Better Off Welfare

Welfare rolls nationwide have fallen by more than 50 percent since welfare reform was enacted in 1996. The goal of reform was not simply to reduce the number of welfare cases, but to move families on welfare – the vast majority of which are headed by single women – from dependency to independence through work. Despite this success, there are two lingering questions: Have the easiest-to-employ women left welfare, leaving the "hardest" cases behind? And are the families who have left welfare better off? Research conducted since 1996 has provided some surprising answers:

• Both national surveys and state data show that the women most at risk for long-term welfare receipt have left the welfare rolls at rates as fast as or faster than women who are much less at risk.
• In the states studied in depth, most of those leaving the welfare rolls have found employment, have increased their incomes relative to those who remain on welfare, and are gradually moving up the income ladder; a majority of those who have left feel they are better off.

It is well established that the single mothers most at risk for lengthy welfare receipt and long-term welfare dependency are those who have never married, are poorly educated, are young (18 to 29 years old), have young children (under age 7) and are black or Hispanic. Yet these are the ones who have been leaving the welfare rolls at the fastest rates. For example:

• Former Congressional Budget Office Director June E. O'Neill and M. Anne Hill, her colleague, found that during the 1990s, and particularly since 1996, welfare participation declined the most for the youngest mothers, for mothers with children under age 7, for those who had not completed high school, for black and Hispanic women and for never-married mothers.
• The Urban Institute found that welfare enrollment for both blacks and Hispanics fell proportionately more than for whites, and the decline was steepest for blacks.
• The Urban Institute data also showed that those who remain on welfare are not suffering the greatest hardships; for example, they do not have greater health problems than do those who have left the rolls.
• A Joyce Foundation study found that some who have left welfare continue to face such financial hardships as having a utility cut off or occasionally being late with a rent check; however, the study found that moving from welfare to work does not necessarily increase such hardships.

What proportion of long-term welfare recipients can be moved from welfare to work? We do not know, but in Wisconsin – which reduced its welfare rolls by more than 70 percent – only 2 percent of women who had been on the welfare rolls in 1990 were either still receiving welfare or had returned to the rolls after some work experience by 1998.

During the booming economy of the late 1990s, welfare reform critics claimed that the good economy was primarily responsible for the fall in welfare rolls. They predicted that if the good times ended, welfare rolls would rise. But a national analysis of welfare leavers by O'Neill and Hill found that:

• The Temporary Assistance for Needy Families (TANF) program created in 1996 has accounted for more than half of the decline in welfare participation since then and for more than 60 percent of the rise in employment among single mothers.
• By contrast, the contribution of the booming economy of the late 1990s to these declines was relatively minor, accounting for less than 20 percent of either change.

The most significant factor in caseload reduction is the policies established by the states – specifically, whether or not the state has an effective sanctions policy. Effective sanctions withhold benefits from welfare recipients who refuse to work or to comply with other program requirements. The states that have been less successful have had ineffective sanctions, high benefit levels and other counterproductive policies. Two of these states, New York and California, account for nearly a third of the remaining welfare caseload.

• In New York State, if welfare recipients refuse to comply with work requirements, their benefits are reduced by a small amount (from $588 per month to $475 for a family of three), but they continue to receive food stamps, public housing and other benefits.
• As a result, recent figures show that only 54 percent of adults perform assigned work.
• California offers high benefits and ineffective sanctions, using discretionary state funds and allowing working families to continue receiving benefits.

Without any changes in federal law, the states that have failed to reduce their welfare rolls by as much as the average could adopt more effective policies. The impact would be significant:

• If the 23 less-than-average states had done as well as the average state, more than 800,000 additional people would have left welfare.
• Instead of a 59 percent reduction in welfare rolls, the United States would have 66 percent fewer welfare recipients. There is no evidence that welfare caseloads increasingly comprise the hardest cases. Rather, it is evident that the most disadvantaged women have left the welfare rolls at rates comparable to, and in some states faster than, less disadvantaged women.

The welfare rolls could be reduced further if more states would adopt effective sanctions – swiftly withdrawing all cash benefits when recipients refuse to work. Requiring more welfare recipients to work, as President Bush and the House of Representatives have proposed, also would encourage more single mothers to make the transition to work. Conversely, loosening time limits and work requirements, as the Senate Finance Committee has proposed, would encourage women to continue collecting welfare. The end result would be more families dependent on welfare for years to come.

[page]

"Welfare reform has reduced the caseload by more than 50 percent."

