Welfare Reform: Meaning and Effects

Lawrence Mead

Below, I briefly summarize what “welfare” and “welfare reform” mean as political issues in the United States and what the effects of reform have been in recent years.  I concentrate mainly on the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, the last and most radical reform to date.  I attempt to avoid saying anything controversial, that is anything that well-informed experts would not agree on, whatever their politics.

Welfare

Controversy about “welfare” in American politics has revolved almost entirely around the family aid program formerly called Aid to Families with Dependent Children.  AFDC, first established by the Social Security Act of 1935, was renamed Temporary Assistance for Needy Families (TANF) by PRWORA and substantially recast, as explained below.  TANF, like AFDC, uses both federal and state funding to provide means-tested aid to needy, mostly single-parent families.

There are other welfare programs in the sense of means-tested aid, such as Food Stamps (coupons for food), Supplemental Security Income (aid for the aged and disabled), Medicaid (health care for welfare recipients and some other poor), and the Earned Income Tax Credit (EITC, wage subsidies for low-income workers), but they have aroused much less controversy.

Welfare reform

In the broadest sense, “welfare reform” has meant proposals to change AFDC/TANF and the controversy about them.  These have been the principal criticisms of the program, and the corresponding recommendations, from both left and right:

·   Low benefits or coverage: Benefits should be raised or coverage expanded.  Welfare now pays an income well below the poverty line.  Higher benefits and coverage of two-parent families would raise incomes.  Secondarily, work incentives would raise work levels.  This was the goal favored by liberal planners in the 1960s and 1970s.^[1]

·   Weak support for work: Provide benefits to “make work pay.” Welfare deters work by deducting earnings from the grant.  Instead, create work incentives by reducing grants by less than earnings.  Add work-connected benefits such as health and child care.  Provide education and training so that recipients can get jobs good enough to escape poverty. The main position of liberals today.^[2]

·   Excessive dependency: Welfare should be reduced or eliminated to cut reliance on government and reduce social problems among the poor.  Anti-government conservatives trace evils such as unwed pregnancy and joblessness to the ready availability of government aid.^[3]

·   Unwed pregnancy:  Welfare should be reformed to reduce illegitimacy, but without denying aid to the needy.  Most welfare mothers are unwed, and these also are the mothers who remain dependent longest.  Conservatives argue that welfare could deter such misbehavior better if it were devolved to localities or the private sector.^[4]

·   Nonwork: Welfare should be reformed to raise work levels among the poor, also without eliminating aid.  The initial reason for most poverty among the working-aged, including welfare recipients, is lack of steady employment.  To raise work levels requires programs to promote employment among poor adults.  Probably the more important conservative reform demand.^[5]

·   Nonpayment of child support: Improve enforcement.  Many single mothers go on welfare for lack of support from absent parents.  To improve child support enforcement and assure payments to families might provide many with an alternative to aid.^[6] 

·   Fraud and abuse: Clean up the system.  Ineligibility and incorrect grant payments were rife in the 1960s and 1970s, when the welfare rolls grew rapidly.  Driven by federal fiscal sanctions, states reduced errors, by the 1990s, to around 6 percent of payments.^[7]

·   Cost: welfare coverage or benefits should be cut to reduce spending and help balance the budget.  This is another theme of anti-government criticism of welfare.

·  Excessive federal control: Responsibility for welfare should be devolved to states or localities. A further theme of anti-government conservatism.  Some of these proposals would maintain federal funding, some eliminate it entirely.^[8]

Serious controversy about welfare first arose in federal politics in the early 1960s.  Since then, all the above meanings of reform have had currency at some point.  But over time, conservative meanings—those stressing work and child support requirements, limits on aid, and devolution to lower levels of government—have tended to win out.  This was partly because American politics as a whole became more conservative after the 1960s.  And of the conservative themes, work enforcement has become the most prominent.  That was because it is popular, but also because it is more feasible than some other conservative goals.  Government has learned how to enforce work, but not how to enforce child support or prevent unwed pregnancy.

Efforts to enforce work in AFDC go back to 1967 but became serious only in the 1980s.  Starting in the Reagan Administration, some states gained special permission to run welfare work programs that were more demanding than usually permitted by federal rules.  The favorable evaluation of some of these programs by the Manpower Demonstration Research Corporation (MDRC) lent further impetus to the work strategy.  In the Family Support Act (FSA) of 1988, Congress toughened and expanded these programs.  The majority of welfare mothers were still exempted, however, and those that did participate more often went into education or training than into available, usually low-paid jobs.

