From Welfare to Work: What the Evidence Shows

Robert A. Moffitt
Robert A. Moffitt Johns Hopkins University

January 2, 2002

The American public has made clear that work by welfare recipients is a defining goal of state and federal welfare laws, the pursuit of which deserves the highest priority in social welfare policy. One of the four goals listed by Congress in the 1996 welfare reform legislation was to encourage job preparation and work. Work among welfare recipients is widely regarded as part of the social contract-a quid pro quo for the provision of income support-as well as a source of self-esteem and self-reliance among single mothers. This in turn is thought to increase the mothers’ chances for long-term economic improvement for themselves and their children.

Now that five years have passed since the 1996 reforms were enacted, the evidence shows that while much success has been attained, there are remaining concerns that Congress should debate during reauthorization. This policy brief reviews both the evidence and the concerns.

Employment Among Single Mothers Has Increased

The overriding single piece of evidence showing that progress has been made on the agenda of helping mothers on welfare work is the dramatic increase in employment rates among single mothers in the last decade. Employment rates among single mothers, the group most affected by welfare reform, have been slowly increasing for over 15 years, but have jumped markedly since 1994 (figure 1). Employment rates rose from 60 percent in 1994 to 72 percent in 1999, a very large increase by historical standards. Among single mothers who have never been married (the group with the lowest levels of education and some of the highest rates of welfare receipt) employment rates rose even more, from 47 percent to 65 percent over the same period.

Not all of this increase can be attributed to welfare reform. Part of the increase has been the result of the robust economy and the longest and strongest peacetime expansion in the last 50 years. Until the recent economic slowdown, employers, desperate for workers, dipped deep into the pool of single mothers and other disadvantaged individuals.

Another factor encouraging employment is the expansion of the Earned Income Tax Credit (EITC), which provides major financial incentives to work. Given the boost in income the EITC provides (up to $4,000 per year for families with two children), many women have been encouraged to try and “make it” off welfare. Other supports for women leaving welfare, as well as for those never on welfare, include increased child care subsidies, food stamps, and health benefits through Medicaid and the State Children’s Health Insurance programs. Nevertheless, despite these other factors, there is no question that welfare reform has played a significant role in increasing employment among single mothers. Even research studies that have attempted to parcel out the relative contributions of different forces on employment rates support this conclusion.

Most Women Leaving Welfare Find Work

These overall trends beg for more details on how individual families have fared in the wake of welfare reform. The largest body of evidence comes from data on women who were on welfare but have left, primarily those who left the Aid to Families with Dependent Children (AFDC) program before 1996 or those who left its successor, the Temporary Assistance for Needy Families (TANF) program, after 1996. Most states have conducted such studies. A recent review of these studies conducted by the U.S. Department of Health and Human Services indicates that the employment rate among welfare leavers is approximately 60 percent just after exiting welfare. Moreover, about three-quarters of welfare leavers worked at some point in the first year after leaving the rolls. When welfare leavers work, they generally work full-time. Their hourly wages range from $7-$8 per hour, somewhat above the minimum wage. Those who work earn about $3,000 per quarter, or $12,000 annually. However, the annual wage is an overestimate because most leavers do not work for four quarters in a row, only a little over one-third do, signaling a potential problem with employment retention and stability.

These employment rates are considerably higher than critics of the 1996 reforms feared; some predicted that families would be made destitute and homeless following the reforms, or that there would not be enough jobs for women leaving welfare. At least on average, this has not occurred. The fact that 60 to 75 percent of welfare leavers found employment is especially remarkable given that, over the decade prior to reform, the employment rate of mothers while they were on AFDC was never more than 9 percent. Equally notable in this light is the fact that almost 30 percent of women currently on the rolls are now employed.

The 60 percent employment rate of welfare leavers is not much different than that of women who left the AFDC program prior to welfare reform. Employment rates over the period 1984 to 1996 ranged from 48 percent to 65 percent, varying by the state of the economy and the area of the country. These rates are similar to the rates following reform. This is surprising because many more women have left the welfare rolls in this era of reform than in any prior period, and many of those who left recently are more disadvantaged than women who left the rolls in prior periods. The fact that employment rates of leavers have not been lower than those experienced by past leavers further supports the strong effect of welfare reform.

In addition, random assignment evaluations of pre-1996 reform programs which had time limits and work requirements and were reasonably close in character to the post-1996 programs put in place by the states also show positive effects on employment and earnings. The employment and earnings gains in these demonstration programs are the average gains for both women who have left welfare as well as women who stayed on the rolls, and they therefore represent a more comprehensive measure than studies of leavers alone.