Welfare rolls nationwide have fallen by more than 50 percent since the enactment of federal reforms in the 1996 Personal Responsibility and Work Opportunities Reconciliation Act (PRWORA). The law set a five-year lifetime limit for future receipt of federally funded welfare and required work by an increasing proportion of able-bodied adult welfare recipients. It also ended the federal welfare entitlement, replacing Aid to Families with Dependent Children (AFDC) with Temporary Assistance for Needy Families (TANF). The act gave states flexibility in reaching the goal of putting 50 percent of able-bodied adult welfare recipients to work or removing them from the rolls. The result was 50 state experiments in reform. Some of these have been very successful in reducing welfare rolls, while others have had little impact.¹

Congress is now debating reauthorization of the law. Since less than half of the caseload of able-bodied adults on welfare has been required to work, and welfare rolls have fallen little in some states, opportunities for substantial further reduction in welfare caseloads abound. Thus the Bush administration proposed, and the U.S. House of Representatives in May passed, a reauthorization bill requiring states to put more of those on welfare to work or remove them from the rolls.²

However, the goal of welfare reform was not simply to reduce the number of welfare cases. It was to move families on welfare – the vast majority of which are headed by single women – from dependency to independence through work. The two principal questions about the success of welfare reform are: Have the easiest-to-employ women left welfare, leaving the "hardest" cases behind? And are the families that have left welfare better off?

Before 1996, there was little interest in answering these questions.³ However, in recent years, separate U.S. Census Bureau and nongovernmental surveys have provided data on the employment and income of those who have left welfare. In addition, academic researchers have examined the experience of welfare "leavers" in specific states. Although research continues, the evidence already gathered and analyzed leads to some surprising conclusions.

[page]

"Have the hardest to employ been left behind?"

An important issue in the welfare reform debate is whether or not the easiest-to-employ women have left welfare, leaving behind those with lesser educations, fewer skills or other problems. Some have claimed to know intuitively that the remaining caseload included those who are hardest to employ.⁴ However, both national surveys and state data show that the women most at risk for long-term welfare receipt have left the welfare rolls at rates as fast as or faster than women who are much less at risk.

Thus the substantially higher spending in the welfare reauthorization bill passed by the Senate Finance Committee⁵ for transitional work, job training, day care, family support, education and housing to help those on welfare achieve independence is hard to justify. The Senate bill would in effect repeal the reforms.⁶ Under the current law, states may exempt up to 20 percent of their caseloads from the five-year federal time limit. The Senate bill would further weaken this lifetime limit by allowing the states to give welfare recipients "supplemental housing benefits" indefinitely. It would partially restore the entitlement to welfare by replacing the level payments to states under the reform law with increased funding when state caseloads or unemployment rise. And, most damagingly, it would effectively allow states to exempt 85 percent to 90 percent of adult recipients from work requirements. These measures would slow the transition of welfare recipients to work by significantly reducing incentives to leave welfare.

At-risk Single Mothers: The O'Neill-Hill Study. It is well established that the single mothers most at risk for lengthy welfare receipt and long-term welfare dependency are those who never married, are poorly educated, are young (18 to 29 years old), have young children (under age 7) and are black or Hispanic.⁷

"The decline in welfare participation has been greatest for the most disadvantaged single mothers."

Former Congressional Budget Office (CBO) Director June E. O'Neill and a colleague, M. Anne Hill, analyzed welfare use by different demographic groups over time, using data from the Current Population Survey (CPS) conducted in March of each year by the Census Bureau. They found that during the 1990s, and particularly since 1996, the decline in welfare participation was largest for groups of single mothers commonly thought to be the most disadvantaged. For instance, from 1993 to 1999,⁸ the percentage of disadvantaged single mothers who received welfare any time during the year fell dramatically – and by more than the fall in welfare receipt by less at-risk single mothers:

• Welfare participation declined for single mothers of all ages, but the decline was greatest for the youngest mothers, falling by 26 percentage points for 18- to 24-year-olds and 16 percentage points for 25- to 29-year-olds, with most of the change occurring since the 1996 reforms
• Welfare participation declined for single mothers with children of any age, but declined the most for mothers with children under age 7, falling by 22 percentage points.
  Welfare participation declined for single mothers regardless of their level of education but declined the most for mothers who had not completed high school, falling by 24 percentage points.
• Welfare participation declined for single mothers of every race and ethnicity but declined more for black and Hispanic women than for white women, falling by 24, 21 and 16 percentage points, respectively.
• Finally, welfare participation declined for single mothers regardless of their previous marital status, but it declined more for never-married mothers (27 percentage points) than for previously married mothers (15 percentage points).

"The steepest decline in welfare participation was among low-income black families."

At-risk Minorities: The Urban Institute Study. The Urban Institute's 2002 National Survey of America's Families looked at changes in welfare receipt by race and ethnicity from another perspective. Welfare enrollment for both blacks and Hispanics fell proportionately more than for whites over the 1996 to 2000 period, and the decline was steepest among blacks.⁹ As Figure I shows:

• The proportion of low-income white families who had received welfare benefits in the last year fell from 17 percent in 1996 to 10 percent in 2000.
• The proportion of low-income Hispanic families fell over the period from 25 percent to 14 percent – closely tracking the decline for all low-income families from 24 percent to 14 percent.
• And the decline for low-income black families was from 37 percent in 1996 to 22 percent in 2000.

