The 1996 legislation provoked extensive research on many facets of the law and its implementation, making it one of the most closely-examined pieces of social legislation in recent decades. The law itself contained numerous provisions funding research and evaluation studies, and many non-profit foundations have also funded studies of the law's major provisions. Moreover, a great deal of research on issues addressed by the legislation was already underway at the time of its passage in 1996. Finally, many states have conducted their own research, especially on what happens to families that leave welfare. The available research has the potential to play a vital role in the reauthorization debate, but it must be synthesized, organized, and interpreted so that members of Congress, their staffs, and the policymaking community can use what has been learned about the effects of the 1996 legislation as a basis for considering changes in old policies or the initiation of new policies. This series of policy briefs from the Brookings Institution will attempt to do just that.
The appendix to this policy brief describes the welfare reform law in some detail. Six major programs were reformed by the 1996 legislation. The programs include AFDC, the Supplemental Security Income (SSI) program for children, child support enforcement, child care, food stamps, and child nutrition. In addition, the new law put considerable emphasis on reducing nonmarital births and welfare benefits for aliens. These changes are also described in the Appendix.
This initial policy brief focuses on results of the most debated and controversial 1996 reform: replacing the AFDC program with TANF. This change converted welfare from an entitlement that paid mostly cash benefits to needy parents into a combined cash benefit/work program with a benefit that was contingent on meeting work or work preparation requirements. Open-ended federal funding was replaced with block grant funding that was fixed over the 1997 to 2002 period, thereby giving states a financial incentive to lower welfare caseloads. States were also required to meet a work participation standard that began at 25 percent of the caseload in 1997 and grew to 50 percent of caseload in 2002. This and the other major reforms are summarized in the appendix.
Results to Date
Given the extensive information available from administrative data and from research, a great deal is known about the effects of the various changes in welfare law. In subsequent policy briefs (see back cover), we will summarize the research as it applies to specific issues and programs in much greater detail. Here, however, our intent is simply to provide a broad overview of results in the TANF program to date based on current research and data. Many studies suggest that a booming economy has played a significant role in producing the results we review below. Thus we should not assume that similar results would occur in a sputtering or contracting economy. Similarly, because of expansions of the Earned Income Tax Credit, Medicaid, child care, and other benefits that help working families, part of the reduction in welfare rolls and increases in work and income should be attributed to the greater availability of these work support benefits.
Generally, we take the view that the combination of work-oriented welfare reforms, aggressive implementation by states, the nation's growing system of work support benefits, the leveling off of nonmarital births, and a strong economy all contributed to the observed results. Existing studies do not allow us to definitively pinpoint the relative contribution of each. Demanding work requirements and new supports for the working poor appear to be more or less permanent features of the nation's new approach to fighting welfare dependency and poverty. Whether they will produce the same outcomes in a weaker economy is less certain.
Change at the State and Local Level
As with any reform, the first question in considering how it affected states and localities is whether the federal law was aggressively implemented. Based on studies conducted by researchers at the State University of New York at Albany and at the Urban Institute in Washington, D.C., it is clear that most states have dramatically revamped their welfare programs. No longer are local offices simply check-writing operations; now they are also programs that help people prepare for and find jobs. The typical welfare office has been transformed, personnel have been retrained, and the activities inside the welfare office—which most states have renamed "Work Centers" or some similar term—have expanded to include job-related pursuits.
Most revealing, the spending patterns found in the budgets of state and local offices have changed dramatically. Before welfare reform, state and local offices typically spent around 80 percent of their welfare money on cash benefits, with the rest going to administration, education, training, and child care. Now states often spend 50 percent or less of their funds on benefits (although few states have reduced their benefit levels for individual recipients) while their spending on job search, education, training, child care, and other work-related activities has expanded and diversified. If budget is policy, important changes have taken place in the welfare program to which clients are exposed at the state and local level.
Decline in Welfare Rolls
One very clear outcome of welfare reform has been a substantial decline in the welfare caseload of every state. In examining these and other data, it is worth noting that over half the states were implementing their own welfare reform programs under waivers from federal law by 1994. Although diverse, most of these waiver programs contained variants of one or more of the five characteristics of the TANF program outlined in the appendix. Thus, in most cases, we present data for the years 1993 through the most recent year, usually 1999. In this way, readers can consider the changes that have taken place over the entire period or various subperiods and come to their own conclusions about the most appropriate years for comparison.
