Maxwell School, Syracuse University

The Maxwell School of Citizenship and Public Affairs


PAI 786 - Urban Policy
Professor Yinger



Case: Welfare Reauthorization[1]


The major federal welfare overhaul passed in 1996, called the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), expired on October 1, 2002.  The House of Representatives and Senate could not agree on a reauthorization package, so they passed a six-month extension of the original legislation.  It is now March 2003, and Congress has taken up the welfare reform once more.  The House got the ball rolling by passing the reauthorization bill it passed in 2002, which is similar to the plan favored by President Bush.  The debate has now moved to the Senate. 

PRWORA eliminated Aid to Families with Dependant Children (AFDC), which was originally created in the 1930s, and replaced it with Temporary Assistance to Needy Families (TANF).  This new program put a five-year limit on welfare receipt.  Moreover, it strengthened the work requirements for welfare recipients, which were first introduced by the 1988 Family Support Act.  Under AFDC, all families that met certain eligibility requirements were entitled to receive benefits.  This entitlement feature was eliminated by TANF.  Not only did TANF cut off payments after five years, it also limited federal responsibility to the provision of block grants to states, with no requirement that states continue to pay recipients if their federal block grants ran out.  More detailed information about TANF is provided in the appendix. 

Although these provisions appear draconian, they were combined with several other provisions that greatly softened the impact of the legislation.  The TANF requirement that welfare recipients find “work” within two years, for example, allows states to define work however they please—one hour of basket weaving each year would do—and there aren’t any penalties for missing the target.  In addition, the TANF provisions specifying the share of their cases that must be in “work” have been largely ineffective because states have been able to count caseload declines toward these targets.[2]  Moreover, up to 20% of each state’s annual caseload can be exempted from the five-year benefit limit on hardship and other grounds.  In addition, programs financed exclusively with state funds are not covered by the five-year limitation on benefits.   

The House Bill 

On February 13, 2003, the House passed the same re-authorization legislation for TANF that it passed in 2002.  The House bill would increase the required level of work from 30 hours per week to 40, and education would no longer count as a core “work activity.”  The bill would, however, allow people to spend up to two days a week in other activities to help them hold jobs, such as vocational training or drug treatment.  The measure would gradually increase the percentage of adults on welfare who must hold jobs from 50 percent to 70 percent by 2008. In addition, the bill would devote $300 million to grants intended to help states foster healthy marriages, and continue a $50 million yearly subsidy to encourage people to abstain from sex before marriage.  It would also continue a ban on cash assistance for many legal immigrants.  Finally, the bill adds $2 billion over five years to the $4.8 billion annual grant the government gives states to help poor families afford child care. Half of that $2 billion could be omitted, if lawmakers decide there is not room for it in the federal budget. Rep. Deborah Pryce (R-Ohio) said the extra money would ensure “recipients can rest in peace that their children are safe as they train for, find and keep a stable job.”


Opinions about the best course of action for the Senate vary widely.  Not surprisingly, House Ways and Means Committee Chairman Bill Thomas (R-Calif.) extolled the bill, saying it would “move people in the right direction: the work direction.”  President Bush commended the House, saying its “compassionate approach will help many more Americans realize a better life of independence, hope and dignity that comes with having a job.” 

Ron Haskins, who is a senior fellow at the Brookings Institution and who helped write the 1996 legislation as the Majority Staff Director at the House Ways and Means Committee, also is a supporter of the House bill.  In an interview on NPR news (2003), he said

The ’96 law had what I think in retrospect we would call a flaw in it, so as a result, the typical state has zero work requirement because the caseload has fallen so much.  So that has to be repaired and the president proposes to repair that.  The president, of course, also wants to have a marriage initiative, and if it passed the way it was written last year, it’d be $300 million a year for states that wanted to try to improve strong and healthy marriages among low-income families in the state.

 Mr. Haskins also defends the House bill’s emphasis on work.

The problem in welfare is people don’t work, not that they don’t go to college.  So the focus of the legislation has not been on education.  That had been tried many times in the past and had failed utterly.  The focus was to help people get in the work force.  So I think the issue of education, it will be a part of the debate, but there’ll be a real attempt by the president and by Republicans in Congress and now by many Democrats as well to limit the amount of education through welfare.  If people want to go and get education through any other number of programs, that’s fine, but not through welfare.  Welfare is a program to help people work, not to send people to college. 

In contrast, many Democrats said the proposed work rules would prompt states to force people into “make-work” positions, at the expense of training that would lead the poor into more meaningful and lucrative jobs. Rep. Steny H. Hoyer (D-Md.) called the measure “hardhearted,” saying that, in an economic downturn when jobs have grown scarcer, “it turns a blind eye to the most obvious question facing any job-seeker today: Where?” 

Peter Edelman, who teaches law at Georgetown University and served in the Department of Health and Human Services in the Clinton Administration, also thinks the House bill takes the wrong approach. 