Thus welfare receipt by low-income families in the major the ethnic groups fell by more than 40 percent.

Furthermore, the Urban Institute data showed a welfare caseload not increasingly made up of the women suffering the most hardships. For example, those who were still on welfare did not appear to have greater health problems than those who had left the rolls.¹⁰

"In Texas disadvantaged single women left the welfare rolls at rates similar to less disadvantaged women."

Other Measures of Disadvantage: Evidence from Texas. Data from the states show that those who have left the welfare rolls are similar to those who have remained. For example, a comparison of Texas enrollees who left in 1997 and 1998 and those who remained up to 2002 shows that disadvantaged single mothers have left the state's welfare rolls at rates similar to those of other women who have left the rolls.¹¹

Hispanic and black women – who made up a larger proportion of the Texas welfare caseload than whites – left the rolls at higher rates than did white women, and as a result the percent of the remaining welfare caseload comprising minorities is smaller, while the percent of whites is larger.

The proportion of the Texas welfare caseload made up of women with less than eight years of education fell from 1998 to 2002, as did the proportion who were high school graduates and above, while the proportion who were high school dropouts increased only slightly.

Thus the Texas welfare caseload does not show an increasing concentration of women with multiple disadvantages or hardships.

Other Measures of Disadvantage: The Johns Hopkins Study. For other disadvantages, researchers using different data sets have come to complementary conclusions. For example, a Johns Hopkins University study of Boston, Chicago and San Antonio found that women under age 25 left the rolls at a faster rate than older women over the 1997 to 1999 period. (Younger women comprised 35 percent of those who left and 30 percent of those who stayed.)¹² The study also found that in terms of education levels, depression and domestic violence, women who remained on the rolls did not differ greatly from those who left.

Long-term Welfare Recipients: National Data. Under the old AFDC program, nearly half of welfare recipients at any point in time had been on the rolls for five or more years. Yet most families who received cash welfare benefits did so for only a few months. About half of those who went on welfare left within a year, 70 percent within two years and almost 90 percent within five years.¹³ However, the longer families remained on the welfare rolls, the more likely they were to stay there for additional years. As a result, long-term welfare recipients accounted for more than half the welfare rolls. Since reform, U.S. Department of Health and Human Services data show how far the proportion of long-term welfare recipients has fallen. [See Figure II.]

• Under the old AFDC system, about 70 percent of recipients at any point in time had received AFDC for more than two years;¹⁴ by 1999, only 47 percent of welfare recipients nationwide had received benefits continuously for two or more years.¹⁵
• Under the old AFDC system, 48 percent of the caseload at any point in time had received assistance for more than five years;¹⁶ by 1997, only 25 percent had received benefits continuously for five years or more.¹⁷

Under AFDC, about one-third of women who ever used welfare would spend longer than five years on the rolls over the course of their lifetimes and 60 percent would receive welfare for at least 24 months. Since the 1996 reforms set a five-year limit on welfare receipt, such long-term receipt will fall over time.

Long-term Welfare Recipients: Evidence from Baltimore. A study of Baltimore, Maryland, welfare recipients found that the long-term welfare dependency of women ages 19 to 24 fell abruptly after welfare reform. In the 1992 to 1996 period, 46 percent of young women were longer-term recipients, while in the 1996 to 2000 period only 15 percent of them were long-termers.¹⁸ The same study found that the labor market skills or earning potential of women on welfare have increased since reform, indicating that the caseload is not increasingly made up of women with less marketable skills.

"Long-term welfare recipients were more than half the caseload under the old program…since reform they are less than half."

Long-term Welfare Recipients: Evidence from Wisconsin. What proportion of long-term welfare recipients can be moved from welfare to work? We do not know, but in Wisconsin – which reduced its welfare rolls by more than 70 percent – only 2 percent of women who had been on the welfare rolls in 1990 were either still receiving welfare or had returned to the rolls after some work experience by 1998.¹⁹

[page]

Disadvantaged women and their children may have left the welfare rolls at rates similar to or even faster than women with fewer disadvantages, but are the ones who have left welfare better off?

"Employment of welfare leavers nationally rose from 75 percent in 1997 to 79 percent in 1999."

More Employment. In a typical year under the old AFDC program, between one-half and two-thirds of those who left welfare found paid employment.²⁰ Under the new TANF program, the proportion that found employment has been increasing. Researchers compared adults on welfare in 1997 with those on the rolls in 1999, using data from the Urban Institute survey. They found that the share of welfare leaver families with at least one employed parent increased over that period – from 75 percent in 1997 to 79 percent in 1999. In former recipient families with spouses or partners, about 90 percent had at least one adult working.²¹

In the states studied in depth, most welfare leavers had found employment, increased their incomes relative to welfare recipients and were gradually moving up the income ladder. For example, Floridians who received benefits from 1996 to 1998 were surveyed in November 1998 and again in March to May 2000.²² Among the survey findings:

• The proportion of former welfare recipients who had found a job after leaving welfare increased from 76 percent in the first survey to 88.8 percent in the second.
• Welfare leavers who currently were employed increased from 56.9 percent to 65 percent, and they were working an average of 35.1 hours per week because most had full-time jobs.