Two of the most important reforms in the 1996 legislation were the imposition of federal time limits on the length of welfare receipt, and the use of more stringent sanctions for not complying with work requirements and other rules. A natural question is how women who hit a time limit or were sanctioned have fared relative to women who left welfare voluntarily or because of different inducements. Time limits have had relatively little effect so far because most states have retained the five-year federal maximum and, as a result large numbers of recipients did not begin to hit time limits until the late fall of 2001. Some states do have shorter time limits than five years, but they have exempted large numbers of families from those limits and have granted large numbers of extensions. These exemptions and extensions have typically been granted to the most disadvantaged families, so that it is primarily those with significant employment and earnings (while on TANF) who hit the time limit in these few states. As a consequence, in the one or two states where significant numbers of families have left welfare because they hit a time limit, post-welfare employment rates of those leavers are quite high (e.g., 80 percent). But in other states where fewer families have hit the limit, employment rates of time-limited leavers are no different than those of other leavers.

More is known about sanctions because they have been in force for most of the time since 1996 and in some cases even before then. Many more women have been sanctioned than have been hit by time limits. The studies of women who have left welfare because of sanctions show that such women are less likely to have jobs than other welfare leavers. This appears to be because sanctioned welfare recipients tend to be less educated, have lower job skills, and are in poorer health than other welfare recipients. Unfortunately, these findings suggest that sanctioning may often occur among women who are the most disadvantaged and have the greatest number of difficulties with work.

Women Leaving Welfare Have Low Incomes

Despite the high employment levels of women who have left welfare, their incomes increase only modestly after leaving the rolls. About half experience an increase in income immediately after leaving, with the other half experiencing a decline. After a year or two off the rolls, earnings gains slightly exceed the losses in TANF benefits. When EITC income is added in, the gains are slightly higher.

However, the major change in income after leaving welfare comes from increased income from other family members (very little from boyfriends and other unrelated persons, however). Such income is a larger component of total household income than either the earnings of the leaver herself or TANF and food stamp income. As a result of additional income from this source, total household income grows by about 20 percent after two years off the rolls. Income from other household members is thus a key ingredient to sustaining the incomes of women leaving welfare.

Random assignment demonstrations measuring the effects of several pre-1996 state welfare reform plans provide additional evidence of the impact of welfare reform on income. For states whose plans most resembled those implemented after 1996 (those with work requirements and time limits), income was essentially unchanged by the reforms three years after they began. However, neither the EITC nor the income of other family members was included in the income calculation, so it is probable that some income gains were in fact attained, possibly in the same 20 percent range found in other studies.

These demonstrations also show that, in the absence of earnings disregards, income is not likely to greatly increase for several reasons. One is that many women work part-time and thus have quite modest earnings, not enough to make up for lost benefits. Another is that many women are sanctioned off the rolls, when they have little or zero earnings, yet they still lose benefits. A third is that many states reduce TANF benefits dollar-for-dollar when earnings increase (at least if women stay on the welfare rolls), thereby canceling out any gain in income that might result from increased work.

The EITC has played a significant role in keeping household income from declining as much as it could. However, many women off welfare do not receive the EITC if they have not been able to achieve steady employment. Others who are working do not have enough earnings to achieve the maximum EITC payment, and others do not apply for it in their tax returns. Thus, the EITC has assisted some families but not all, and families with income declines tend to be those that have benefited from it the least.

Studies also show welfare leavers experience declines in their receipt of food stamps and Medicaid. It appears that this decline is not a result of loss of eligibility so much as it is a result of lower participation despite eligibility, possibly because access to offices that determine eligibility is difficult to sustain. For whatever reason, low rates of food stamp and Medicaid receipt are a significant problem among TANF leavers.

Women who have left welfare are not the only single mothers whose income has changed since the reform legislation of 1996. Low-income single mothers who choose to stay off welfare to try to make it in the labor market have had increases in income as well. The fact that the incomes of low-income single mothers as a whole have risen at the same time that incomes of welfare leavers have been relatively stagnant suggests that the incomes of such “non-entrants” have indeed risen, probably because they work more hours.

Some Leavers Are Not Doing Well

The flip side of the high employment rates of 60 to 75 percent of women who have left welfare is that 25 to 40 percent of those women are not working. Indeed, some studies have indicated that as many as 18 percent of leavers in some areas did not work at all for a full year after leaving the rolls.

This group is of some concern. Because they have lost their welfare benefits and do not have earnings, they have lower incomes than non-working women who are still on TANF. A fraction of these non-working leavers have a relative, spouse, or partner who brings some income to the household, and others supplement their income with benefits from other government programs.

One of the most common program benefits received by this group are disability benefits from either the Supplemental Security Income program or the Social Security Disability Insurance program for either the mother or her children. That many families leaving welfare receive disability benefits is a reflection of the high prevalence of health problems and disabilities that hinder work. Nevertheless, even with income from other family members and from government programs, non-working leavers have considerably lower income than they did when they were on welfare. Consequently, leaving welfare has been particularly disadvantageous for these women and their children.

The existence of such a group shows that there is great diversity in the experiences of welfare leavers, for while some have fared reasonably well, others have not. Not surprisingly, employment rates of less educated leavers are considerably below those of more educated leavers, and poverty rates are higher, as are the employment and poverty rates of those leavers who are in relatively poor health.

