The New Anti-Welfare States
While states experiment with real change, Clinton threatens to end welfare reform as we know it.
If you want to see what a real "two years and out" welfare-reform plan might look like, look to Madison, not Washington.
Madison, Wisconsin, not Washington, D.C., that is. That conclusion only became unmistakably clear once President Clinton finally unveiled his long-awaited, much-anticipated welfare-reform plan in June. The Clinton plan promises that, by the end of the century, about 7 percent of families receiving federal cash assistance will have to leave the rolls after two years unless they participate in some form of jobs program.
In the state of Wisconsin, Republican Gov. Tommy Thompson's reform agenda is moving along at a less-glacial pace: A pilot program approved for his state last year will put welfare families to work after 30 days, and cut off cash assistance to families in two years whether they are working or not. Other non-cash assistance such as medical coverage will end after another year. Thompson's plan, in other words, actually makes good on candidate Clinton's pledge to "end welfare as we know it."
Not coincidentally, the Wisconsin plan puts into action the "new consensus" among welfare-policy analysts: Long-time assistance destroys initiative and family formation; regular work is essential to building the skills and self-respect needed to make the leap to self-sufficiency; education and training are far less important than actual work experience; and spending lots of money to "reform" welfare is counterproductive.
Unfortunately for serious welfare reformers in Wisconsin, Massachusetts, Oregon, and other states, the Clinton plan promises to be the end of welfare reform as they know it. Not only does Clinton fail to deliver the goods, but he will keep the reformers from doing so as well. Analysts say a federal bill introduced by or acceptable to the administration will trump plans that call for stronger work requirements. Combined with heavy-handed federal oversight of state experimentation and recent anti-reform court decisions, Clinton-style welfare reform will be a public-policy catastrophe.
That's probably not what many Clinton voters expected after their candidate's campaign promises during 1992. His calls for welfare reform, which New York Sen. Daniel Patrick Moynihan (D) has dismissed as "boob bait for the Bubbas," helped Clinton separate himself from his party's traditional political establishment and contributed mightily to his reputation as a "New Democrat."
His plan, though, was kept deliberately vague. After the election, an internal war in the administration—between paleoliberals aghast at the prospect of cutting off benefits to anyone and political advisers arguing for some sort of bill they could sell to voters as tough welfare reform—took 17 months and some 200 grueling meetings to resolve. It wasn't until June that the details of the bill finally made the papers.
But while the administration was conducting an extended excursion into welfare wonkery, individual states were hammering out and attempting to implement actual reforms. In Wisconsin, Thompson has been assailing the welfare state for years. Since his election on a welfare-reform platform in 1987, the state has reduced its Aid to Families with Dependent Children caseload by almost a fifth, while all but two other states have seen their caseloads rise. But even in Wisconsin, welfare remains in need of a complete overhaul, not just a tune-up. It guzzles hundreds of millions of dollars in state and federal money. And the existing programs do little to help thousands of residents get off of welfare and become self-sufficient members of society. So last year Thompson announced his latest broadside, called Work Not Welfare, which, under a waiver granted by the federal government, initially affects two Wisconsin counties.
Perhaps the best way to get a handle on how Clinton's and Thompson's plans differ is to compare their effect on an average welfare family—a mother with two kids receiving AFDC, food stamps, Medicaid, and a smattering of other programs.
Under the Clinton plan, the 80 percent of current welfare mothers born before 1972 can breathe easy—they're not affected. Younger mothers will have a two-year grace period, during which they can participate in education and training programs to prepare them for the job world. After two years, those who haven't already found a private-sector job would have to enroll in a work program run by state and local agencies. A mother might be placed in a community-service program—picking up trash at a public park, stuffing envelopes at the public library—or in a government-subsidized job with a private company. Wherever she is placed, the mother would be required to work only 15 hours per week.
Robert Rector, a welfare-policy analyst at the Heritage Foundation, points out that for a welfare recipient working 15 hours, the average state's total cash, food, and medical benefits will work out to about $15 an hour, several times the minimum wage she would presumably garner in the market. By the year 2000, the Clinton administration predicts that some 400,000 welfare households, out of 5.7 million AFDC households, would be participating in work programs created by the president's plan.