Welfare politics

Politicians tend to polarize over welfare, with liberals favoring higher benefits or coverage while conservatives favor restrictions or tougher work tests.^[9]  The public, in contrast, divides more inwardly than outwardly.  The voters are largely agreed about welfare, but what they want is complex.  They strongly oppose the unreformed aid system, largely because most of the adult recipients appear to them to be undeserving.  Yet they accept the principle of aid.  They want government to be generous to children and other poor people in need.  But they also insist that poor parents and other adults work in return for support.^[10]  Popular support is much stronger for enforcing work than for attempts to enforce marriage or reduce unwed pregnancy.

The core issue in welfare politics is entitlement.  One meaning of that term is budgetary: In an entitlement program, government is legally obligated to pay benefits to all persons eligible for a program, regardless of available funds; in a non-entitlement program (for example, training programs or public housing), funds are limited and some eligible persons may be denied benefits.  The more important meaning of entitlement is behavioral: Entitlement means that people qualify for aid on the basis of impersonal economic criteria, such as income, regardless of their behavior or lifestyle.  They need not work or otherwise appear as “deserving.”  Conservative politicians and, it seems, most voters want to end entitlement and condition aid on good behavior, especially work effort, at least for the employable.  But liberal politicians and community groups, and some voters, defend entitlement as realistic given the problems that poor families face.  They support employment as a goal, but they hesitate to enforce it as a condition of aid.

In general, controversy over welfare has been to the advantage of conservatives and Republicans.  Growing dependency, like crime, raises popular concerns about social order, shoving off the agenda the concerns of liberals and Democrats about economic fairness and equality.  Con­versely, when social problems abate, economic issues return to prominence and the political agenda becomes more liberal.^[11]  Currently, both welfare and crime are falling, and this has favored the left.

PRWORA

The Republicans’ takeover of Congress in 1994 gave conservatives more power to reshape welfare than they had had since the inception of the program.  A sense of urgency also arose from a sudden rise of the welfare rolls by about a third between 1989 and 1994.

PRWORA thus goes beyond the efforts to enforce work and child support and save money that marked earlier conservative reform proposals.  The act also attempts to promote the family and devolve control of aid to states and localities in radical new ways.  These aspects were obnoxious to most Democrats and lacked a clear popular mandate.  However, President Clinton had a history of criticizing welfare, which had been politically damaging to Democrats.  He also feared that welfare might become an issue in his reelection campaign.  So he felt bound to sign the bill and did so.^[12]

These were PRWORA’s main provisions.  Their main brunt was to end entitlement in both a budgetary and behavioral sense:

·   Aid was  unentitled: Under AFDC, the federal government shared in the cost of local aid programs however many eligible persons they served.  Under TANF, instead, federal funding is limited to a fixed block grant to each state.  That grant, however, was set at the levels of 1994-5, which turned out to be generous due to the unexpected fall in the caseload.  Additional money was added for child care, and TANF spending was supposed to rise above earlier levels.

·   Aid was time-limited: Only 20 percent of families are allowed to remain on TANF for longer than five years after the signing of PRWORA in August 1996.  This includes repeat spells.  However, states may support families beyond five years using their own resources.  Some liberal states, including New York, plan to do so.

·   Work requirements were ostensibly toughened: Cases are supposed to work within two years of going on the rolls even to continue to get aid.  States are required to raise the share of their cases that are working in steps, until half are in work activities by 2002, on pain of cuts in federal funding.  However, states were also allowed to count against these targets any percent by which their caseloads fell after 1995.  In most states, the fall was enough virtually to negate the new work standards.  Parallel but less controversial steps were taken to strengthen child support enforcement.

·   Control of welfare was further devolved: While states controlled benefit levels under AFDC, under TANF they also got greater control over eligibility and other policy details.  These include whether to cover two-parent families (largely excluded under AFDC) and whether to institute work incentives (rules that allow recipients who take jobs to keep part of their benefit, to give them more reason to work).

·   Marriage was promoted: States are allowed, although not required, to deny coverage to unwed mothers under 18 and to children born on the rolls.  If covered, teen parents must live with caretakers and go to school.  States also receive bonus funding if they reduce the incidence of unwed pregnancy or do well relative to other states in promoting job retention and earnings gains among recipients going to work.  They are not, however, penalized for failure in these respects.

·  Other aid was cut: Legal aliens were largely barred from TANF, SSI, and Food Stamps, and these programs were cut in other ways.  PRWORA’s budgetary savings were entirely due to these cuts, not to cuts in AFDC/TANF.  Congress later restored most benefits for aliens who received them prior to the signing of PRWORA, but other aliens remain largely excluded.  SSI, Food Stamps, and Medicaid were not otherwise reformed and remain entitlements.