Employment Rates by Race and Ethnicity: State Studies. The Florida surveys found that a higher proportion of blacks and Hispanics than whites had found employment:

• White people were less likely than nonwhites²³ to have found a job (85.1 percent compared to 91.2 percent for minorities) and whites had been employed less often in the preceding 12 months than nonwhites (7.4 and 8.2 months, respectively).
• Minorities had higher employment rates although the level of education attainment for the minority groups was lower – 33.5 percent of whites had education beyond high school graduation, compared to 23.1 percent of nonwhites.

Another analysis, based on Florida administrative data for those who left TANF in the second quarter of 1997, found that 21 months afterward:

• The average quarterly employment rate of blacks was higher (at 59.7 percent) than that of Hispanics (47.9 percent) or whites (45.9 percent).
• The percentage that had been employed at any time in the preceding two years was higher for blacks (78.2 percent) and about the same for Hispanics (65.5 percent) and whites (65.9 percent).²⁴ [See Figure III.]

Other studies sponsored by the U.S. Department of Health and Human Services found similar results. Of these leaver studies, only six reported employment rates by racial and ethnic groups. Five of the six studies found that black leavers had higher employment rates than white leavers. [See Table I.]²⁵

• At any time in the year after exit, the black employment rate exceeded the white rate among welfare leavers by 6 percentage points in Arizona and 14 percentage points in Florida.
• Arizona, Missouri and South Carolina reported similar differences for employment in the fourth quarter after exit.
• Only in Illinois were the employment rates slightly higher for whites (56 percent) than for blacks (54 percent), and this difference may have been due to differences in employment rates between Chicago and downstate Illinois rather than to race.²⁶

Of the three studies that reported separate employment results for Hispanics, two (Arizona and Illinois) found employment rates of Hispanic leavers were similar to those of black leavers. In Florida, the employment rate of Hispanic leavers at any point in the year after exit (48 percent) was similar to that of whites and lower than that of blacks.

"Five of six state-level studies found minority leavers had higher employment rates than whites."

More Income. Researchers compared adults who had been on welfare in 1997 with those on the rolls in 1999, using data from the Urban Institute survey. They found that a year after exit the average household income of welfare leavers was about 125 percent of the poverty level, and 10 percent of households were above 250 percent of the poverty level.²⁷ Furthermore, the monthly average poverty rate of more recent leavers declined significantly, from 48 percent in 1997 to 41 percent in 1999. In Baltimore, researchers found that poverty rates (not including noncash benefits) were 70 percent for leavers, compared to 85 percent for "stayers."²⁸

Between 1998 and 2000, the proportion of Florida welfare leavers earning low wages (between $5.15 and $7.99 per hour) declined by 7 percent, while those earning more than $8 an hour increased by 24.1 percent.²⁹ In Iowa, after two years, more than 60 percent of family heads earned an average wage of $8.16 per hour – more than enough to lift them above the poverty level.³⁰

Earnings by Race and Ethnicity: State Studies. Of the state leaver studies, only six reported earnings by racial and ethnic groups. The evidence on earnings was mixed. [See Table I.]³¹

• The Florida and Georgia studies both found that mean annual earnings a year after exit were higher for blacks than for whites – about $700 higher in Florida and $900 higher in Georgia.
• Earnings for Hispanic leavers in Florida were significantly higher than for blacks or whites, at $7,732 a year.
• In the fourth quarter after exit, Missouri found that all nonwhite leavers (nearly all black) had mean quarterly earnings almost $400 higher than white leavers.
• In Arizona, Hispanic leavers had lower earnings in the fourth quarter after exit ($2,487) than black leavers, and white leavers had higher earnings than blacks.
• South Carolina found that working white leavers had higher mean monthly earnings than working black leavers.

Less Recidivism. Those who have ever received welfare are at higher risk for returning to the welfare rolls than nonrecipients. As noted previously, under the old AFDC program, many families left welfare after a few months, but returned to the rolls.³² Nationally, researchers compared adults on welfare in 1997 with those on the rolls in 1999, using data from the Urban Institute survey:

• In both 1997 and 1999, about one-quarter of adults on welfare were new to the rolls, about one-fifth had received cash assistance intermittently during the previous two years, and almost half were long-term recipients.³³
• By comparison, under the old AFDC program, about 45 percent of welfare leavers returned within a year and 70 percent returned by the end of five years.
• State studies also found that recidivism rates had fallen; for example, one study found that in Iowa, only one in five families was back on welfare after two years.³⁴

In California, researchers found that compared to the old AFDC program, fewer returned to the rolls within 18 months of leaving welfare. Among single parent families, 17 percent of 1998 welfare leavers returned within a year, compared to 27 percent of 1993 leavers and 25 percent of 1988 leavers. (Researchers note that economic growth in California in 1988 was at least as strong as in 1998.)³⁵ [See Figure IV.]

