and cities are providing services to low-income fathers, these efforts remain quite limited.
These problems are exacerbated by child support policies that often are not designed with low-income fathers in mind. Too many low-income fathers face child support obligations that exceed the amount they are realistically able to pay (even though the child support award may fall short of what it costs to raise a child). In some cases, the current monthly support order is set too high; in others, the current monthly support order by itself may be reasonable, but added obligations, including birthing costs paid by Medicaid, fees, interest charges, and required additional payments toward past child support debt, may lead to an overall obligation that is impossible to meet. 
The federal government should provide states with incentives to extend employment services and other necessary services to low-income fathers, while addressing other limitations in the way the child support system works with low-income fathers. States also should be encouraged to adopt and test new approaches to subsidizing the payment of support by low-income fathers. For example, states could reward the payment of child support by partially subsidizing it, similar to the way the earned income tax credit rewards low-wage work for families with children. 
- States should be given some credit toward meeting their TANF work participation rate for low-income fathers of TANF children who are engaged in TANF work activities or pay a sufficient amount of child support. This would provide an incentive for states to extend employment services to more low-income fathers and increase child support collections for low-income children. A state would not be eligible for the credit if the number of mothers receiving employment services declined or there was other evidence that resources were being diverted from low-income mothers.
- States should be given one-time federal grants to review their child support policies and to develop programmatic recommendations to extend employment and parenting services to low-income fathers. As part of the review process, states should develop child support policies that prevent the build-up of unmanageable child support debt and allow for the forgiveness of child support owed to the government when merited. States also should be encouraged to address child support and employment issues in a comprehensive and integrated fashion across a broad array of state programs — child support, employment, criminal justice programs — and to implement programs that match or otherwise subsidize the payment of child support by low-income parents.
- Recovery from low-income fathers of birthing costs paid by Medicaid should be prohibited. This practice can saddle low-income fathers with large child support debts even if they consistently keep up with their current monthly support obligations.
- States should have the option to extend access to federally funded health care coverage to low-income non-resident parents on the same basis as coverage is available to low-income resident parents.
Reducing Teen Pregnancy
Though the teen birth rate fell in the United States in the 1990s, it remains a serious concern. The teen birth rate is still higher in the United States than in any other industrialized democracy. Children born to teenage parents — whether they are married or unmarried — are at a greater risk of growing up without the benefits of living with two parents. Teenage pregnancy leads to a host of problems that are best prevented rather than addressed after they arise.
A growing body of rigorously conducted research points to specific programs that have been shown to reduce teen pregnancy and childbearing. For example, a random assignment multi-site evaluation found that young teens enrolled in the Children's Aid Society-Carrera Teen Pregnancy Prevention Program were 46 percent less likely than teens in a control group to become pregnant or give birth.  This intensive program included both youth development and reproductive health components. Funding should be provided to states to replicate this and other effective programs and evaluate new initiatives. As described in more detail below, Congress should establish a research and development fund for this purpose.
Congress also should revisit the limitations placed on programs funded under the new abstinence education program.  The few studies of abstinence-only programs that have been completed to date do not show any reductions in sexual behavior or contraceptive use. Absent evidence of positive results, states should have broader flexibility in how they use abstinence education funds.
There is substantial interest in developing programs that further reduce nonmarital births, foster and strengthen healthy two-parent families, and increase the proportion of children cared for by both parents. Some have argued for an earmark of TANF funds for marriage promotion activities. Based on the evidence available, however, too little is known about policies and programs that will produce desirable results in this area to warrant such a federal mandate.
Further research in these areas is critical to developing a knowledge base on which to build future successful programs. States are unlikely to do substantial research in this area on their own without federal support and assistance. To support and systemize research in the states, a Family and Child Well-Being Research and Development Fund should be established, using monies previously allocated to the out-of-wedlock bonus, to encourage the replication of proven policies and to conduct research on programs designed to enhance the well-being of families and children.  The Secretary of HHS, with recommendations from a panel of welfare administrators and experts from a range of disciplines, could use this fund to provide technical assistance or to competitively fund evaluations of demonstration projects that would be proposed by states or localities.