The promise of welfare reform was that if people went to work, they would
get help with child care and Medicaid and other work supports.  There were many places where the promise wasn’t kept, and in the current fiscal crisis, states are reneging on the deal right and left.  And time limits, a looming issue from the outset, are hitting in more and more states, disqualifying families at the worst possible time, right in the middle of a recession.  Especially in a recession, we should be increasing the funding for people who are down on their luck.  President Bush doesn’t think so.  He proposes to force nearly three-quarters of current recipients into costly make-work programs.  As a result, states will have to give up the very incentives that have helped move people into real jobs.  More flexibility for recipients to go to college, not allowed.  More flexibility to help legal immigrants who are still struggling, not available.  More help to continue health coverage for low wage workers coming off welfare, not.  More attention to child well-being by reforming child support rules—Who are we kidding?  (Edelman, 2003).  

Evidence on the Impacts of the 1996 Legislation 

Since the 1996 legislation was passed, the number of people on welfare plunged to five million in September 2002 from 12.2 million in 1996.  There is widespread disagreement about the meaning of this plunge, however.  Some commentators attribute it mainly to the healthy economy of the late 1990s, while others attribute it mainly to welfare reform.


Moreover, people disagree about the impact of welfare reform on the well-being of the families receiving welfare in 1996.[3]  Some people argue that welfare reform produced large income gains for these families.  According to Mr. Hastert (2003), for example,  

 Low income, female-headed families have increased their overall income by around 15 to 20 percent, and the reason is that there earnings have increased dramatically and the welfare income has fallen.  If you think of this—welfare down, earnings up—that is the very definition of ending welfare dependency, so the law has worked very, very well.  Now there are some families that are undoubtedly worse off, because in the old days they used to be able to stay on welfare for many, many, many years.  Now it’s difficult to do that because the states run more rigorous programs. 

Rep. Rahm Emanuel (D-Ill.), who helped write the original welfare overhaul as a senior aide to President Bill Clinton, said the law had succeeded by “linking a generation of children to the culture of work.”[4]

Moreover, a recent study financed by the National Institutes of Health and recently published in the highly regarded journal Science found that poor children suffer no psychological damage when their mothers move from welfare to work, as millions of women have in recent years.  Among adolescents in such families, the study found, mental health may actually have improved.  More specifically, this study observed preschoolers and adolescents in 2,400 families in Boston, Chicago and San Antonio. They found that the mothers' transition from welfare to work did not harm the emotional development, intellectual achievement or mental health of the children.

The researchers said they had detected neither harmful nor beneficial effects on preschool children when their mothers went from welfare to work.  “The positive and negative aspects of going off welfare or getting a job may cancel each other out” in these families, said P. Lindsay Chase-Lansdale, a professor of psychology at Northwestern University, who directed the study.  When mothers of preschoolers went to work, the study found, family incomes increased, but the mothers' time with the children, 2 to 4 years old, decreased. “These two effects may have offset each other,” the report said.

 The researchers also said adolescents whose mothers had begun working reported small but “statistically significant declines in psychological distress,” including symptoms of anxiety. Some adolescents also reported significant declines in delinquent behavior like the use of drugs and alcohol after their mothers left welfare.  These results reflect a different pattern of effects in families with adolescents; family income increased when their mothers went to work, but the women did not substantially reduce the time spent with the children, ages 10 to 14. 

In interpreting these changes, the study said that “The self-esteem of mothers of adolescents often significantly increased when those mothers went to work and often significantly decreased when they left work.”  Adolescents seemed to thrive when their parents had a positive self-image.  Moreover, “adolescents are perceptive and sensitive to the pressures of poverty and economic hardship in their families,” so their “anxiety levels may decrease as they see their mothers going to work each day.”  Conversely, the researchers said, when mothers leave the labor force, adolescent children are more likely to display behavioral problems. “Teenagers may express their feelings of disappointment or worry about finances as depression or anger when their mothers leave employment.” 

Finally, this study found that when mothers went to work, their household income nearly doubled, and most of the families were lifted out of poverty. Before they went to work, their average income was 35 percent below the poverty level. After they took jobs, their income was more than 25 percent above the poverty level.  “For a family of four, that represents an increase of $10,000 a year,” Prof. Chase-Lansdale said. 

“This study provides reassurance that mothers may leave welfare and enter the job market without harmful effects to their preschoolers or young adolescents,” said Dr. Duane Alexander, director of the National Institute of Child Health and Human Development, a unit of the N.I.H.  Dr. Chase-Lansdale also said, however, that the changes in the lives of the families in the study had occurred in a booming economy. In a weak economy, she added, it might have been more difficult for women to leave welfare and find jobs, and the jobs might have paid less.   