Less Poverty. Since the 1996 welfare reforms, the number of Americans living in poverty has fallen 21 percent; and the annual incomes of the poorest women have increased nearly $1,000 – after adjusting for inflation and not including the Earned Income Tax Credit, which was doubled in 1996. The proportion of their income from personal earnings jumped from 26 percent to 36 percent, while the proportion of their income from welfare fell from 53 percent to 37 percent. More importantly, the number of children in poverty declined from 16 million in 1993 to 11.5 million in 2000.³⁶

Other Evidence. Welfare leavers themselves believe they are better off. [See Figure V.] Regardless of their employment status – and with no significant variation by race or ethnicity – 18 months after leaving welfare nearly 60 percent of of Florida welfare leavers said they were better off. Asked if their children were better off since leaving welfare, almost two-thirds of leaver parents said they were; a little more than a quarter said they were about the same, and only 5.5 percent said they were worse off.³⁷

Overall, the evidence is that welfare leavers do not face significantly greater problems than they did while on welfare. A Joyce Foundation study found that some who have left welfare continue to face such financial hardships as having a utility cut off or occasionally being late with a rent check. However, the study found that moving from welfare to work does not necessarily increase such hardships. Michigan researchers found that those who had left welfare and were relying on wages were better off by several criteria than those who were still on welfare or who had left welfare but had no jobs.³⁸ Florida researchers found that whites were less likely than blacks to have employer-provided health insurance (31.5 percent to 39.6 percent).³⁹

[page]

As welfare rolls fell by more than 50 percent during the years following reform, those who had opposed TANF had difficulty explaining the apparent success of the program. In the past, welfare rolls grew during good economic times as well as recessions. And new federal initiatives, such as job training programs and disregarding wages from work for eligibility purposes actually increased the number of welfare recipients. The answer may be found in the strategies some states used to reduce their rolls by more than 90 percent, while other states were reducing theirs by less than 25 percent.

"During the 1990s, the good economy accounted for less than 20 percent of the drop in the welfare rolls."

Did the Booming Economy Explain the Drop in Welfare Rolls? During the boom years of the late 1990s, welfare reform critics claimed that the good economy was primarily responsible for the fall in welfare rolls. They predicted that if the good times ended, welfare rolls would rise as newly unemployed families returned to welfare. Prior to reform, welfare caseloads rose even during periods of increasing employment, whereas after reform welfare rolls continued to fall in most states and increases were negligible in others. This was true even after unemployment rose with the onset of recession in 2001. In fact, rigorous statistical analyses confirmed the common wisdom that welfare rolls in the past had grown during both recessions and booms, and welfare receipt had little to do with the economy at large.

Economists June O'Neill and Anne Hill, using data from the Census Bureau's CPS, found that:⁴⁰

• The TANF program created in 1996 has accounted for more than half of the decline in welfare participation since then and for more than 60 percent of the rise in employment among single mothers. By contrast, the contribution of the booming economy of the late 1990s to these declines was relatively minor, accounting for less than 20 percent of either change.
• Employment gains have also been the largest among disadvantaged single mothers, and TANF accounts for much of these gains: 40 percent of the increase in work participation among single mothers who are high school dropouts, 71 percent of the increase in work participation among single mothers ages 18 to 29, and 83 percent of the increase in work participation among black single mothers.

There is also evidence that those states with relatively lower benefit levels have had the most success in reducing caseloads. This is not unexpected, since the point of welfare reform was to change incentives between welfare and work. Cato Institute researchers found that "those states with the strongest sanctions and the lowest benefit levels had the most success in reducing their caseloads."⁴¹

"Potential recipients diverted from welfare are found to be ineligible, or they qualify for other programs or opt to receive a lump-sum cash
payment."

Are Potential Welfare Recipients Being Diverted to Other Programs? Some allege that caseload reductions have been achieved by diverting families from applying for welfare. Actually, diversion is an explicit and appropriate policy in most states. Many of those diverted are ineligible, usually because at least one parent is working. Some are found to be eligible for more generous federal programs, such as Supplemental Security Income (SSI), which provides cash assistance and free medical care to disabled children, regardless of their parents' work history. Other divertees opt to receive a lump-sum cash payment of $1,000 to $2,000 in lieu of monthly cash benefits – an alternative authorized by the 1996 reform law. Follow-up surveys of those who were diverted show that most soon found employment.

Are Sanctions Discriminatory? Welfare reform critics have claimed that minorities are disproportionately sanctioned, or "kicked off" welfare. However, there is evidence that many sanctioned families – families that have lost their benefits due to noncompliance – were actually voluntary welfare leavers.