Finally, states' ability to implement a policy agenda that supports marriage and strengthens families will depend in large measure on the availability of sufficient resources. Congress should provide states with the TANF block grant funding necessary to continue their existing commitments and to expand efforts in the other areas detailed here.
 A version of this paper will appear in a forthcoming edition of Focus, the newsletter of the Institute for Research on Poverty at the University of Wisconsin-Madison. The analysis and recommendations presented here were developed in close collaboration with the Center for Law and Social Policy. This paper draws in part from an earlier internal working paper on family formation co-authored by Theodora Ooms of the Center for Law and Social Policy and Shawn Fremstad.
 Three of the four purposes of the TANF program relate to family formation — ending the dependence of needy parents on government benefits by promoting job preparation, work and marriage; preventing and reducing the incidence of out-of-wedlock pregnancies; and encouraging the formation and maintenance of two-parent families. The law also includes other provisions related to family formation. For more information on TANF's family formation provisions and current state efforts in this area, see Testimony of Theodora Ooms, Senior Policy Analyst, Center for Law and Social Policy, House Committee on Ways and Means, Subcommittee on Human Resources, May 22, 2001 and American Public Human Services Association, State Efforts to Promote Marriage and Family Formation, Washington Memo, 2001.
 Megan Twohey, Getting Hitched to Promoting Marriage, National Journal, March 24, 2001.
 According to Rector, "[t]hese funds should be allocated directly by the federal government, or, to a lesser extent, through the state governments." Robert Rector, Using Welfare Reform to Strengthen Marriage, American Experiment Quarterly, Summer 2001.
 Wade Horn, Wedding Bell Blues: Marriage and Welfare Reform, Brookings Review, Summer 2001.
 Daniel T. Lichter, Marriage as Public Policy, Progressive Policy Institute, September 2001; Isabel Sawhill, What Can Be Done to Reduce Teen Pregnancy and Out-of-Wedlock Births?, The Brookings Institute, October 2001.
 Sara McLanahan, Irwin Garfinkel, and Ronald B. Mincy, Fragile Families, Welfare Reform, and Marriage, The Brookings Institution, November 2001.
 CTS Facts At A Glance , Child Trends, August 2001.
 National trends in nonmarital birth rates from 1988-89 through 1993 are affected by substantial underreporting of births in Michigan and Texas. According to the National Center of Health Statistics, if nonmartial births had been accurately reported by these states, nonmarital births would have peaked at an earlier point in the 1990s. Stephanie Ventura and Christine Bachrach, Nonmarital Childbearing in the United States, 1940-99, National Center for Health Statistics, October 18, 2000.
 Wendell Primus, Child Living Arrangements by Race and Income: A Supplementary Analysis, working paper, Center on Budget and Policy Priorities, January 2002.
 In addition, there is at least one troubling trend in this area. Data from the National Survey of America's Families suggest that between 1997 and 1999, there was a small but statistically significant increase in the percentage of children living with neither of their biological parents. Greg Acs and Sandi Nelson, "Honey, I'm Home." Changes in Living Arrangements in the Late 1990s, Urban Institute, June 2001.
 Richard Bavier, Recent Increases in the Share of Young Children with Married Mothers, unpublished paper, December 21, 2001. In fact, Bavier finds that the increase in the share of young children living with married mothers was actually larger in states that did not have especially rigorous sanction and time-limit policy than in the 17 states that did.
 Robert I. Lerman, Marriage as a Protective Force Against Economic Hardship, October 2001, paper presented at 23rd Annual Research Conference of the Association for Public Policy and Management, Washington, DC, November 1-3, 2001.
 Sara McLanahan and Julien Teitler, The Consequences of Father Absence, Office of Population Research, Princeton University, September 1997.
 Sara McLanahan, Parent Absence or Poverty: Which Matters More?, in Consequences of Growing Up Poor, Greg J. Duncan and Jeanne Brooks-Gunn, editors, 1997.
 See Wendy D. Manning, The Implications of Cohabitation for Children's Well-being, paper presented at "Just Living Together: Implications for Children, Families, and Social Policy," October 30-31, 2000 at Pennsylvania State University.