Other scholars do not see the impacts of the 1996 reforms in such a positive light.  They emphasize that many of the previous recipients remain in poverty or that the apparent gains in income among recipients will disappear when the economy moves into recession, as it now apparently has.  One study (Scott, et al., 2002) conducted in-depth interviews of 75 women in Cleveland and Philadelphia in 1998 and 1999.  These women had all been AFDC recipients.  On first glance, these women appeared to be doing quite well under welfare reform.  As the authors put it,  

Overall, 70% of the mothers saw an income gain by the second interview in 1999, with an average monthly gain of $233. Income includes all sources: earnings from work, cash assistance, food stamps, child support, Social Security, survivor’s benefits, disability, emergency assistance, heating assistance, unreported work, the Earned Income Tax Credit (EITC), and support from family and friends. Not surprisingly, women who were working (either working fully, or combining work and welfare or cycling) saw higher income gains than other mothers. Among women in the work-only category, the average change was a $482 increase. The average change for cyclers was $215, and $195 for combiners. Those neither working nor receiving welfare were least likely to gain and the most likely to have lost income between the interviews.

These scholars also concluded, however, that “Looking at income alone masks the reality of daily life for single mothers complying with welfare reform requirements.”  In particular, these scholars found that earnings for these women were erratic and incomes were highly variable.

Only 17% of the sample of 75 women interviewed worked steadily in the first year of the study. Others either cycled between two income sources, or combined sources (44% of the sample). Nearly one-third relied on welfare and did no work in the formal sector.  Income remained extremely low, even for the best-case scenarios. On average, incomes at the second interview were just above the federal poverty line (112% of poverty) for a single parent with two children (but many of these women had more than two children). Few women were able to purchase the items that, at the outset of welfare reform, they had believed would make their children’s lives measurably better, most notably moving to a better neighborhood with better schools.

In addition, these scholars found that the income increases experienced by these women did not primarily reflect higher earnings.  Instead, they came from the EITC, larger earned income disregards, transitional welfare benefits, child care, and, in a few cases, Supplemental Security Income (SSI).  “In fact, those who reported the most work and commanded the highest wages were also those with the greatest cash and in-kind network supports.”

Another group of scholars argues that the apparent gains from the 1996 reform were largely a consequence of the economic boom of the late 1990's, which generated many jobs, and that problems are starting to return in the current recession.  According to Sheldon Danziger, a noted poverty researcher at the University of Michigan,

Recessions always increase poverty, but poverty is likely to rise more in this recession than in the last one because the 1996 reform changed the nature of the social safety net….  Many women who have been unable to get or keep jobs during the boom are now at risk of reaching their time limits.   These women have more barriers to employment than those who have left the rolls; many suffer from depression, are in poor health, and have been victims of domestic violence.  They have spent many months on welfare during the last five years because they could not work steadily when employers were scrambling to fill vacancies.  They are unlikely to get steady jobs not that many firms are downsizing (Danziger, 2002).

Moreover, these scholars continue, not enough progress was made during the boom years.[5]  Over all, the official child poverty rate fell to 15.8 percent in 2001, but it is still well above the lows of the late 1960's and 1970's of around 14 percent, it increased between 2000 and 2001, and it probably moved up in 2002. Child poverty for blacks and Hispanics, also down, is still an unconscionable 30 percent and 28 percent, respectively.

The financial state of America's children becomes a stark national embarrassment when compared with that in other rich nations. Timothy Smeeding of Syracuse University, based on work done in collaboration with his Harvard colleague Lee Rainwater, has completed probably the most comprehensive study to date on the subject.  Because poverty rates are measured differently across nations, they are not typically comparable. To adjust for this, Professor Smeeding draws a poverty line at a specific percentage of a nation's median income (including government cash and near-cash transfers, like America's food stamps).  In Europe, official poverty lines are typically about 50 percent of median income, or even higher. In the United States, it is somewhat above 40 percent.

Based on a poverty line of 40 percent, Prof. Smeeding calculates that America still has a child poverty rate of 14.8 percent, the highest among the 19 rich members of the Organization for Economic Cooperation and Development for which he has data.  Only Italy comes close, at 14.6 percent.  The next closest is Canada at 9.6 percent, then Britain at 8.4 percent.  France's child poverty rate is only 2.9 percent, Taiwan's is 2 percent, and Sweden is at the bottom of the list, at 1.3 percent.  The American data is updated through 1997, but the fall in the poverty rate since then does not alter the national relationships because the differences are so large.

Prof. Smeeding also adjusts these income levels for what they can actually buy in each nation, or “purchasing power parity.” By doing this, he can estimate the actual standard of living of children in different countries.  It turns out that America's poorest children — say, those in the bottom 10 percent — have a lower standard of living than those in the bottom 10 percent in any of the other nations measured except Britain.  Moreover, the gap from rich to poor is far higher in the United States than anywhere else. In America, he finds, a child in the 90th percentile — one whose income is higher than 90 percent of all children — has an adjusted income five times that of one in the 10th percentile. In all other nations, it is an average of three times.