Figure 1 shows that by December 1999, the national cash welfare caseload had declined by well over 50 percent from its 1994 peak. In addition, the rolls have now declined for five consecutive years, the most ever. Because a major feature of the TANF block grant is that states receive a fixed amount of federal funds for a period of years, an important consequence of the caseload decline is that the typical state now has much more money per family left on welfare. We estimate that the number of federal dollars available per family on cash assistance, including both TANF and the child care block grant, has increased from around $3,500 to almost $8,000 (many TANF and child care dollars are being spent for services to a broader group of families that includes welfare leavers and the working poor). In addition, because states are required to maintain their own spending at 75 percent of its level in the early 1990s when the caseload was much higher, states also have substantial money of their own that must be spent on poor and low-income families. Of course, caseload decline is only one measure of what has happened to families on welfare since 1996. It is good that so many families have left welfare, but only if there is evidence that their financial status has not seriously deteriorated and that they are moving toward self-sufficiency. Thus, it is important to examine the employment, income, and poverty status of these families.
Mothers' Employment
Two types of information are available to assess whether families leaving welfare are getting jobs. First are studies of mothers who have left welfare. Generally, these studies find that about 60 percent of mothers are employed at the time of the interview and about 75 percent have been employed at some time since leaving welfare.
Second are studies that have tracked overall changes in women's employment. After a decade of relative stability, the number of single mothers working rose by about 25 percent between 1993 and 1999 (Figure 2). Even more impressive, there was a 50 percent increase in the number of never-married mothers who had a job.
Never-married mothers, who tend to have less education and work experience than other single mothers, are the most likely to go on welfare and the most likely to have long spells on welfare. That these are precisely the mothers who have had the biggest increase in employment in recent years suggests that even poorly educated mothers of the type that used to stay on welfare for long periods are proving themselves capable of productive work in the private sector, at least during a period of low unemployment. Nevertheless, important questions remain about the employment stability and earnings of those who have left welfare and especially about the circumstances of those who do not have jobs.
Poverty
A major issue of the reauthorization debate will be whether those leaving welfare are better off. Figure 3 shows the annual changes in the TANF caseload, in poverty among all children, and in poverty among black children between 1995 and 1999. The figure shows that as the caseload has declined each year, so has overall child poverty and black child poverty. In fact, black child poverty declined more in both 1997 and 1999 than in any previous year and reached its lowest level ever in 1999. Similarly, poverty among Hispanic children is at its lowest level since 1979. Overall child poverty also declined substantially and by 1999 was also at its lowest level since 1979. But these results are averages which mask extreme scores. Even where averages are encouraging, the condition of those on the bottom, such as those who leave welfare and do not hold jobs, remain a focus of research and debate. Critics such as Wendell Primus of the Center on Budget and Policy Priorities and others have pointed out that the decline in poverty has not been as steep as the decline in the caseload and that deep poverty, defined as income below 50 percent of the poverty level, has been growing. Thus, even though child poverty is at its lowest level since 1979, and poverty for female-headed families is the lowest ever, there is still concern about the most disadvantaged families.
Earnings and Total Income
Moving from welfare to work is no guarantee that mothers will be able to lift their families out of poverty. Most mothers leaving welfare receive low wages, on average about $6.75 per hour. Nevertheless, one reason that many mother-headed families are leaving poverty is that they are earning more money than in previous years. If we array all mother-headed families from those with the lowest to those with the highest incomes and then divide the distribution into five equal parts, the bottom two-fifths will contain all or nearly all the mothers on welfare and most of the mothers who have left welfare. If the bottom two-fifths of the distribution in 1993 are compared with the bottom two-fifths in 1999, we find that annual earnings for the bottom fifth have increased from $1,331 to $2,417, an increase of nearly 82 percent, while earnings for the second fifth have increased from $4,815 to $9,603, an increase of almost 100 percent (Figures 4 and 5). Logically enough, more work has led to increased earnings.
Not surprisingly, mothers in both bottom fifths of income have lost welfare income. The bottom fifth lost almost $900 in cash and $300 in food stamps between 1993 and 1999. Even so, because their earnings and their income from the Earned Income Tax Credit (EITC) have increased so much, the total annual income of these poorest mother-headed families has increased by about $900 on average since 1993. The second fifth has lost even more welfare than the bottom fifth, but the combination of earnings and EITC has boosted their income by nearly $3,000 per year or over 20 percent since 1993.
Like the information on poverty, the data on the income of mothers in the bottom 40 percent of the income distribution are generally favorable. Even so, serious concerns remain that should receive further attention from researchers and policymakers. There are families at the bottom —say in the bottom 10 percent or 5 percent of the income distribution—that appear to be worse off without welfare. In addition, many families that go to work lose their food stamps, despite the fact that they remain eligible for them. Finally, Census data do not take adequate account of work expenses, especially child care and transportation, that may leave some families worse off or no better off, despite increased earnings.
