From Welfare to Work: Making Welfare a Way Station, Not a Way of Life

In 1996 Congress reformed a welfare system that was deeply unpopular with the American public. Although Republicans pushed hard for reform, many Democrats—led by former President Bill Clinton—went along, and in the end both parties took credit for the new law. The system was transformed from one that handed out cash to one that required work and penalized with a loss of benefits those who failed to comply with the requirement. What had been called Aid to Families with Dependent Children (AFDC) became Temporary Assistance for Needy Families (TANF). The use of the word "temporary" was noteworthy. Welfare was to be a way station, not a way of life.

Although most politicians supported these changes, many scholars and advocates for the poor complained that the new law was a voyage into uncharted waters—an experiment with the lives of some of the nation's most vulnerable citizens and their children. Noncitizens who were legal residents of the United States were dealt an even bigger blow, as they and their children lost many benefits previously available to them. In this special issue of the Brookings Review, our contributors assess the results of the experiment to date and reflect on next steps, including what modifications may be needed when Congress reauthorizes the law in 2002 and what directions all this suggests for antipoverty policy in the United States. This introductory essay builds on their insights but contains my own assessment as well.

Is Welfare Reform Working?

Not surprisingly, the answer to the question of whether welfare reform is working depends on whom you ask. That said, even the law's critics point to at least some successes while its supporters acknowledge some limitations.

All our contributors agree that caseloads have declined far more than expected (from 5 million families in 1994 to 2.2 million in June 2000), that about 60 percent of those who have left welfare are working, that employment rates among single mothers have increased dramatically, and that child poverty has declined. They also agree that a substantial minority of mothers who have left the rolls are not working and that many of those who remain on the rolls may have difficulty finding or keeping jobs because of poor schooling, substance abuse, depression, or other barriers to employment. Finally, those who are working tend to earn low wages (about $7.00 an hour on average) making it hard for them to support their families.

Another concern, noted by Wendell Primus, Mark Greenberg, and Tom Downey in particular, is that the poverty rate has not fallen as much as the caseload. Fewer poor children are receiving assistance. In addition, the incomes of the poorest one-fifth of single-parent families have continued to fall, many families remain in deep poverty, and according to some reports, requests for emergency assistance have grown. Overall, 700,000 families were significantly worse off in 1999 than their counterparts in 1995.

Because welfare reform was implemented during an unprecedented economic expansion, questions also must be raised about how much of the good news should be attributed to the 1996 law and how much to a strong economy or to the growth of other programs such as the earned income tax credit (EITC) over this same period. Research on this question doesn't permit firm answers, but almost everyone agrees that all three have been important In his essay, Ron Haskins compares the 1990s to the 1980s, when employment also expanded strongly but without comparable declines in caseloads. The data he cites suggest that welfare reform played an important, and probably a critical, role.

Still, many people remain uneasy about what will happen should the economy sink into a prolonged recession. An economic downturn would not only swell the caseload, according to this view, but strain state budgets, perhaps forcing some states to cut benefits. At a minimum, states would likely reduce spending on an array of supportive services for those making the transition from welfare to work and for other low-income families.

Mickey Kaus considers the longer-term implications of welfare reform. He notes that reform was at least partially motivated by a view that the old system of no-strings entitlements had encouraged the growth of an urban underclass, characterized by idleness and out-of-wedlock childbearing. For many, dismantling this system is the key to reconnecting an isolated and stigmatized group of the poor to work and marriage. Even if the short-term result were to make this group worse off, Kaus argues, these longer-term goals should be pursued. The major purpose of the 1996 law, after all, was not to end poverty or improve material well-being, but to end dependency and encourage the formation of stable two-parent, married families. Although the family formation goals have not yet been achieved, some signs are encouraging. Teen pregnancy and birth rates have plummeted, and the proportion of children born outside marriage has leveled off. Lack of more definitive progress in family formation should not be surprising, for such cultural changes are deep-seated and likely to be decades in the making. In the meantime, the dramatic increase in employment among never-married mothers has had one unintended consequence: by creating greater sympathy for their lives, it has made new supports for this group politically feasible and thereby transformed the debate about poverty. No longer is the debate primarily about how much cash assistance and support services to provide to mothers who stay home with their children; it is about how much assistance to give them once they are working to support these same children.