For example, in Arizona, African-Americans were 13 percent of cases closed due to sanctions and just 9 percent of cases closed due to earnings. Hispanics were 39 percent of cases closed due to sanctions and just 9 percent of cases closed due to earnings. However, a survey of welfare leavers in Arizona found a much lower percentage reporting that they left welfare due to sanctions than was indicated by administrative data, supporting the conclusion that they found work, and having no incentive to do so, did not report their new status. They just stopped showing up for appointments with caseworkers.⁴²

A survey of families whose benefits had been terminated a second time found that two to four months later, despite the loss of their welfare checks, half had experienced an increase in household income averaging $758 a month – more than double their income before the sanction. Only one-third reported a reduction in household income (averaging $384).⁴³

[page]

"Less than half of welfare recipients have been subject to enforceable work requirements."

As previously noted, although national welfare rolls have fallen by more than 50 percent under the federal reforms and some states have reduced their rolls substantially, other states have reduced their caseloads by less than 25 percent. [See Table IIA.] [See Table IIB.]

How Much Do Sanctions Matter? The most significant factor in caseload reduction is the policies established by the states – specifically, whether or not the state has an effective sanctions policy. Effective sanctions terminate benefits for noncompliance. Sanctions may be imposed when recipients refuse work, fail to keep appointments or refuse to document illnesses they claim resulted in unexcused work absences. Some states sanction recalcitrants but reduce benefits by only a token amount. Some states have such an extensive appeals process (during which welfare payments continue) that their sanctions are ineffective.⁴⁴

Yet effective sanctions rather than state unemployment rates, demographics or the proportions of disadvantaged women on welfare best correlate with caseload reductions.⁴⁵ Only effective sanctions compel welfare recipients to comply with program requirements. More than half of welfare recipients have not been subject to enforceable work requirements. The less successful states have had ineffective sanctions, high benefit levels and other counterproductive policies. For instance, through prolonged appeals processes, very modest benefit reductions or transferring cash benefits to the children of sanctioned parents, 17 states still allow recipients to receive welfare even if they refuse to participate in required activities.

Two of these states, New York and California, account for nearly a third of the remaining welfare caseload.

New York. In New York state, if welfare recipients refuse to comply with work requirements, their benefits are reduced by a small amount (from $588 per month to $475 for a family of three), but they continue to receive food stamps, public housing and other benefits. As a result, "recent figures show that, of those adults who are assigned to an activity, only 54 percent participate as their assignment requires."⁴⁶

"New York and California have ineffective sanctions and account for nearly a third of the remaining caseload."

California. California offers high benefits and ineffective sanctions. Much of the fall in federal welfare rolls (TANF) in California was due to the state transferring all two-parent families to a program that uses discretionary state funds, does not have to meet federal requirements and does not count in the TANF caseload. If the state had adopted the welfare policies of an average state, 200,000 fewer Californians would be receiving welfare.⁴⁷

California and some other states also allow recipients to combine benefits with a high level of wage income. This has created incentives to remain on welfare. The Urban Institute found that 20 percent of welfare recipients combined welfare and work in 1999, compared to 5 percent in 1997.⁴⁸

Without any changes in federal law, the states that have failed to reduce their welfare rolls by as much as the average could adopt more effective policies. The impact would be significant:

• The average (and median) state caseload decline from 1993 to 2000 was about 59 percent.
• If the 23 less-than-average states had done as well as the average state, more than 800,000 additional people would have left welfare.
• Instead of a 59 percent reduction in welfare rolls, the United States would have 66 percent fewer welfare recipients.⁴⁹

This would amount to a reduction of about 15 percent in the more than five million Americans who now collect welfare.

[page]

Welfare reform has changed the incentives governing the trade-off between welfare and work. Requiring able-bodied adult recipients to work, sanctioning those who do not and imposing time limits for welfare receipt – all have helped to make welfare less attractive. The welfare reforms are the principal reason why the rolls nationwide have declined by more than 50 percent since 1996.

"If the 23 states that had smaller caseload declines than the national average had met the average, 800,000 more people would have left the welfare rolls."

Most of the women who have left welfare have found work and have been able to earn more than they collected. A minority of the women who have successfully left welfare have not gone to work. Some have returned to school or returned home to live with their parents in order to raise their children. At present, nearly one-quarter of families on welfare are "sub-families" living with relatives, other families receiving welfare or an unrelated employed householder. Thus many of them have resources not captured by most administrative data.

There is no evidence that welfare caseloads increasingly comprise the hardest cases. Rather, it is evident that the most disadvantaged women have left the welfare rolls at rates comparable to, and in some states faster than, less disadvantaged women.