The United States also has the greatest gap between the child in the middle of the pack — the median — and the poor child in the 10th percentile. Guy Stevens, a senior economist at the Federal Reserve until his recent retirement, supplements income data with an analysis of a couple of dozen noneconomic indicators of childhood well-being. He finds that the state of America's poor children is every bit as bad as the low incomes suggest.  Infant mortality rates in America are only slightly better than in Cuba, Dr. Stevens notes, making it 33rd in the world. Eighteen percent of American women have minimal or no prenatal care, higher than in any other rich nation. Fourteen percent of children have no health insurance. Only 60 percent of 3- and 4-year-olds go to child care, well below the European rate, and many of those go to centers that are inadequate.  “Too bad children can't vote,” Dr. Stevens laments. Child poverty remains one of America's most stunning failures.

A related closely related development is the change in the number of child-only welfare cases.[6]  The number of welfare households in which the only recipients of benefits are children peaked nationally at 978,000 in 1996 after rapid growth during the late 1980's and early 1990's. This number dropped to 743,000 in 1998, as other welfare cases fell much more steeply, then began to climb again. Even in New York, where regular cases have continued to drop in tough economic times, child-only cases have been edging up every year since 1998, and are now nearly 37 percent of the caseload. The most recent national welfare figures, released to Congress in May, show child-only cases accounted for 35 percent of the caseload in 2000.  These developments have crept up almost unnoticed on policy makers focused on using welfare-to-work strategies to shrink caseloads and combat poverty. And recent studies show high rates of hunger and hardship among the young recipients of such aid.

Officials have barely begun to address the different challenges posed by these cases, including the amount of the grants themselves, as little as $68 a month in some states. But the share of child-only cases, up from 10 percent nationally in 1988, is becoming too big to ignore. One reason for the rise is that some of the same rules and practices that sharply reduced the regular welfare rolls are gradually expanding the pool of children eligible for child-only grants.  Some live with disabled parents who receive Supplemental Security Income for themselves, often after having been found unable to work for family aid. Some are children born in the United States to immigrant parents, either illegal or legal immigrants barred from welfare by strict new rules in the 1996 federal overhaul.  In a few states, though not yet in New York, some live with parents who have exhausted welfare time limits or violated the rules for aid.  But most of the children — those in 54 percent of cases nationwide — live apart from their parents, reared by relatives who have no legal responsibility to support them

A recent detailed study of New Jersey's child-only caseload found that 63 percent involved family members other than parents. Half were older than 60, and many were in ill health.  Indeed, a stark picture emerged from the New Jersey study, which the state paid Mathematica Policy Research to conduct. One in four grants went to a family headed by a disabled parent on Supplemental Security Income, and more than half of those families had trouble getting enough to eat last year, the study found. Three in 10 went hungry. That is double the rate of food shortage that is typical at their income level, the researchers said, suggesting that high medical costs or parental handicaps may have played a role.

Among immigrants, who accounted for one in 10 New Jersey cases, half the families had incomes below 50 percent of the federal poverty threshold, 84 percent had no health insurance, and many were in severely crowded households, though they had lived in the United States for nine years on average.  “Clearly the work-first approach is not relevant for them,” Robert G. Wood, a Mathematica researcher, said of child-only welfare recipients. “They need to be thought of as a separate population with their own problems and needs.”

Not everyone agrees. Though few participants in the welfare debate have drawn attention to child-only cases, some critics see them as the last bastion of unlimited aid, and describe many adults caring for such children as people exploiting the system to avoid work.  Those critics urge that all able-bodied caretakers of children receiving child-only benefits be required to work, and that illegal immigrants who seek aid for their children born in the United States be deported.

Nationally, more than 60 percent of all families headed by non-parent relatives have low incomes. Nearly half of these had trouble buying enough food, and about 40 percent had trouble meeting housing costs, according to a recent study by the Urban Institute, a nonpartisan research center.  Rob Geen, one of the study's authors, estimates that a quarter of all the nation's welfare cases involve caretakers who are relatives and collect welfare benefits themselves or on behalf of the children. But he stresses that in the larger category of children living without their parents, three out of four households headed by relatives receive no public cash aid, though virtually all the children are eligible. Some relatives do not want welfare because of the stigma, he says; some are unaware that help is available; and some have been wrongly denied help.

Several states are experimenting with programs that pay relatives more than welfare, but less than foster care, if children would otherwise be in state custody. But some economists suggest that approach has its pitfalls, too. If meager grants already encourage hard-pressed parents to leave their children with relatives — a contention that grandparent advocates vigorously dispute — higher subsidies for relatives could compound the problem.  Moreover, such programs do not address the needs of welfare children living with disabled or immigrant parents, who experience greater financial hardship, studies find. 