One important feature of the new law was the enormous discretion it gave the states. Several contributors (Richard Nathan and Tom Gais, Mark Greenberg, and Bruce Katz and Katherine Allen) address the question of devolution and how welfare reform has played out differently in various parts of the country. Most states have embraced the goal of getting recipients into jobs as quickly as possible, a strategy that appears more cost-effective than providing substantial up-front education and training. But it is also a strategy dictated in large measure by federal rules that require states to place an increasing fraction of their caseloads in jobs and that reward them for doing so by freeing up funds for other, related purposes. Many states also supplement the limited earnings of those moving into jobs by allowing them to keep a portion of their welfare check. And as Gordon Berlin notes, these efforts to make work pay, although more costly than mandating work alone, can produce more positive effects not only on employment and earnings but also on some measures of child well-being. A key issue for the future will be whether welfare reform is about improving the lives of families and children or about ending dependency and saving money for taxpayers. Thus far, different states have adopted different approaches, with Minnesota, for example, emphasizing making work pay while Texas and Wisconsin have put more emphasis on ending dependency. And whatever formal policy choices states may make, local practices vary substantially not only from one state to another but even from one county to another within a state.

For these and other reasons, caseloads have declined unevenly. On average nationwide they have dropped by more than half, but that average reflects sharp variations—from 90 percent in Wyoming and 79 percent in Wisconsin to 46 percent in California and 26 percent in Hawaii between January of 1994 and June of 2000. And as Katz and Allen note, declines have been slower in the cities than outside of them, so that more and more of the caseload is concentrated in urban counties.

Not surprisingly, states are happy with their new role and are likely to want a simple extension of the law when it is reauthorized in 2002. But the wide latitude given states by the law must be balanced against the need for accountability. Today policymakers lack good information about what states are doing, where federal funds are being spent, who is receiving what services, and what happens to families after they leave welfare. Data on such matters have never been adequate, but as Nathan and Gais, as well as Greenberg, note, devolution has entailed real losses in information.

Does the Law Need to Be Changed?

Vin Weber observes that neither of the political parties in Washington is advocating major changes in welfare law. Both are more likely to fine-tune or modify it in ways that build on, rather than reverse, its current thrust. From a local perspective as well, some stability could be useful; too much change at the federal level makes life difficult for those with day-to-day responsibility for implementing the law.

Nonetheless, debates will undoubtedly arise about funding as well as other matters. Conservatives may argue that states no longer need so much money in the face of falling caseloads while liberals will want to expand funding to serve both those who remain on the rolls and a broader group of low-income working families. Currently, states receive $16.5 billion a year from Washington to assist needy families with children. States are required to maintain their own spending at 75—80 percent of what they spent during the early 1990s, or around $11 billion a year. Because caseloads have declined sharply, states have reaped a windfall and now have much more to spend per family on welfare than they did before 1996. But if caseloads should rise again in response to a recession, states are likely to need more than the small contingency fund now available for this purpose. Because the allocation of federal funds among states is based on commitments that historically favored high-benefit states, another issue is the fairness of providing more than twice as much federal money for a poor child in a wealthy state as for one in a poorer state.

One of the new law's most contentious issues has always been its five-year limit on the time families may receive federally funded benefits. Because the time limit will not begin to kick in until this coming fall, its full effects have not yet been felt. Should states start implementing it in the midst of a recession, the resulting harm to families could produce an outcry that might well change the debate. But states have other options. They may use their own funds to keep families on the rolls longer or may put families into community service jobs. Federal rules already allow states to exempt 20 percent of their caseload from the limit, and states may ask Washington to liberalize those rules.

The issues of funding and time limits will almost surely be linked to a broader debate about the purposes of the law. Once the goal of welfare reform is broadened beyond providing cash assistance to needy families to include encouraging marriage, making work pay, and helping families with problems ranging from substance abuse to mental health, the adequacy of any particular level of resources is unclear. Indeed, one danger is that funding will be set based on historical experience while the objectives expand in new directions. In particular, several contributors argue that the law should give much more attention to reducing poverty and finding strategies and performance measures that relate to this larger goal.