Under Thompson's Work Not Welfare plan, a typical welfare family's situation would be radically different. Unlike the Clinton plan's emphasis on job training, its focus is on jobs—immediately giving welfare recipients the sort of work experience that would make it possible for them to enter (or reenter) the economy with marketable skills. In the affected counties, all AFDC recipients, regardless of age, will be subject to work requirements by July 1, 1995. The plan cashes out the food stamp program in the two target counties and adds the money to AFDC to create a so-called Independence Account for each recipient. This account consists of 24 months of (not necessarily consecutive) cash assistance provided in exchange for work, plus an additional 12 months of Medicaid and child-care benefits.
While the Clinton plan provides full benefits if a recipient works a couple of days a week, the Wisconsin plan requires her to earn the full value of her AFDC or food stamp payments by work at the minimum wage of $4.25 (though the total workload can't exceed 40 hours a week). In Wisconsin, a mother with two children on welfare receives a total cash grant of $729 ($517 in AFDC and $212 in food stamps). To earn her benefits at the minimum wage, the mother will have to work 172 hours during the month, or roughly 40 hours a week. If she works only 30 hours, her monthly cash assistance will be cut by a fourth. To further motivate recipients, the account must be drawn down within a four-year period. After either drawing down the account or reaching the four-year limit, that's it. No more assistance. And if you don't participate in Work Not Welfare, you immediately lose AFDC. While the enforcement measures may appear draconian, they are designed not to punish people but to spur them into self-sufficiency.
Wisconsin isn't the only state poised to try serious welfare reform. In Massachusetts, Gov. William Weld's welfare-reform proposal similarly imposes a work requirement: All able-bodied AFDC recipients would have to work to continue receiving assistance. Work would help break the cycle of dependency by changing "the daily physical routine of the recipients," said Weld at a recent American Enterprise Institute seminar. "Many AFDC mothers do not get out of the house. As a result of their isolation, they develop major self-esteem problems, and it grows harder and harder for them to become contributing members of society." Weld's program excludes teen parents, parents with infants, and disabled recipients, so the state expects only about 50 percent of AFDC families to participate. To smooth the transition to the wage economy, the plan converts $800 million in cash-assistance programs (primarily AFDC) into day-care subsidies for parents working in private or community-service jobs.
On the West Coast, Oregon has set the pace in welfare reform. In 1990, voters passed a ballot initiative that would have required AFDC, food stamp, and low-income unemployment insurance recipients to earn their benefits by working at 90 percent of the state's minimum wage. Implementing that initiative, however, would have required extensive waivers from the federal government that were either unlikely, in the case of food stamps, or, in the case of unemployment insurance, simply impossible, says Jim Neely, assistant administrator of adult and family services in Oregon's Human Resources Department.
In place of the initiative's reforms, Oregon is experimenting with a less-radical program called JOBS Plus, passed last year. Like the Wisconsin and Massachusetts plans, it places work at the center of reform. Jobs Plus establishes a six-month period of employment in either a private or community-service program in six target counties, but participation is optional. The state will collapse AFDC, food stamp, and unemployment insurance payments for recipients into payment for work at no less than the minimum wage ($4.75 in Oregon). The state expects to pay for the extra costs of the program (child-care and transportation expenses) with $2.7 million in start-up money from a state lottery and future savings from getting people employed and off public assistance. That could be tough, Neely says, because no one is sure "where the jobs [for welfare recipients] will come from and what it will cost. We think JOBS Plus can be a method for answering that question."
The current wave of state reforms reflects a new appreciation of the shortcomings and unintended consequences of conventional welfare policy. The new consensus that short-term assistance and work experience are the real keys to getting and staying off the dole is rooted in important scholarship published during the 1980s.
The 1984 publication of Charles Murray's Losing Ground directly challenged the efficacy of anti-poverty programs. The reaction from traditional welfare scholars and politicians alike was generally scathing, but at the same time Murray's straightforward rebuttal of the War on Poverty shifted the welfare debate substantially. Analysts could attack the system while eschewing Murray's more controversial assertions and policy prescriptions. "Until the mid-1980s welfare reform meant more benefits," explains Ron Haskins, a Republican staff member on the Ways and Means Committee of the U.S. House. "After Murray, reform also meant reducing dependency."