The welfare rolls can be reduced further by the adoption of effective sanctions when recipients refuse to comply with work requirements. Exposing more of the welfare caseload to work requirements, as President Bush and the House of Representatives have proposed, also would encourage more single mothers to make the transition to work. Conversely, loosening time limits and work requirements, as the Senate Finance Committee has proposed, would encourage women to continue collecting welfare. The end result would be more families dependent on welfare for years to come.

NOTE: Nothing written here should be construed as necessarily neflecting the views of the National Center for Policy Analysis or as an attempt to aid or hinder the passage of any bill before Congress.

[page]

1. Many states had begun implementing reforms as early as 1993, using federal waivers from AFDC program requirements. One of the earliest and best-known reform efforts was in Wisconsin, under the leadership of then governor and current U.S. Health and Human Services Secretary Tommy Thompson.
2. The Personal Responsibility, Work, and Family Promotion Act, H.R. 4737, passed the House on May 16, 2002.
3. For example, "Because relatively little attention was paid to welfare leavers in the past, there are no directly comparably [sic] figures from before welfare reform." Elizabeth Lower-Basch, "'Leavers' and Diversion Studies: Preliminary Analysis of Racial Differences in Caseload Trends and Leaver Outcomes," rev. December 2000, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, page 7.
4. See Thomas Corbett, "The New Face of Welfare; From Income Transfers to Social Assistance?" Focus, Vol. 22, No. 1, Special Issue 2002, Institute for Research on Poverty, University of Wisconsin, page 5.
5. The Work, Opportunity, and Responsibility for Kids Act was approved by the Senate Finance Committee in June 2002. Action on the bill by the full Senate was pending as of September 16, 2002.
6. Robert Rector, "The Bacus 'WORK' Act of 2002: Repealing Welfare Reform," Backgrounder No. 1580, September 3, 2002, Heritage Foundation.
7. June E. O'Neill and M. Anne Hill, "Gaining Ground: Women, Welfare and Work," NCPA Policy Report No. 251, February 2002, National Center for Policy Analysis.
8. Welfare receipt peaked in 1994, and some states began implementing federal government waivers soon thereafter.
9. Kenneth Finegold and Sarah Staveteig, "Race, Ethnicity, and Welfare Reform," in Alan Weil and Kenneth Finegold, eds., Welfare Reform: The Next Act (Washington, D.C.: Urban Institute Press, 2002), pages 205-06.
10. Pamela J. Loprest, "How Are Families Who Left Welfare Doing Over Time? A Comparison of Two Cohorts of Welfare Leavers," Economic Policy Review, September 2001, page 11, Federal Reserve Bank of New York.
11. Author's calculations, based on "Adult TANF Client Demographics" for 1997, 1998 and 2002, Texas Department of Human Resources, and Deanna Schexnayder et al., Texas Families in Transition, January 2002, Texas Department of Human Services, Appendix A, Tables A-4 and A-5, pages A-18 to A-22.
12. Robert Moffit et al., "The Characteristics of Families Remaining on Welfare," Policy Brief 02-2, March 2002, Welfare, Children, and Families Study, Johns Hopkins University, Table 1, page 2.
13. LaDonna A. Pavetti, Urban Institute, "Time on Welfare and Welfare Dependency," Statement to the Human Resources Subcommittee of the House Committee on Ways and Means, May 23, 1996.
14. Pavetti, "Time on Welfare and Welfare Dependency."
15. Pamela J. Loprest, "Making the Transition from Welfare to Work: Successes but Continuing Concerns," in Weil and Finegold, eds., Welfare Reform: the Next Act, page 23.
16. Pavetti, "Time on Welfare and Welfare Dependency."
17. Loprest, "Making the Transition from Welfare to Work," page 23.
18. "Long-term" in this study was defined as 21 months or more during the five-year period. See Robert A. Moffit and David W. Stevens, "Changing Caseloads: Macro Influences and Micro Composition," Economic Policy Review, September 2001, pages 43-44, Federal Reserve Bank of New York.
19. Sammis B. White and Lori A. Geddes, "Disappearing Wisconsin Welfare Recipients: Where Did They Go?" WPRI Report, Vol. 14, No. 9, December 2001, Wisconsin Policy Research Institute, page 4.
20. Pavetti, "Time on Welfare and Welfare Dependency."
21. Loprest, "How Are Families Who Left Welfare Doing over Time?" page 13.
22. Robert E. Crew Jr., Steve McDonald and Johnny Johnson, "Following WAGES Clients Over Time," September 2000, College of Social Sciences, Florida State University, pages 1-17.
23. The category nonwhite includes both African-Americans and Hispanics, who may be of any race. The study separately tallied results by primary language and, unsurprisingly, non-English speakers had lower rates of employment.
24. Robert E. Crew Jr. et al., "Tracking the Outcomes of Welfare Reform in Florida for Three Groups of People," October 2000, College of Social Sciences, Florida State University, Table 4, page 14.