Most states have implemented the act so as to combine continued aid with rising work demands, as the public apparently wants.  Most are instituting tougher work requirements.  Few have cut benefits, as liberals feared, or paid much attention to the family provisions.  This is because these steps are much more controversial than enforcing work, and because programs that clearly prevent unwed pregnancy are not yet available.^[13]

The liberal side of reform

Recent welfare policy also has a liberal side that is less often noticed.  In the last decade, the following steps have raised support for poor families, especially if they are working:

·     Expanded wage subsidies: The Earned Income Tax Credit was sharply increased in 1990 and 1993.  The EITC now subsidizes the earnings of low-income parents by as much as 40 percent.  Some states have added wage subsidies of their own.  In 2001, Congress added a tax credit for children that is aimed at full-time workers and is partially refundable even to families without tax liabilities.

·     Strengthened work incentives: The majority of states have liberalized work incentives in TANF.  These rules make it more advantageous for recipients to go to work.  Families in the typical state can now escape poverty even working at the minimum wage, provided they work at least half-time.  On the other hand, the subsequent loss of these benefits as earnings rise further makes it harder to move up to higher incomes.^[14]

·     Higher minimum wage: In 1996-7, the federal minimum wage was raised from $4.25 to $5.15 an hour, thus enhancing the earnings among the low-paid, although few workers at the minimum wage are heads of household.

·     Expanded child care: Federal financing for child care increased sharply.  In 2000, Washington paid about $15 billion for child care under Head Start and several other programs, both in and outside of welfare.^[15]

·     Expanded health coverage: Although plans for a universal health system have failed, Congress in the 1980s and 1990s expanded coverage for poor children and families under Medicaid and the Child Health Insurance Program (CHIP).  Today, most poor mothers and children can get at least some coverage whether on or off welfare.

The enactment of these measures reflects the popularity of aiding poor families—provided the adults work.  The new benefits already represent a renewed redistribution of advantages to the working poor, part of the liberalizing political shift mentioned above.

Effects of reform

“Effects” here means the consequences, not just of PRWORA, but of the overall process of welfare reform going back to the 1980s.  I describe below the key changes surrounding welfare and poverty since 1994, when the welfare rolls peaked.  How far these effects are really due to reform, rather than other changes, is addressed further below:

·   Cash welfare rolls have fallen sharply:  From a peak of over 5 million cases and over 14 million recipients in 1994, AFDC/TANF fell to 2.2 million cases or 5.8 million recipients by June 2000.  That decline is 56 percent for cases, 59 percent for recipients.  This decline is the greatest in the history of the program and considerably more than experts expected.  It more than reversed the welfare growth of1989-94.^[16]

·   Noncash rolls have fallen less sharply: The Food Stamps rolls have come down by around a quarter since 1993.  The Medicaid rolls have not fallen, but neither have all children and families claimed the expanded coverage that has been offered.^[17]  Many TANF leavers do not claim these benefits, even though they remain eligible for them, either because the bureaucracy fails to offer them or because people prefer to leave all welfare programs when they exit TANF.

·   Work levels on welfare have sharply risen: In 1998, 23 percent of AFDC/TANF adults were working in unsubsidized jobs, compared to about 8 percent in 1994.  In those same years, the proportion in work activities within welfare work programs rose from 19 to 35 percent.^[18]

·   Most people who left welfare are working: According to state studies of the leavers, most left welfare because of going to work or increased earnings.  From 55 to 64 percent of adult former recipients are employed when surveyed, and from 63 to 91 percent have worked since leaving welfare.^[19]  These work levels are much above what most experts considered feasible.

·   Nonworking leavers usually have other support: They often have benefits from other programs, such as disability.  Some surmise that they also have support from spouses or other family members who are working, but this is unclear.^[20] 

·   Earnings remain low: Among working leavers, average wages run from $5.50 to $8.80 an hour.  Quarterly earnings average from $1,999 to $3,868.  This is below what one would expect from the wage levels, because few leavers work normal hours.^[21]

·   Whether reform has raised overall incomes is unclear:  In the short run, leavers’ earnings usually exceed prior welfare benefits, but families also tend to have less noncash coverage than before.  There is some evidence that incomes among the poorest single mothers have fallen, but at the same time their consumption is rising.^[22]

·   Earnings and incomes rise slowly over time: Earnings tend to increase gradually the longer a family has been off welfare. This could bring many above poverty eventually.  It appears that wel­fare mothers glean higher wages with experience, but only if they work steadily, as many do not.^[23]

·   Profamily effects appear likely:  The share of welfare mothers living with a partner rose from 7 to 14 percent between 1997 and 1999.^[24]  A welfare reform evaluation (see below) also found positive effects on marriage and less spousal abuse among families exposed to the reform.^[25]

·   Hardship appears limited: As much as 59 percent of leavers report problems paying bills or affording food and housing, but many had similar problems before leaving welfare.^[26]  Most leavers appear to be struggling, but few suffer acute hardship such as homelessness or a need to give children over to foster care.