About the same time Losing Ground came out, researchers Mary Jo Bane and David Ellwood of Harvard University published several influential studies about the composition of the welfare population. They found that while most families leave welfare within two years, at any given moment more than 65 percent of the families on welfare are in the midst of spells lasting eight years or more. The implications from Bane's and Ellwood's studies, and from research along similar lines by Baruch College economist June O'Neill, are that lots of families move back and forth between welfare and work and that the number of families in long-term dependency builds up in the welfare caseload over time.
O'Neill has gone on to quantify the relationships among welfare, work, and marriage that Murray had originally postulated. She points out that welfare caseloads rose rapidly during the 1960s and early 1970s but then leveled off and even fell as a percentage of the population during that horrible decade, the 1980s. Only recently, since 1989, have caseloads again been going up in most states.
Interestingly, there is a close correspondence between growth in caseloads and growth in total welfare benefits (including food stamps, Medicaid, and other programs). O'Neill speculates that the current increase in caseloads is probably the result both of the recent recession and of the 1988 Family Support Act, which increased the value of welfare benefits.
And she notes that when the new Reagan administration and Congress decided to reform the old Work Incentives program (WIN) in 1981, lowering the amount a person could earn and still remain on AFDC, caseloads fell in many states. The 1981 reform "may be the only federal legislation ever to have been enacted which resulted in a decline in welfare participation," O'Neill observes.
A General Accounting Office analysis of the 1981 legislation found another interesting result: By limiting eligibility for AFDC, it also reduced Medicaid coverage for many low-income workers. But the GAO found that these workers did not have higher rates of unemployment than similar recipients did before the change, calling into question the popular assumption that for welfare reform to succeed you must guarantee Medicaid coverage during the transition from dependency to a job with health insurance.
In a 1993 study funded by the U.S. Department of Health and Human Services, O'Neill and M. Anne Hill discovered that a 50 percent increase in monthly AFDC and food stamp benefits led to a 75 percent increase both in the number of women enrolling in AFDC and in the length of time spent in the program. More important, they found that higher AFDC benefits in a given community suppressed employment of young adult men by reducing the probability of marriage (and the responsibility it creates for providing for a family) and by providing a source of income through girlfriends receiving benefits. The study also found that children raised in families that receive welfare assistance are themselves three times more likely than other children to go on welfare once they become adults, and that a 50 percent increase in benefits led to a 43 percent increase in the number of out-of-wedlock births.
Yet another body of research inspires current reform plans, and it comes from the experience of previous state efforts. When Congress reformed WIN in 1981, it also gave states more flexibility in designing and implementing job-training-and-placement programs. Of the 30 states that responded with demonstration programs, eight contracted with Manpower Demonstration Research Corporation to help design and evaluate their programs. These programs were set up to be true research tests, with random assignment of participants into experimental and control groups, so the results are fairly solid. MDRC found that the states that combined job search and work programs could boost incomes or employment rates somewhat and slightly shorten welfare spells. Programs with extensive amounts of education and job training didn't pan out. Generally speaking, the cost of the successful programs was small ($800 per participant) and was offset by their benefits, albeit modest.
Chuck Hobbs, who headed an important Reagan administration commission that advocated greater flexibility for state experimentation, sums up the research and experience of the past decade or so this way: Put welfare recipients to work massively with few exceptions. Merge the 76 or so welfare programs (costing local, state, and federal taxpayers around $300 billion annually, according to the Heritage Foundation) and use the money to finance job searches and work. Limit the time recipients can receive public assistance, even when it is received in exchange for work. And, stresses Hobbs, "Don't waste time on education and training—work is the best training." Hobbs maintains that work builds self-reliance, which in turns builds self-respect and reduces the need for ongoing social programs.
Given the quality and quantity of the supporting research, work-driven welfare reform is eminently plausible. But it remains to be tested wholesale in the real world—and it may never be. The Heritage Foundation's Rector has carefully read the text of Clinton's own welfare bill and points out two provisions that would trump existing and proposed state reforms alike.
First, the bill appears to limit the conditions on which a state can impose work requirements to that specified in the bill itself—in other words, says Rector, Clinton's real message is that, "If you don't meet these conditions, you don't have to work." Second, the bill stipulates that anyone made to work as a condition for receiving benefits must be paid the prevailing wage for the job they take. This provision could well serve as a sort of "quasi-Davis-Bacon Act," Rector says, protecting unionized workers and pricing low-skill welfare recipients out of the market for available work. In short, the bill, which he calls "tough on the outside, gooey on the inside," will set the welfare-reform debate back about 15 years.