Issues to Consider in Reauthorization 

Any legislation to reauthorize welfare much address a wide range of issues.  The most important are time limits and work requirements (along with the associated sanctions and exemptions), links to Food Stamps and Medicaid, out-of-wedlock births and marriage, training and job advancement, child care, immigrants, and funding. 

Time Limits and Work Requirements 

Time limits are of course, the most dramatic feature of the 1996 legislation.  The House bill does not alter these limits.  Nevertheless, the range of views on this issue is quite large, as clearly expressed by an exchange between Mr. Haskins and Rebecca Blank, Dean of the Ford School of Public Policy at the University of Michigan (Haskins and Blank, 2001).[7]  Dean Blank believes that five years is not long enough. Many former welfare recipients have little work experience, and the jobs they find are unstable and low-wage. As a result, many cycle between work and welfare, or combine the two, eating up the five years of allowed benefits. Optimally, Dean Blank would eliminate the time limit, but if that is politically unfeasible, she would advocate for more state discretion in how time limits are applied. One option would be to stop the clock for families that are working 25 or more hours a week.  Mr. Haskins believes the time limit sends a clear message that work is required and that public assistance should not be a way of life. Moreover, states are allowed to use federal funds to support benefits for up to 20% of their caseload even after individuals have reached the five-year limit.

Some experts support preserving the time limits and strengthening the work requirements.  According to Prof. Meade (2001), for example, “To restore pressure on states to enforce work, the credit for caseload decline should be withdrawn. All states should be required to reach the original 50% standard, perhaps by 2005. Higher levels should be required for later years, until a clear majority of cases is working. “

Other experts, including Prof. Danziger (2001), support weakening the time limits and work requirements.

Recipients who have reached time limits or who cannot meet work requirements could be offered an opportunity to perform community service in return for cash assistance.  A more costly option, but one that would have a greater antipoverty impact, would be to provide them with low-wage public service jobs of last resort.  Welfare recipients who were willing to work could then combine wages with the EITC and support their families even when there was little employer demand for their skills.

A similar proposal is made by Mark Greenberg, a senior staff attorney at the Center for Law and Social Policy in Washington, D.C.  According to Mr. Greenberg (2001),

Congress should require that before a sanction is imposed, the state must inform the individual how he or she failed to meet program requirements and what needs to be done to correct the problem. Moreover, states should make reasonable efforts to contact sanctioned families, offer needed services, and allow assistance to be reinstated when families comply with program requirements.   


Links to Food Stamps and Medicaid

The 1996 legislation imposed work requirements on Food Stamps recipients but, in principle, it did not change the eligibility requirements for federal health insurance through Medicaid.  In practice, however, the tightening of the requirements for the receipt of welfare appears to have caused a dramatic reduction in participation for these two programs, even among those who are eligible.  According to Mr. Haskings and Dean Blank, Food Stamp caseloads have fallen nearly 40% since 1994, and health insurance coverage of children in former welfare families fell in 1996 and 1997, but has been growing somewhat since then.  Many recipients and apparently even many case workers do not realize that Medicaid eligibility is retained for many people who no longer receive welfare payments.

These problems will be difficult to solve with the current administrative structure.  Mr. Haskins would like to transform Food Stamps into a block grant program similar to TANF with a guarantee of eligibility for certain families.  Dean Blank would prefer to clarify the eligibility rules and simplify the administrative procedures.  To solve the Medicaid problem, Mr. Stevens would like to see a serious effort to provide universal health insurance for all children.


Out-of-Wedlock Births and Marriage

Although programs to reduce out-of-wedlock births and to promote marriage are politically popular, Mr. Haskins and Dean Blank point out that there is very little evidence that these programs work.  Under these circumstances, Mr. Haskins’ preferred approach is to fund and evaluate demonstration programs that might prove to be effective.  He also approves of programs to abstinence education.  Dean Blank would rather introduce programs for improving schools and making work pay.  She bases this position on recent research suggesting that giving teenagers a sense of promising future alternatives is a key to reducing out-of-wedlock births.  In addition, Dean Blank opposes a federal mandate that states do more to reduce out-of-wedlock births and encourage abstinence. 

Education and Training 

The issue of education and training is one of the most contentious issues in the table.  A strong bi-partisan consensus supported a heavy emphasis on education and training in the 1988 Family Support Act, but this consensus has sense evaporated.  As noted earlier, conservatives such as Mr. Haskins, approve of a shift in focus away from education and training toward work.  This position is supported by extensive recent research showing that most education and training programs do not have significant job-market impacts.  In contrast, Dean Blank recommends that education or job training programs count as a valid fulfillment of the work requirement.  This position is supported by the obvious point that many, if not most, welfare recipients lack the education and training needed to obtain and hold a job with decent wages.