Trends in Nonmarital Births
Because research suggests that children are better off, both economically and psychologically, in two-parent families, one objective of the 1996 law was to reduce the proportion of children born to unmarried parents. Figure 6 shows the trends in nonmarital births since 1940; the figure includes the number of nonmarital births, the rate per thousand women aged 15 to 44, and the percentage of all births that are to unmarried women. The pattern is striking: more or less continuous increases, slow at first but picking up steam after the mid-1960s, in all three measures until the mid-1990s, at which point they level off. The trends for teen births are even more striking: after a surge in the 1980s, the rate has declined every year since 1991 and has now reached the level of the 1970s. The timing of these trends suggests that more is at work here than welfare reform. The teen birth rate has declined due both to increased abstinence among youth and to increased use of birth control, especially Depo-Provera. However, the leveling off in nonmarital births for all women does begin in about 1994, the same year that most states were beginning to implement welfare reform.
A few studies have suggested that tougher child support enforcement reduces nonmarital childbearing, and one study of an experimental program in Minnesota that rewarded work and treated two-parent families similarly to those with only one parent found that the program increased both entry into marriage and the stability of existing marriages. Still, there are few studies that directly link the various policies contained in the 1996 legislation and reductions in nonmarital births. Thus, although the trends are favorable, we still do not have persuasive research on the specific policies that could maintain or deepen these trends.
Well-Being of Children
As a result of changes in income or their mother's employment, children in welfare families could be affected by the new law. Some argue that they will be better off once their mothers are working because regular employment will engender a more structured home environment, more work-oriented values, and greater exposure of the children to good quality out-of-home care or education. Others argue that the effects could as easily go in the other direction if families have fewer material resources, if the time pressures on mothers lead to greater stress, less adequate parenting, and less child supervision, or if any substitute care provided is of poor quality.
Data from a variety of experiments that were conducted under waivers prior to enactment of the new law suggest the effects on children were small. Child participation in organized activities, center-based child care, and health insurance programs generally increased. Academic achievement, behavior, overall health, and the home environment of the child, however, changed very little, if at all, during these TANF-like programs. Although about half of the child outcome studies found significant improvements in child behavior or academic achievement, the other half found no effects, and any improvements were very small. Effects for school-age children are the most positive while teens were more likely to be negatively affected and preschoolers were generally not affected at all. The biggest and most positive effects so far have been associated with more children being placed in child care, after-school programs, or other structured activities outside the home. These results are preliminary, but even with more data it may never be possible to determine the precise impacts on children of these changes in maternal employment and income, since the pathways through which they influence children are many, and different families are likely to react in different ways.
Conclusion
The overall pattern of welfare and work associated with welfare reform seems clear. The welfare rolls have declined greatly, more mothers than ever are working, the average income of female-headed families—especially those who are never married—is increasing, and poverty has dropped substantially. This pattern of outcomes suggests that welfare reform, along with the strong economy and supports for working families, produced some notable successes. However, researchers and advocates have also raised a number of issues and problems that they believe need to be addressed in reauthorization. We turn our attention to these issues in the second policy brief in this series.
Appendix
Appendix
Appendix
Overview of the 1996 Welfare Reform Law
There were eight major programs or policy domains addressed in the 1996 law: TANF, Supplemental Security Income (SSI) for children, child support enforcement, welfare for noncitizens, child care, nonmarital births, child nutrition, and food stamps. A summary of the major reforms in each program or domain follows:
TANF TANF was the most fundamental of the reforms because it replaced the entire AFDC program. The most important elements of the TANF reforms can be captured in five provisions. First, the individual entitlement to benefits provided by AFDC was repealed. The signal sent by this policy change was that no longer could able-bodied parents qualify for cash assistance simply by being destitute. The right to cash welfare was replaced by a system of mutual responsibilities in which cash benefits are conditioned on attempts to prepare for self support. Second, the funding mechanism of open-ended federal payments for every person added to the welfare rolls by states was replaced by a block grant with a fixed amount of funding for each state for 6 years. Thus, if states help families leave welfare, they retain the funds that had been used to pay the family's welfare benefit. This feature of the block grant was intended to provide states with a financial incentive to help families leave welfare. States were also given far more discretion than under AFDC law, to spend funds for purposes other than cash assistance, such as transportation, wages subsidies, pregnancy prevention, and family formation. Despite the new flexibility, states were required to maintain a fixed percentage of their former spending effort in order to qualify for full TANF funding. Third, states were required to place an escalating percentage of their caseload in work programs. Fourth, financial sanctions were placed both on states and individuals who fail to meet the work standards. In the case of individuals, states must reduce the cash TANF benefit and sometimes the food stamp benefit of adults who fail to meet the work requirements designed by states. Similarly, the federal government will reduce the block grant of states that fail to meet the percentage work requirement. This requirement stipulates that, by increasing at a rate of five percentage points per year beginning at 25 percent in 1997, states must have 50 percent of their caseload involved in work programs for a minimum of 30 hours per week by 2002. Finally, with a few exceptions, states are not allowed to use federal dollars to pay the benefits of families who have been on welfare for more than 5 years. Taken together, these five characteristics rendered TANF a radically different program than the AFDC program it replaced.