Beyond Welfare: The Unfinished Agenda

Certainly, the new welfare law, and any modifications to it in 2002, will not end the debate about poverty in the United States. That debate is ongoing and will always involve disputes about the balance between individual and social responsibility. In 2002 the reauthorization of TANF and related legislation, such as food stamps and the child care block grant, will provide an opportunity to reengage that debate.

I would argue that the three most important aspects of that broader debate are what society owes to those who work but remain poor; what, if anything, can be done to reduce childbearing outside of marriage; and how to break the cycle of poverty for children growing up in low-income communities.

Supports for the working poor have expanded over the past two decades, especially since 1990. The earned income tax credit is now the nation's largest antipoverty program by far. Future debates are likely to center on whether it can be further expanded to encourage work and marriage while minimizing error and fraud. Other support strategies likely to be discussed are subsidized child care or health insurance for low-income adults, a higher minimum wage, refundable tax credits for children in families that work, or cash supplements paid through the welfare system for those in low-paid jobs. Concerns about whether the least skilled will be able to find work in a less robust economy may also be addressed. Although few states have chosen to back up the work requirement in the new welfare law with community service jobs for those unable to find jobs in the private sector, linking the two has broad appeal. Liberals have always liked the idea of guaranteeing a job to those unable to find work and conservatives have come to understand that last-resort jobs in the public sector make a tough work requirement more politically acceptable as well as easier to enforce.

Reducing childbearing outside of marriage is a more daunting challenge. Not only is there less consensus about this as a goal, but also there are far fewer workable remedies. Nonetheless, reducing out-of-wedlock childbearing was a key goal of the 1996 legislation and for good reason. The growth of single-parent families, driven almost entirely by out-of-wedlock childbearing in recent years, has substantially worsened child poverty. Wade Horn suggests that public policy might encourage marriage by reducing the marriage penalty in the EITC, by helping fathers earn enough to support their children, and by funding programs that enhance the marital and parenting skills of high-risk families. But he admits that such strategies may not change people's behavior. The fact that many women in these communities begin childbearing at a very early age—typically in their teens—suggests the need to prevent early sex and pregnancy. The 1996 law provided funding for abstinence education programs whose efficacy remains unproven. Bearing in mind that the typical teenager spends about 7 hours a day watching television, surfing the Internet, and reading popular magazines, reaching them through these popular media may be far more cost-effective than school or community-based programs regardless of whether the message is abstinence or safer sex. In short, emphasis needs to be placed as much on changing cultural norms through creative use of the media as on offering more sex education programs. And with so many mothers now working, and some evidence that this has created problems for their adolescent children, more after-school programs may be needed as well.

Breaking the cycle of child poverty is a high priority for many but often gets lost in discussions of the welfare system. To the extent it is discussed, it is usually in the context of providing child care for mothers leaving welfare for work. Most observers recognize that the children of these mothers will need some sort of care, but they cannot agree whether the available care is adequate and whether improving it could be one way to break the cycle of poverty. Thus far, there is little evidence that requiring mothers to work has harmed their children. But the goal of leaving no child behind goes beyond doing no harm. In fact, many believe that if every young child of a mother leaving welfare had access to Head Start or another high-quality early education program, more children would reach school age ready to learn and, with appropriate follow-up in the elementary grades, go on to greater success as adults.

Looking Ahead

Welfare reform has been far more successful than most people anticipated. Caseloads are down, employment among single mothers is up, and poverty rates have fallen. In the absence of a major recession that reverses all this good news, it is unlikely that the direction of current policy will be reversed or that major features of the 1996 law will be modified. That doesn't mean there won't be some changes and that reauthorization won't or shouldn't catalyze a broader debate about antipoverty policy in the United States. The most important challenges in this regard are to provide greater supports for low-wage work, to reduce childbearing outside of marriage, and to ensure that more children have a good start in life.