In fact, states seeking to try out radical welfare experiments already face significant barriers. The Oregon JOBS Plus proposal, for example, illustrates the need for more flexi-bility in federal welfare oversight. The plan's relatively cautious design—eschewing tough work mandates—is a response to federal intransigence about the state's original, voter-approved initiative.
"The current waiver process is difficult at best and not very satisfactory for anyone," Oregon's Neely says. "It is intended to test new ideas on a very small scale for use as information by the Congress to draft national policy. It doesn't work for states seeking to change their policies on their own."
Jack Svahn, the chairman of MAXIMUS Inc., which has contracts in 20 states to run job-placement and child-support-collection programs, has viewed the process from both the state and federal sides. He agrees that states lack the flexibility they need to change their welfare policies. "What happens in New York City is not necessarily going to work in Wyoming," says Svahn, a former state welfare director in California and high-ranking official in the Reagan administration. "The more power you can devolve, the better off you will be." But the Clinton welfare bureaucracy (which includes researchers Ellwood, who helped design the president's welfare plan, and Bane, the deputy HHS secretary in charge of granting waivers) has been seeking ways to squelch state reforms that go "too far," Svahn alleges.
Congress has been getting into the act, too. Both Massachusetts senators, Ted Kennedy (D) and John Kerry (D), oppose Weld's tough workfare plan and want the administration to deny it waiver approval. Sen. Patrick Leahy (D-Vt.), chairman of the Agriculture Committee, tried in July to strip the U.S. Department of Agriculture's authority to grant waivers to states that want to convert food stamps to cash for use in welfare reforms. "Providing cash undermines the character of food stamps as a nutrition program," Leahy said. It also undermines their more-fundamental character as an agriculture subsidy, a fact that surely did not escape the farm-state senator's notice.
The Senate, in spite of Leahy, voted to continue the waiver policy. But a statutory problem remains: the USDA will not let states make food stamps conditional on work. Wisconsin's Work Not Welfare originally proposed that both AFDC and food stamps be cut off to welfare recipients who refuse to work, but the USDA killed the idea.
Over at HHS, a policy change announced in May would require all states to follow an intrusive, time-consuming process for allowing public comment on waiver requests before they are sent to Washington. In defending this policy, Bane suggested that state welfare-reform experiments need to be examined on the basis of, among other things, that "protected groups"—minorities—"are not adversely affected." The National Governors Association notes that such a policy is clearly designed to give public-employee unions, civil rights organizations, and others opposed to serious welfare reform a "vehicle for national appeal of a waiver application," with which they can bog down a reform proposal for months or even years. Even reform opponents without big pockets can gum up the works, thanks to a Senate vote in late July to allow Legal Services Corporation representation for poor clients seeking to overturn welfare reforms that reduce their benefits.
Court challenges led by Legal Services lawyers, in fact, already pose a significant risk to the welfare-reform movement. For example, the U.S. Supreme Court in January rejected an appeal by state officials in California and Minnesota of a federal ruling preventing states from limiting welfare benefits to new residents. Then in July, a federal appeals court in Sacramento struck down millions of dollars in California welfare cuts implemented under a 1992 waiver from the Bush administration. The court stated that the state government had failed to consider the "increased risk of homelessness, inadequate nutrition, and variety of emotional and physical problems" the cuts could create among the state's poor. Observers on both sides of the issue expect the California ruling to affect experiments in other states that reduce benefit levels.
Even if the feds weren't prepared to limit state welfare experimentation, success would still be far from guaranteed. A big question is how to line up private-sector jobs for welfare recipients who, lacking them, will end up working for the government or a government-approved non-profit. Weld argues that since immigrants, many uneducated and unable to speak English, seem to find work in most job markets, welfare recipients should be able to as well.
But states with work requirements are having problems lining up jobs, often because employers aren't willing to hire welfare recipients due to concerns about their skills, commitment, and honesty. And, even though states are making more use of private job-placement contractors whose profits are tied to performance-based incentives, they're slower to reduce state and local taxes, remove excessive regulations—including restrictions on firing people—and streamline other government impediments to job creation for low-skilled workers. High federal payroll taxes and the national minimum wage also make it unattractive for employers to retain workers once their welfare subsidies end.