SSI for Children The SSI provisions for children were intended to tighten the standards by which children qualify for cash SSI benefits. A test called the Individualized Functional Assessment that many, including the General Accounting Office, thought too generous in judging children to be disabled, was banned and the definition of childhood disability was clarified. These and similar measures were designed to ensure that only poor children with serious disabilities were admitted to the SSI program. The Congressional Budget Office estimated that these reforms would reduce the number of children on SSI by around 180,000, although by late 2000 the reduction was actually more on the order of 100,000.
Child Support Enforcement The child support enforcement amendments, by sheer number alone, were the most extensive provisions in the 1996 legislation. They were exceptionally comprehensive and amended nearly every aspect of the child support enforcement program. The general themes of the reforms were to provide access to new sources of employment and salary information for state programs, to streamline and automate as much of the child support program as possible, and to reform the state programs so that enforcement activities were based on mass processing of information and cases. The major goal of these reforms was to improve the performance of the child support program so that more noncustodial parents would be located, more paternities established, more child support orders put in place, and more money collected to help custodial parents, especially mothers leaving welfare, as well as to reimburse taxpayers for providing cash welfare payments for needy children. Welfare for Noncitizens
Welfare for Noncitizens Next to TANF, the most controversial provisions in the legislation were those that sharply restricted welfare for noncitizens. When legislation on welfare for aliens was first enacted by Congress in the 1880s, the legislative text made it clear that no alien should be admitted to the U.S. if immigration officials thought the alien could become a public charge. Moreover, if aliens did become a public charge, they were subject to deportation. Even so, along with the striking growth of welfare programs in the 1960s and 1970s, aliens began to qualify for welfare. The goal of Republicans who sponsored the 1996 welfare reform legislation was to return welfare policy on aliens to its roots in which the underlying presumption was that aliens should not receive welfare. Thus, the 1996 legislation all but eliminated welfare for the first 5 years after aliens entered the country and made access difficult even after the fifth year. Although Congress modified the alien provisions in 1997 so that aliens already receiving SSI and some who would become eligible in the future could receive benefits, the five-year ban on welfare for all aliens who enter the U.S. after Aug. 22, 1996 was maintained. Even after 5 years, aliens are still banned from receiving SSI and food stamps, although states may at their discretion allow aliens to qualify for benefits under the TANF, Medicaid, and Social Services Block Grant (Title XX) programs. There are a number of exceptions to the five-year ban, such as emergency medical assistance, disaster relief, and assistance for abused children. In addition, aliens continue to be eligible for means-tested education and training programs so that they have the opportunity to improve their marketable skills and maintain self-sufficiency. Once aliens become citizens, they qualify for welfare on the same basis as any other citizen. Refugees continue to be eligible for welfare benefits, although their eligibility ends once they have been in the country for seven years, the approximate amount of time required to become a citizen.
Child Care The child care provisions of the 1996 law were straightforward and relatively noncontroversial. The first major change was that several programs that provided child care for low-income and welfare families were terminated in favor of increased funding for the Child Care and Development Block (CCDBG) grant, a program that provided states with greater flexibility in funding child care for low-income families. This reform had the effect of simplifying state administration of child care by allowing states to run a single program that covered all poor and low-income families, including those leaving welfare. The second major reform was to increase total child care funding by around $4.5 billion over 6 years. Finally, states were given even more flexibility in funding child care by provisions that allowed them to spend money from the TANF block grant on child care or to transfer funds from the TANF block grant into the CCDBG to spend on child care. By 2000 states had about $4 billion in combined federal funds from the CCDBG and transferred from TANF to help poor and low-income working families purchase child care. By contrast, in 1997 states had only about $2 billion of federal child care funds to help these families.