·  Many social problems remain: In cities with extreme caseload falls, such as Milwaukee, journalists observe more people working and paying taxes.^[27]  Much less change is apparent in the more private problems common among the poor, such as family conflicts or sexual or substance abuse.^[28]

Alongside these changes, several recent evaluations of welfare work programs have found effects on children, mostly small.  These programs mostly date from the early 1990s, prior to PRWORA, but they are enough like the programs states are now implementing that most experts take this evidence as applying to the current reform^[29]:

·   Effects on young children appear positive: When their parents are in welfare work programs, young children tend to show cognitive gains and do better in school.  These effects appear stronger for boys than girls.  The reason may be that a working mother is a better parent and role model, and/or that children receive more stimulation in child care than they would at home.

·   Effects on adolescents appear more negative: These children tend to have more behavioral problems, such as drinking or school misbehavior, perhaps because parents are now working and less able to supervise them.

·  Health coverage falls: Fewer children have health benefits, due to families leaving Medicaid along with cash welfare.  Again, how this happens is unclear since health eligibility has expanded.

The effects appear most positive with programs that combine work requirements with income supplements such as work incentives.^[30]

These welfare changes, in turn, are related in unclear ways to the following positive changes occurring in the society at large:

·   Work levels have risen sharply: More poor adults are working than before welfare reform, especially single mothers.  In 1999, 64 percent of poor female heads of family with children were employed, compared to only 44 percent in 1993.^[31]

·   Poverty rates are falling more slowly: The overall poverty rate was 12 percent in 1999, down from 15 percent in 1993.  Poverty rates specifically for children have also fallen.^[32]

·   Unwed childbearing is declining: The share of births out of wedlock is falling, both overall and for teenagers, and the teen birthrate is falling.  These trends, however, are gradual and began in the early 1990s, prior to PRWORA.^[33]

·  Child abuse and neglect is falling: Again, trends are gradual and predate PRWORA.^[34]

Doubts

Several uncertainties surround the role that welfare reform has played in these developments.

The role of PRWORA: As mentioned above, the 1996 act’s bark is worse than its bite.  Its severe work standards were largely obviated by the caseload fall credit, and its family goals are not mandatory for states, which have mostly ignored them.  It is thus unclear how much of “welfare reform” is really due to PRWORA.  There is a vogue for work requirements at the local level, but this seems to stem mostly from earlier federal requirements or state and local decisions. 

The role of welfare reform: It is also unclear how important welfare reform is in producing the above effects, as against other changes that may be responsible.  The surprising fall in the welfare rolls and the rise in work levels seems due to all of the following:

·   Work requirements: That is, the demands that adult recipients work in return for aid that stem, in varying degrees from PRWORA, earlier federal policy, and state and local decisions.

·   A superb economy: The later 1990s saw the best economic conditions for poor adults in thirty years—not only rapid economic growth and job creation but rising real wages for the low skilled.

·   Better work-connected benefits: As mentioned, the liberal side of welfare reform has raised the EITC, the minimum wage, and other benefits for the working poor.

·   Better child support enforcement: This problem is far from solved, but there is progress.  In 1997, 53 percent of poor single mothers had child support judgments, compared to 38 percent in 1978.  The share receiving support from absent parents rose from 18 to 25 percent in the same years.^[35]

·  A political climate against dependency: Aside from policy changes, public opinion is less accepting of dependency by the employable than it once was, and this also motivates needy families to leave the rolls or avoid them entirely.

Most experts agree that all these forces help to explain the dramatic welfare and work changes, but they disagree about their relative importance.  Some economists argue that the changes are mostly due to the economy and new benefits, not work requirements.^[36]  Journalistic observations of reform, however, suggest that work requirements are most important, followed by the economy.  In a study of welfare decline in Wisconsin, officials told me that work requirements and expectations were the leading cause, followed by the economy and improving child support; they never mentioned new benefits.^[37]  One reasonable estimate is that 35-40 percent of the work rise for low-income single mothers is due to work enforcement, 25-30 percent to the economy, and 20-30 percent to expanded benefits.^[38]

What if recession comes? Experts agree that the principal question hanging over the reform is what would happen if economic conditions worsened.  Would the leavers then lose jobs and return to welfare, causing the rolls to rise?  The economy has recently weakened, and at the same time the fall in the rolls has leveled off.  Is there a connection, or is the stabilization simply due to the fact that the remaining recipients are those least able to work?