Without available jobs, work-based welfare reform will short-circuit. As a Pensacola, Florida welfare administrator remarked to The Wall Street Journal about that state's newly imposed work requirement for some recipients: "If we can't find them a job at the end of two years, we haven't held up our end. Then the contract is broken"—and the recipient goes back on welfare.
Another concern is that any reform program promising generous child-care, transportation, and medical benefits to ease the transition from welfare to work might instead entice more low-income workers to apply for welfare in the first place, gutting any cost savings or reductions in overall dependency. Heritage's Rector observes that if a state or national welfare reform is expected to cost more to operate than current programs do, "it is destined to fail."
That's one reason why an alternative to the Clinton plan, designed with help from Rector and introduced by Sen. Lauch Faircloth (R-N.C.) and Rep. James Talent (R-Mo.), would be a sound first step toward serious welfare reform. The plan would require 50 percent of adult welfare recipients to be working by 1996. For AFDC recipients under the age of 21, the bill takes a different tack. It would collapse about 60 different federal welfare programs into a bloc grant, which would be given directly to states with only two major provisos: They must use the money to assist low-income people, and no money could be given directly as cash assistance to young women who have children out of wedlock. States could use the money to expand adoption programs, chase down deadbeat dads, create group homes for young women and their children, or whatever else they want to try. And welfare would no longer be a matching-funds program: Federal spending would be strictly limited to 3.5 percent annual growth.
The Faircloth-Talent bill faces long odds—"If it goes any place, be surprised," says a Senate staffer—but it has served to unite many House and Senate Republicans behind real reform and, just as important, against Clinton's sham reform. Bill Kristol, head of the Project for the Republican Future, says that while serious reform may not pass this year, the GOP's goal should be "to shape the debate and position ourselves to advance more thoroughgoing changes to the welfare system next year."
It may well be possible to toughen the bill next year, especially if Republicans are successful in November House and Senate races. A Republican Senate or Republican-moderate Democrat alliance in the House could rewrite Clinton's bill, or at the very least force the administration to preserve and expand state authority to experiment with work requirements and other reforms, a position that Clinton himself has endorsed in the past. State reformers have demonstrated their seriousness about working out the problems of moving people off of welfare, and should be allowed to go to it.
Will we ever really see an "end to welfare as we know it"? It's a strong possibility, if Washington will get out of the way. In Madison, they even know the date: Wisconsin passed a law last year that will officially end the AFDC program statewide on December 31, 1998.
Who Are the Real "Welfare Magnets"?
For years, one of the most controversial issues in the welfare-reform debate has been the role of states in setting benefit levels for recipients. Self-styled "advocates for the poor" claim that allowing states to vary the size of welfare benefits is inequitable and discriminatory. Advocates point to the fact that, depending on where a poor person lives, benefits can differ dramatically. Fiscal conservatives often complain about "welfare magnets," states that attract migrants who come for the dole.
In every case, however, those arguments use nominal-benefit figures that do not account for the obvious cost-of-living differences between, for example, Boston, Massachusetts, and Biloxi, Mississippi. In nominal terms, the gap in benefits (including AFDC, food stamps, and Medicaid) is indeed great between Massachusetts, the most generous of the 48 contiguous states and the District of Columbia, and Mississippi, the least generous—$15,872 vs. $7,621 in 1991. But when cost-of-living adjustments are factored in to the comparison, the gap between the two benefit levels shrinks from $8,251 to $4,731. And when you compare benefits to the median family income in each state, you find that 42 states fall within the relatively narrow band of 25 percent to 35 percent.
Several states change rankings dramatically after adjusting for the cost of living. Some states traditionally thought of as welfare magnets turn out to be slightly below the middle of the pack—California plummets from 6th to 37th, the District of Columbia drops from 7th to 39th, and New Jersey slides from 9th to 38th. On the other hand, states in the Great Plains lacking the reputation of welfare generosity end up near the top of the list—Utah (2nd), North Dakota (4th), Oklahoma (6th), South Dakota (8th), and Iowa (9th). Even after the adjustment, state and regional differences persist in average-benefit levels, but their magnitude has been consistently exaggerated.
Contributing Editor John Hood is research and publications director at the John Locke Foundation in Raleigh, North Carolina, and a syndicated columnist.