Reducing Nonmarital Births Each of the provisions outlined above was contained in a separate title of the 1996 legislation. However, the law also contained a host of provisions designed to reduce nonmarital births that were found in several titles of the law. President Clinton and Congressional Republicans agreed that nonmarital births were a major social problem for the nation and even that the high level of such births was a motor driving many other social problems such as welfare use, delinquency and crime, poor school performance, and illegitimacy in subsequent generations. Many influential Republican policy intellectuals and conservative members of Congress, such as Rep. Jim Talent (R-Missouri) and Sen. Lauch Faircloth (R-North Carolina), argued that unless something were done about nonmarital births, the emphasis on work, even if successful, would make little difference in the long run. In the absence of Senate support for mandating the measures most favored by social conservatives to reduce nonmarital births (a ban on cash payments to unmarried teen mothers and a ban on additional payments for children conceived or born while a mother was receiving welfare payments, otherwise known as the family cap), Republicans decided to throw in virtually any provision that might reduce nonmarital births. The most important measures were the illegitimacy bonus given to states that achieved a reduction in their nonmarital birth rate, a grant program for every state designed to teach children that sexual abstinence was the expected standard for youth, mandatory paternity establishment requirements in the child support program, and an explicit state option to impose family caps.
Child Nutrition The major change in this section of the 1996 law lowered the cost of the program providing nutrition aid to family day care facilities by reducing subsidies for homes not located in low-income areas or not run by a low-income provider. The reforms also reduced the federal subsidy for the summer food service programs, changed the "rounding rules" in child nutrition programs, ended the program of grants for expanding school breakfast and summer food service programs, and provided more flexibility to state and local child nutrition providers.
Food Stamps Considered the most extensive food stamp reforms since 1977, the 1996 welfare reform law expanded state options and control of food stamps, especially by allowing states to operate a simplified program that permits coordination with TANF rules. The 1996 reforms also strengthened work requirements for 18-50 year old recipients without dependents, tightened financial eligibility tests, greatly restricted eligibility for aliens, reduced the basic food stamp benefit across the board by about 3 percent, increased the penalties for fraud, and expanded the use of electronic cards to deliver food stamp benefits.
Appendix Table 1
Overview of TANF Provisions for Reauthorization |
| Provision |
Description |
Funding |
In Baseline? |
|
Basic TANF Grant
|
Block grant to states to help needy children, to reduce nonmarital births, and for other purposes
|
$16.5 billion annually, FY1996-FY2002
|
Yes
|
|
Illegitimacy Bonus
|
Bonus grant to reward up to five states for greatest reduction in out-of-wedlock birth rates
|
$100 million annually, FY1999-FY2002
|
Yes
|
|
Performance Bonus
|
Bonus grant to reward high performance by states for attaining goals of TANF
|
$1 billion for FY1999-FY2003; Average annual bonus grants are $200 million
|
Yes
|
|
Population and Poverty Adjustor
|
Supplemental grants for 17 qualifying states with above-average population growth and low welfare spending per poor person
|
Up to a total of $800 million for FY1998-FY2001
|
No
|
|
Contingency Fund
|
Matching grants for needy states
|
$2 billion for FY1997-FY2001.
|
No
|
|
Indian Tribes
|
Grants for Indian Tribes
|
$7.6 million annually, FY1997-FY2002
|
Yes
|
|
Territories
|
Grants to territories for TANF, foster care, adoption assistance programs, and other programs
|
About $116 million, FY1997-FY2002, for TANF, Title IV-E, and aid to the aged and disabled
|
Yes
|
|
Loan Fund
|
Interest-bearing loans for state welfare programs
|
Total amount of loans outstanding may not exceed $1.7 billion
|
N/A
|
|
Medicaid for Families Leaving Welfare
|
Federal payments for families leaving welfare be given up to 12 months of Medicaid
|
About $.5 billion per year
|
Yes
|
|
Additional Medicaid Administrative Costs
|
Funds provided to compensate for increased costs of computing Medicaid eligibility
|
$500 million total without fiscal year limit
|
No
|
|
Research: Census Bureau
|
Census Bureau study to evaluate impact of TANF on national sample of low-income families
|
$10 million annually, FY1996-FY2002 (plus supplementals)
|
No
|
|
Research by HHS on effects, costs, and benefits of state TANF programs
|
Funds for DHHS to use to evaluate and conduct research on welfare reform
|
$15 million annually, FY1997-FY2002
|
No
|
Note: If TANF provision is in the baseline, then Congress will not need to find a funding mechanism (either a tax increase or a program cut) to reauthorize the provision. If funding is not in the baseline, Congress must find a funding mechanism. For example, if Congress decides to reauthorize the Census Bureau study for five years, it will need to find $50 million in funding offsets. |