A related issue is what to do about job stability and job advancement.  As noted earlier, many former welfare recipients have a difficult time holding a job, so policies that promote job stability and job advancement would be desirable.  Because not much is know about such programs, one approach would be to encourage experimentation (and evaluation) of demonstration programs designed to achieve these objectives.  A complementary policy would be a set of programs to help hard-to-employ parents overcome obstacles such as drug abuse and domestic violence.  This type of policy has been emphasized by a few Democratic senators, by Prof. Danziger, and by Scott et al. (2002), who write “The variety of different needs among families points to the importance of flexible and comprehensive programs that are responsive to the changing needs and circumstances of working poor women and children. Those with low human capital, fewer network resources, and children with health or behavior problems may need considerably greater intervention.”


Child Care

Virtually all observers agree that child care is a critical element of welfare reform, but there is enormous disagreement about the amount of resources needed.  According to Mr. Haskins and Dean Blank, conservatives see few problems with either the availability or cost of care. Liberals, in contrast, see affordability and convenience as key issues to address during reauthorization.  Mr. Haskins and Dean Blank also agree that the quality of child care is a serious problem.  Dean Blank sees a need for improved access to subsidies and services to help women find stable, quality care. She favors enforcing standards by requiring states to demonstrate that their subsidies are adequate to purchase care that meets certain standards. Mr. Haskins believes current funding levels are adequate, especially because most states have surplus TANF funds that can be used for child care, and is wary of further legislation, especially concerning standards.



In the post 9/11 era, policies concerning immigrants are bound to be controversial.  Indeed, many people now strongly support the dramatic restrictions on benefits to immigrants that were contained in the 1996 legislation.  However, others argue that restoring some types of benefits, such as Medicaid coverage to children and pregnant women who entered the country after August 1996, would clearly be in the national interest. According to Mr. Greenberg, for example, “An estimated 27% of poor children are immigrants or the children of immigrants, and the nation cannot implement an effective anti-poverty strategy that excludes this group.”

Dean Blank supports this restoration of benefits and believes Food Stamps and Medicaid should be made available to all legal immigrants.  Further, she believes that SSI should be available to immigrants who become disabled after entering the country.  She finds it unacceptable that some residents should be hungry or unable to receive basic medical care.[8]  Mr. Haskins, in contrast, believes that immigrants should come to this country in search of opportunity, not welfare. He believes the current law provides for reasonable exceptions that cover emergencies and other instances in which assistance should not be denied. Moreover, assisting needy non-citizens should be the responsibility of their sponsors, not taxpayers.


Finally, commentators disagree on the amount of federal funding that is needed for welfare reform.  Some people point to the huge drop in caseloads and argue that the federal contribution could be cut back significantly.  However, most people argue that the drop in caseloads has had the beneficial effect of allowing states to implement a wide range of new programs addressing issues in, for example, education, training, and child care.  As Mr. Haskins and Dean Blank (2001) put it, “Many of those entering the workforce—the majority in low-wage jobs—are likely to need continued supports if they are to rise out of poverty and advance in jobs.” 

Indeed, many commentators believe that the block grant amount should be increased.  As Prof. Edelman (2003) put it, “The amount of money set aside for welfare in 1996 hasn’t changed, and now, with inflation, is worth 11 percent less than it was then.  The welfare roles have been cut in half, to be sure, but the point should be to get people out of poverty, not just off of welfare.  This costs money.”  Mr. Haskins argues that the amount should be linked to inflation, whereas Dean Blank prefers a link to economic indictors, such as poverty.  A key question with any such link is whether it would vary by state, that is, whether federal funding should recognize shifts in need between states over time.      

The Decision

The Work, Opportunity, and Responsibility for Kids (WORK) Act was passed by the Senate Finance Committee on June 26, 2002, but was never considered by the full Senate.  The WORK Act:[9]

Now the Senate must act again.  The starting point of the Senate debate will be, of course, the WORK Act.  To meet it responsibility for designing reauthorization legislation, however, the Senate Finance Committee has decided to consider alternative approaches and to solicit recommendations from a variety of sources. 

As the Director of the Income and Benefits Policy Center at the Urban Institute, a highly respected, nonpartisan research organization, you have been asked to present your recommendations at a hearing on reauthorization to be held by the Senate Finance Committee.  You have been asked to state and defend your views on the most important features of reauthorization legislation.  If you wish, you may also identify provisions in the House bill or the WORK Act that should not be included in a reauthorization package and briefly explain why they are not appropriate.


Aid to Families with Dependent Children (AFDC) was the main federal welfare program from the 1930s to 1996.  It provided aid for needy children and their caretakers.  In general, families with children were eligible to receive AFDC if one parent was absent or disabled or both parents were unemployed, and if the family met an income and resource test.  In New York in 1996, total need for a family of four was about $600; the actual grant equaled this amount minus any income the family earned. 

The Family Support Act of 1988 (FSA) placed some work requirements on AFDC recipients through the JOBS Program, but the requirements were flexible and could be satisfied through education and training as well as through employment.  Moreover, FSA mandated that states provide child care to JOBS participants. 