25. Gregory Acs, Pamela Loprest and Tracy Roberts, Final Synthesis Report of Findings from ASPE's "Leavers" Grants, December 2001, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, Table III-11.
26. Acs, Loprest and Roberts, Final Synthesis Report of Findings from ASPE's "Leavers" Grants, Chapter 3, "Cross-State Findings on Employment and Earnings."
27. Press Release, "Characteristics of Recent Welfare Population Dispel Some but Not All Concerns as Time Limits Approach," April 27, 2001, Urban Institute.
28. Moffit et al., "The Characteristics of Families Remaining on Welfare."
29. Crew Jr. et al., "Tracking the Outcomes of Welfare Reform in Florida for Three Groups of People," Table 4, page 14.
30. Jacqueline Kauff, Thomas Fraker and Julita Milliner-Waddell, "Iowa Families That Left Welfare: How Are They Faring Two Years Later?" 2002, Mathematica Policy Research.
31. Acs, Loprest and Roberts, Final Synthesis Report of Findings from ASPE's "Leavers" Grants, Chapter 3.
32. Pavetti, "Time on Welfare and Welfare Dependency."
33. Loprest, "Making the Transition from Welfare to Work," page 23.
34. Kauff, Fraker and Milliner-Waddell, "Iowa Families That Left Welfare."
35. Charles Lieberman and David C. Mancuso, "Are California's Welfare Leavers Staying Off Public Assistance? Welfare Recidivism and Use of Non-Cash Aid Since Welfare Reform," California Policy Review, October 2001, SPHERE Institute, page 4.
36. Joe Barnett, "Welfare Reform: Reasons to Stay the Course," Brief Analysis No. 401, June 24, 2002, National Center for Policy Analysis.
37. Crew, McDonald and Johnson, "Following WAGES Clients Over Time," page 18.
38. "Welfare to Work: What Have We Learned?" April 2002, Joyce Foundation.
39. Crew, McDonald and Johnson, "Following WAGES Clients Over Time," page 18.
40. O'Neill and Hill, "Gaining Ground."
41. Michael J. New, "Welfare Reform That Works: Explaining the Welfare Caseload Decline, 1996-2000," May 7, 2002, Cato Institute, Executive Summary.
42. For example, the extensive appeals process set up by New York state is a problem in New York City. See Jo Anne Barnhart, Deborah Chassman and Sandie Hoback, "Moving from Full Engagement to Full Employment: A Program Review of New York City's Welfare Reform," November 5, 2001, American Institute for Full Employment, pages 14-15.
43. Robert E. Rector and Sarah E. Youssef, "The Determinants of Welfare Caseload Decline," CDA99-04, May 11, 1999, Center for Data Analysis, Heritage Foundation.
44. Asked why their cases were closed, the majority of Arizona respondents of all races said they got a job or a better job – 63.8 percent of African-Americans, 60 percent of Native Americans, 55.7 percent of non-Hispanic whites and 52.9 percent of Hispanics. In Washington state, two-thirds of both whites and nonwhites said that increased income through employment was their primary reason for leaving welfare. This reason was cited far more frequently than the second most prevalent, which was that their income from other sources had increased. Elizabeth Lower-Basch, "'Leavers' and Diversion Studies," page 7.
45. Douglas J. Besharov and Peter Germanis, "Welfare Reform After Five Years," Policyforum No. 1, October 2001, page 10, Acton Institute.
46. Jason Turner, "Don't Let California and New York Undermine Welfare Reform's Work Requirements," Executive Memorandum No. 819, June 13, 2002, Heritage Foundation.
47. Thomas E. MaCurdy, David C. Mancuso and Margaret O'Brien-Strain, "Does California's Welfare Policy Explain the Slower Decline of Its Caseload?" 2002, Public Policy Institute of California; summarized in "What Accounts for Recent Declines in Welfare Caseloads?" Research Brief Issue No. 53, January 2002, PPIC.
48. "Characteristics of Recent Welfare Population Dispel Some but Not All Concerns as Time Limits Approach," Urban Institute.
49. Calculations by Devon Herrick, NCPA research manager, based on U.S. Department of Health and Human Services data.

[page]

Joe Barnett is Director of Publications at the National Center for Policy Analysis, where he is responsible for the institute's publishing program, working with academic authors, independent experts and the NCPA policy staff in the writing, editing and printing of NCPA books, studies and periodicals. He edits Executive Alert, the NCPA's bimonthly compilation of the best in public policy research.

Barnett also edits Daily Policy Digest, the NCPA's e-mail and Web-based newsletter summarizing public policy research news for opinion leaders. As an NCPA analyst and editor, he has conducted research, and written and edited studies covering economic, education, welfare, tax and regulatory policies.

Prior to joining the NCPA in 1995, Barnett held various positions with Ernst & Young and McGraw-Hill. He also served three years in Washington, D.C., as a legislative assistant to U.S. Representative Ron Paul (RTexas). He received a Bachelor of Arts degree in English from the University of Texas at Arlington.