Random assignment studies of time-limited pre-1996 welfare reforms show some evidence that welfare reform results in a larger fraction of families ending up with below average incomes. The presence of a group of women who have left welfare and are not doing well is consistent with broader trend studies indicating that the poorest single mother families have experienced declines in income in the post-reform period.

The Number of Women Going onto Welfare Has Declined

As noted previously, women who were once welfare recipients and have left welfare are not the only ones affected by welfare reform. Some women have chosen not to apply for welfare subsequent to reform, possibly discouraged by the work requirements and other new mandates that come with being on welfare, and possibly encouraged enough by the good economy to stay off welfare and work. Other women have applied for welfare but have been rejected.

Over twenty states have formal diversion programs, which encourage women through financial inducements and other means to not come onto the welfare rolls. More than thirty states have either diversion policies or have imposed work requirements that must be fulfilled prior to eligibility for benefits.

The decline in the number of women joining the TANF rolls has been very large in the post-reform era. In some states, the decline in entry onto welfare has been more important quantitatively than the increase in exit rates in accounting for the caseload decline. This finding casts a different light on the caseload decline and demonstrates that there is an important group of women other than leavers whose employment, earnings, and income should be of interest to policymakers.

Unfortunately, no studies have been conducted to date that examine this group, so their employment status and well-being remains unknown. However, the studies which have showed large post-reform increases in employment rates of single mothers as a whole, and which necessarily combine both those who have left welfare and those who have not come onto the rolls, strongly suggest that employment rates of women who choose not to enter the welfare system are high.

Issues for Reauthorization

The overall picture of employment among single mothers in the wake of welfare reform is a favorable one, indicating widespread work among former welfare recipients and among low-income single mothers as a whole. With this accomplishment a given, reauthorization should focus on policies that address the remaining problems.

There are two major problems that deserve attention. One is the broad issue of how to improve the income gains of women who have left welfare for work. Income gains are too modest for too many families, with earnings gains insufficient to counter reductions in benefits and with poverty rates-though lower than for families staying on welfare-remaining high. Aside from the need to increase the income of former welfare families for its own sake, income gains from leaving welfare will be necessary, in the long term, to provide financial incentives for women to leave welfare for work. While sanctions and work requirements can continue to be used to push women into the work force, they will operate much more successfully if the financial incentives operate in the same direction.

More supports for working families in the form of increased child care assistance, assistance with transportation, and other work-related services can substantially increase the incentive to work. Moving more women from part-time work to full-time work would be another direction to pursue, but this approach has limits if adequate child care and transportation are not available. Providing stronger financial incentives with state EITCs and enhanced TANF earnings disregards are also possible, although the latter policy will keep families on the TANF rolls longer. Major improvements beyond this are likely to come only from increased earnings. This calls for expanding policies aimed at job retention, skills enhancement, and job training. States are only now beginning to think about these types of policies and have a long way to go before such policies are widespread and have a major impact on incomes.

The second major issue is how to develop policies to assist families that have special difficulties in establishing employment. These families are sometimes called the “hard-to-serve,” although that term begs the question of what types of services are needed. One important result of the studies reviewed here is that many of these families are found not to be on TANF or on any other major welfare program. Rather, they are already on their own, off welfare, and have very low incomes. Any set of services that is directed mainly to TANF recipients alone on the presumption that the most disadvantaged families are still on the rolls, will not reach these families. This fact requires a major expansion of assistance to the non-TANF population. Some states, notably Wisconsin, have made such an expansion a major goal, but most states are far from having penetrated this population deeply with services and programs.

Most observers already recognize that designing successful policies to move non-employed families into stable work will be very difficult, given the severity of the difficulties these families face. These difficulties include low levels of education and job skills, significant health problems (both physical and mental), substance abuse, and domestic violence. The multiple interlocking and overlapping sets of problems faced by these families should give pause to any optimistic view that easy solutions will lead to steady employment and significant earnings gains.

Given these difficulties, a more open discussion is needed of assistance policies for floundering families who are unlikely to achieve significant employment gains in the short-term or even medium-term. Long-term cash assistance accompanied by job training, health insurance, and better programs aimed at reducing substance abuse, mental health problems, and domestic violence need to be directed toward this population independent of employment considerations. While the strong work incentives that are currently in place should remain, thereby continuing to provide financial incentives to families to work at higher levels than they currently do, simply strengthening work supports and further increasing work incentives will not, by themselves, provide much help to these families.

Virtually all states have already taken steps to develop programs for these families. States have started to identify families with serious problems that are barriers to work, and then design an appropriate set of services for them. However, the states must further develop these programs before a successful set of identification procedures and an adequate service delivery structure is established. When coordinating treatment for these families with employment programs, states can use their ability to exercise exemptions from work requirements and time limits as a short-term strategy. But more active and aggressive policies should be implemented to address the needs of these floundering families, both on and off TANF.