If welfare reform is ultimately to be considered a success, the new system should do a better job than the old one at enabling low-income children to grow up to become healthy, productive adults. Protecting children from destitution was, after all, the original mission of the welfare system when it was included in the Social Security Act in 1935. But during the 1996 debate on welfare reform, poverty reduction was pushed to the margins by calls for reducing welfare dependency.
As Temporary Assistance for Needy Families (TANF) comes up for reauthorization this year, however, we have the opportunity to reconcile the conflicting goals of the welfare system: Evidence from rigorous new research shows that we can reduce poverty and promote work -- and improve outcomes for children, especially their school performance -- all at the same time. In particular, recent studies of three programs -- the New Hope Project in Milwaukee, Wisconsin; the Canadian Self-Sufficiency Project; and the Minnesota Family Investment Program -- whose results have been published over the last two years show that children's outcomes improve when policies increase parents' employment and their income.
Each of the three programs used a different method to get more money to low-income working parents each month. (All of them drew some public subsidy.) New Hope operated outside the welfare system, offering low-income parents who worked full time a cash supplement, child care, and health benefits, with a guaranteed job for those who couldn't find private employment. The Canadian program provided a generous cash supplement to welfare recipients who left the welfare system to work full time after receiving assistance for at least a year. The Minnesota program supplemented welfare recipients' earnings by allowing them to keep part of their welfare checks when they worked.
All three succeeded in increasing employment and income for single-parent families, and all three produced concrete, if modest, improvements in child development (roughly corresponding to an increase on a standardized test from about the 25th percentile to the 30th percentile). Two of the three studies also showed evidence of improvements in children's behavior. Interestingly, the improvements were greatest for children of long-term welfare recipients, a group whose capacity to sustain employment is of particular concern to policy makers.
These findings are significant because until now, though we knew that low-income children fared worse than their middle-income peers in school, we did not know whether these disadvantages could be reversed by policies that increase family income. If the problems faced by low-income children were not caused by money per se but rather solely by parents' low education or other characteristics, we would have expected no change in children's outcomes when parents' employment and income levels improved. These results demonstrate, with as much certainty as social science can muster, that financial supports for low-income workers can actually improve children's development. And because these studies looked at programs that target income supplements only to parents who work (two of the programs required full-time work while the third combined the supplements with work requirements for long-term recipients), their results render moot the conventional wisdom that income supports inevitably lead parents to cut back their work efforts.
When you compare these programs with welfare-to-work programs that required parents to work but did not provide earnings supplements, the results are even more striking. A multisite study found that while simply requiring work did lead to increases in parental employment, it often did not increase family income because under the old Aid to Families with Dependent Children system, welfare grants were reduced nearly one dollar for every dollar parents earned. As a result, low-wage jobs left them no better off financially than welfare did, and with no overall net benefit for children. Thus, while these welfare-to-work programs did not show consistent harm to children from parents' employment, as some opponents of welfare reform feared, neither did they show benefits for children from maternal employment, as proponents had hoped.
The strong evidence that work-based supports can help improve child well-being means that the Bush administration -- if it means what it says in promoting child well-being as a central goal of the welfare system -- now has the potential to fundamentally alter the TANF system for the better. In some states there is already considerable movement in a positive direction. States' fledgling work-support systems include tax-based Earned Income Tax Credits (EITCs) and expansions in child-care subsidies [See Roberta Wolfe and Deborah Lowe Vandell, "Welfare Reform Depends on Good Child Care" ] as well as supports for working families that are not specifically contingent on work, such as expansions in Medicaid and health insurance for children. The evidence that such policies can benefit children while at the same time encouraging parents to work should bolster policy makers' resolve to continue them, even in the face of tightening budgets.
But states -- backed by the Bush administration -- should recognize that there is still considerable room to benefit children by expanding work-based supports. The new earnings supplement programs in Minnesota and Wisconsin improved child well-being even though those states already had EITCs and other supports for low-income workers that were more extensive than in most other states. In addition, while many states have followed Minnesota's lead and adopted policies allowing parents to keep a partial welfare check when they go to work, most of these supplements are time-limited along with other welfare benefits -- and some studies have found that when time limits truncate income gains, the gains to children's education fail to materialize.
One strategy would be for states to provide work supports outside of the welfare system, as in the New Hope program. Federal officials could encourage the development of these work-based supports by taking concrete steps to increase states' flexibility in using TANF funds. Since TANF began, states have often come up with ideas for providing work supports outside the welfare system -- a new program in Texas to provide job retention bonuses to former recipients is just one example -- only to meet with mixed signals about whether TANF funds can be used for these purposes. Clarifying these rules would provide a framework for state decision making more conducive to creative approaches.
There is still much we don't know, of course, and not all experiments have been successful. We do know, for instance, that under several different welfare-reform approaches, adolescents have fared worse than younger school-age children, and we have very little information about how infants and toddlers have been affected as work requirements increasingly apply to single mothers with very young children. We also need to know more about how children have been affected by state policies that eliminate welfare benefits altogether when parents do not follow work rules or other requirements.
But there will always be more to learn, and as we continue to reshape the nation's safety net our deliberations should be guided by the knowledge that financial supports to low-income workers can help us achieve our farthest-reaching goal: to propel the next generation forward on a positive track toward adulthood.