In 1996, AFDC was repealed by the Personal Responsibility and Work Opportunity Reconciliation Act of 1996.  This act replaced AFDC with the Temporary Assistance for Needy Families Program (TANF).  The major provisions of TANF are: 

1.  Federal Funding and Regulation 

The 1996 law eliminated the matching formula that funded AFDC.  Instead of a matching grant, each state now receives a block grant based on the money it received under the pre-1996 formula.  Thus, the amount a state spends will no longer influence the amount of money it receives from the federal government.  The law has no provision for updating a state’s block grant as the eligible population in the state changes. 

The states have vastly increased discretion under the new legislation.  The federal government must determine whether a state’s plan is “complete,” but the law does not make it clear what happens if the plan is found to be incomplete or if the federal government disagrees with a state on a particular issue.

2.  Lack of Entitlement 

Under AFDC, every family that met the eligibility criteria was entitled to receive benefits.  This is no longer true.  If a state has used up its block grant from the federal government and does not want to raise any more money, it is not obligated to provide benefits to families that meet its TANF eligibility criteria.  Some analysts anticipate that, particularly during recessions, many families may lose their benefits for the last few months of a state’s fiscal year.

 3.  Work Requirements 

Unless a state specifically opts out of this provision, all recipients of cash assistance are required to participate in community service after receiving benefits for six months.  This provision applies to all parents or caretakers receiving assistance.  Recipients of cash assistance face a work requirement after receiving assistance for two years.  The law required 25 percent of single-parent recipients to work in 1997 and raised this limit to 50 percent in 2002.  States that fail to comply are penalized. 

The law required 75 percent of two-parent recipients to work in 1997 and raised this limit to 90 percent by 2002.  Again, states are penalized for failure to comply.  Participation in education and training programs cannot generally be counted as satisfying these work requirements.

4.  Child Care 

The 1996 law does provide a limited amount of money for child care to help parents, particularly single parents, when they go to work.  However, a state is not required to provide child care for families who receive benefits.  In fact, parents of children age six or above are not guaranteed child care and are not entitled to an exemption from the law’s work requirements because they lack child care.  This type of exemption will be given only for parents of children under six who can demonstrate an inability to obtain child care. 

5.  Five-Year Limitation 

TANF imposes a five-year (60 month) limit on the receipt of welfare benefits.  Specifically, states are prohibited from using federal funds to provide benefits to any family that has received benefits for 60 months.  This limit, which applies to any family in which an adult resides, counts any month in which the family receives federal assistance.  This assistance is not necessarily limited to cash assistance, but could also include the receipt of subsidized housing, counseling, child care, and so on. 

The five-year limit does not include any month in which the adult(s) in the family does not receive assistance.  However, once the five-year limit has been reached for an adult in the family, the entire family is ineligible for assistance, not just the adult.  To be specific, TANF prohibits the use of funds “to provide assistance to a family that includes an adult who has receive assistance under any State program funded under this part attributable to funds provided by the federal government for a period of 60 months.” 

Because of these provisions, some families may have to move out of subsidized apartments to avoid adding time toward their five-year limit and some children who are living with caretakers, who are, along with parents, covered under the law, may have to move into foster care to prevent a family from losing its benefits.  However, up to 20 percent of each state’s annual caseload can be exempted from the five-year limit.  Exemptions can be granted on the basis of hardship or if the family includes an individual who has been battered or subjected to extreme cruelty. 

6.  Coverage of Immigrants 

As originally passed, TANF, excluded immigrants arriving in the U.S. after 8/22/96 from eligibility for cash assistance, food stamps, and other means tested programs, including SSI, for a period of five years.  After five years, their sponsor’s income was “deemed” to be available to them, until they become citizens or had worked for 10 years (as measured by Social Security Administration criteria).  In other words, their sponsor’s income was considered in determining whether they are eligible for benefits.  (No one can immigrate into the U.S. without a sponsor.)  No exemption to these provisions was provided for immigrants who are 75 years of age or older or who are too disabled to go through a naturalization process. 

The 1996 law also ended the benefits of noncitizens who were already receiving welfare, but a bipartisan coalition repealed this harsh provision in 1997.  In addition, eligibility was restored for many needy immigrants. As one report puts it, 

Many children, elderly, and disabled immigrants have had their eligibility for food stamps and SSI restored.  But most working-age adults remain ineligible for food stamps and most immigrants entering the United States after the law’s passage are ineligible for the major federal assistance programs for at least five years.  (Zimmerman and Tumlin, 1999)           

The long term prospects for immigrants also are not clear.  The 1996 legislation turned over to the states decisions about immigrants’ eligibility for TANF and Medicaid.  Most states have retained benefits so far, but may back away from this position when faced with the budget challenges of 2003. 