 See, e.g., Jeffry H. Larson and Thomas B. Holman, Premarital Predictors of Marital Quality and Stability, 43 Family Relations 228 (1994); Matthew D. Bramlett and William D. Mosher, First Marriage Dissolution, Divorce, and Remarriage: United States, National Center for Health Statistics, May 31, 2001.
 Daniel T. Lichter, Deborah Roempke Graefe, and J. Brian Brown, Is Marriage a Panacea? Union Formation Among Economically-Disadvantaged Unwed Mothers, paper presented at the 2001 annual meetings of the Population Association of America, April 2001. The authors do not speculate on why this is the case. It may be that unwed mothers who marry work less during the marriage or are less likely to pursue additional education or training while married compared to unwed mothers who do not marry.
 Richard M. Tolman and Jody Raphael, A Review of Research on Welfare and Domestic Violence, 56 Journal of Social Issues 655-681 (2000).
 For a discussion of this issue, see Catherine T. Kenney and Sara S. McLanahan, Are Cohabiting Relationships More Violent than Marriages?, Center for Research on Child Wellbeing Working Paper #01-2, June 2001.
 One study finds no difference in marital outcomes between couples who have participated in premarital education and those who have not. Kieran Sullivan and Thomas Bradbury, Are Premarital Prevention Programs Reaching Couples at Risk for Marital Dysfunction?, 65 Journal of Consulting and Clinical Psychology 24 (1997). There is some evidence that the Prevention and Relationship Enhancement Program (PREP) reduced divorce rates over a five-year period in a few sites, but methodological concerns leave the findings open to various interpretations. For a discussion of this and related research, see Scott M. Stanley, Making a Case for Premarital Education, 50 Family Relations 272 (2001).
 Virginia Knox, Cynthia Miller, and Lisa A. Gennetian, Reforming Welfare and Rewarding Work: A Summary of the Final Report on the Minnesota Family Investment Program, Manpower Demonstration Research Corporation, September 2000.
 Sara McLanahan, Parent Absence or Poverty: Which Matters More?, in Consequences of Growing Up Poor, Greg J. Duncan and Jeanne Brooks-Gunn, editors (1997).
 For a review of research on this point, see Irwin Garfinkel, Theresa Heintze, and Chien-Chung Huang, Child Support Enforcement: Incentives and Well-being, Joint Center on Poverty Research Working Paper, February 2001.
 Tabulations of HHS data by Vicki Turetsky, Center on Law and Social Policy.
 Vicki Turetsky, What if All the Money Came Home?, Center for Law and Social Policy, June 2000.
 Written Testimony of Vicki Turetsky, Senior Staff Attorney, Center for Law and Social Policy before the Subcommittee on Social Security and Family Policy, Senate Finance Committee, October 11, 2001.
 While we generally use the term "fathers" to refer to non-resident parents, the policies proposed here would apply equally to male and female non-resident parents.
 See Elaine Sorenson and Chava Zibman, A Look at Poor Dads Who Don't Pay Child Support, Urban Institute, September 2000.
 See Paula Roberts, An Ounce of Prevention and a Pound of Cure: Developing State Policies on the Payment of Child Support Arrears by Low Income Parents, Center for Law and Social Policy, May 2001.
 For a fuller discussion of this concept, see Wendell Primus and Kristina Daugirdas, Improving Child Well-Being by Focusing on Low-Income Noncustodial Parents in Maryland , September 2000
 Douglas Kirby, Emerging Answers: Research Findings on Programs to Reduce Teen Pregnancy, The National Campaign to Prevent Teen Pregnancy, May 2001.
 For background on the federal abstinence education program, see Jodie Levin-Esptein, "Abstinence unless Married" Education, Center for Law and Social Policy, February 1999 and Reauthorization Issues: Abstinence Education, Center for Law and Social Policy, November 2001.
 The out-of-wedlock bonus provided $100 million a year to be split among up to five states that had the greatest reductions in out-of-wedlock births while also reducing abortion rates.