7.  Food Stamps 

Food Stamps remains an entitlement program, but the amount of Food Stamps benefits is cut in various ways.  For example, the shelter deduction cap will be frozen at $300 in 1997.  No matter what happens to housing costs in future years, shelter expenses above this amount cannot be deducted from income in calculating Food Stamps benefits. 

All “able-bodied” recipients between the ages of 18 and 50 will have to work 20 hours per week or else be limited to 3 months of Food Stamps in any 36 month period.  (CBO estimates that as many as 1 million unemployed individuals who cannot find work will be denied Food Stamps in an average month because of this provision.)

8.  Medicaid 

Medicaid remains an entitlement program, and eligibility for Medicaid will still be determined according to the AFDC that applied as of July 16, 1996, and there is no five-year limit on Medicaid eligibility.  People eligible for Medicaid through routes other than AFDC, such as being medically needy, will still be eligible, for the most part, according to the old rules.  However, people can lose Medicaid eligibility for failure to meet work requirements.  Termination of Medicaid to an adult for violation of the work rules does not extend to children in the family.


Bernstein, Nina.  2002.  “Child-Only Cases Take up a Bigger Share of Welfare.”  The New York Times, August 14. 

Danziger, Sheldon.  2002.  “Recession Threatens Welfare Reform’s Success.”  Poverty Research News, vol 6, no. 2, (March/April). 

Edelman, Peter.  2003.  “Proposed Welfare Reform Measures Will Only Increase the Hardship People Face.”  Op-Ed Piece, National Public Ratio, February 14.

Goldstein, Amy and Juliet Eilperin.  2003.  “House Passes GOP Welfare Plan:  Tougher Work Rules, Funds to Promote Marriage Included.” The Washington Post, February 14, p. A4.

Greenberg, Mark.  2001.  “From Caseload Reduction to Poverty Reduction.”  Poverty Research News, Vol. 5, No. 6 (November/December), pp. 6-7.

Haskins, Ron.  2003.  “Interview on Morning Edition.”  National Public Radio, February 14. 

Haskins, Ron, and Rebecca M. Blank.  2001.  “Welfare Reform Reauthorization.”  Poverty Research News, Vol. 5, No. 6 (November/December), pp. 3-5. 

Madrick, Jeff.  2002.  “A Rise in Child Poverty Rates is at Risk in the U.S.”  The New York Times, June 13. 

Mead, Lawrence M.  2001.  “Rebuilding Welfare into a Work-Based System.”  Poverty Research News, Vol. 5, No. 6 (November/December), pp. 8-9. 

Moffitt, Robert.  2001.  “A Return to Categorical Welfare.”  Poverty Research News, Vol. 5, No. 6 (November/December), pp. 12-13.

Pear, Robert.  2003.  “A Welfare-to-Work Study Finds No Harm to Children”  The New York Times, March 7. 

Scott, Ellen, Kathryn Edin, Andrew W. London, and Rebecca Joyce Kissane.  2002.  “Welfare Recipients Struggle to Balance Work and Family.” Poverty Research News, vol 6, No. 4 (July/August), pp. 12-15 

Zimmerman, Wendy, and Tumlin, Karen C.  1999.  “Patchwork Policies:  State Assistance for Immigrants Under Welfare Reform.”  Washington, D.C.:  The Urban Institute, May. 



[1] This case was written by Professor John Yinger solely for the purposes of class discussion.  Many of the quotations and some of the sentences come from the newspaper articles in the bibliography.

[2] As Professor Lawrence Meade of New York University puts it (2001), this provision has “knocked the bottom out of the new work standards. In 1999, 35% of cases were supposed to be working, but the caseload credit for declines cut the standards for most states to trivial levels—in 23 cases to zero. Virtually all states met these lowered standards, but 23 failed to reach the original 35%.”

[3] Scholars also disagree about the quality of the available evidence.  According to one leading poverty scholar, Robert Moffitt (2001) of Johns Hopkins University, “making recommendations for reauthorization is difficult … because the evidential base to support specific policy recommendations is weak. Despite the great volume of data analysis conducted on welfare reform since 1996, a careful review of what has been done reveals, unfortunately, that there is very little strong research evidence on many of the key reforms, or modifications of reforms, that Congress is likely to consider…. Indeed, it is probably not going too far to say that there is almost a complete research vacuum on these issues.”

[4] Rep. Emanuel also said, however, that that more aid for child care is necessary.

[5] The paragraphs on child poverty are taken from Madrick (2002) with only minor editing and some updating in the first paragraph from the Census website (

[6]The paragraphs on child-only welfare households are taken from Bernstein (2002) with only minor editing.

[7] The views of Mr. Haskins and Dean Blank are taken directly from this publication with editing to fit the flow of this case (but not for content).

[8] A more unusual argument in favor of benefits for immigrants is provided by Professor Mead (2001), who says that these benefits would provide a way to bring child-only cases “under the work test.”

[9] This list of provisions comes from the National Association of Social Workers (


Trustee Professor of Public Administration and Economics