Three years into America's grand welfare reform experiment, policymakers nationwide are taking stock. They are learning from their mistakes, building upon their successes, and anticipating and planning for future challenges.

We have come a long way from President Clinton's 1992 campaign promise to "end welfare as we know it" and the enactment of federal welfare reform legislation four years later. Still, statistics and anecdotes suggest the effort to change welfare from a creator of dependency into an engine of opportunity is far from over. The biggest unanswered question is: Will successful reform be defined simply as reducing caseloads or doing the hard and sustained work often necessary to help former recipients lift themselves into the economic and social mainstream?

Between 1994 (when many state welfare caseloads peaked following the 1990-93 recession) and June 1998, state welfare rolls declined by an average of 44 percent, dropping from a total of 14.3 million to under 8 million. In Wisconsin, a welfare reform leader, rolls declined by a remarkable 87 percent from 1993 to 1998. Even in California, where a prolonged recession meant caseloads reached their record highs a few years later than most other states, rolls have fallen by almost 20 percent in the last three years. Despite fears that early declines had "creamed" those recipients most able to make it on their own, caseloads continued to fall at a steady or even accelerating rate in most states until quite recently.

Surprisingly sharp reductions in welfare rolls have exceeded reformers' expectations and confounded critics. According to the National Conference of State Legislatures, between 50 percent and 60 percent of those leaving welfare are finding jobs. In a recent telephone survey, 54 percent of recent welfare recipients in New York City said they went off the rolls because they had found full-or part-time work; their median wage was $7.50 an hour. A second survey of recent recipients in Wisconsin found that 62 percent were presently employed and 83 percent had worked "sometime" in the past few months; the median wage for both groups was $7.00 an hour. Nation-wide, people leaving welfare for jobs earn, on average, only slightly more than minimum wage, rarely enough to raise a family out of poverty.

"Making work pay for people who play by the rules" is likely to be a potent rallying cry when inevitable adjustments to welfare reform are made. Fortunately, a number of mechanisms to make work pay already exist, including food stamps, Medicaid, child care and transportation subsidies, and the Earned Income Tax Credit. Some will need to be readjusted. The EITC, for example, currently undermines two-parent families by paying far more to two single earners than it does to married earners. In many states, income transfers supplementing part-time work count against a person's lifetime welfare limit as if the person were not working. Despite such problems, most of these programs are pointed in the right direction: They support work financially, not just rhetorically.

The Sources of Success

What is the secret to welfare reform's success? Obviously, the strong national economy has created tight labor markets pulling more people into the work-force. The nation's unemployment rate is near a 28-year low; consumer confidence is approaching a 29-year high; the home-ownership rate is an unprecedented 67 percent; and stock-price indexes hover near their record highs. Recent monthly surveys by the National Federation of Independent Business show that labor shortages now tie with taxes as the major problem cited by member businesses.

But a strong economy is only half of the explanation. The other part of the story is that the country's expectations about welfare have changed. Those expectations are being institutionalized in welfare rules emphasizing work, self-sufficiency, and reciprocity. Americans have discovered the importance of "job readiness" (as opposed to "job training") programs, often run by nonprofit or for-profit intermediary firms rather than by government. These programs stress the "soft skills" that employers consistently say are so important: showing up on time, getting along with fellow workers and supervisors, and having a positive attitude. The best programs -- such as America Works and STRIVE, each with several locations across the country -- continue to coach ex-recipients on how to keep their jobs and progress in their new places of employment. They also help them overcome such barriers to work as transportation, child care, housing, disability, and substance abuse.

The national Welfare-to-Work Partnership is working with approximately 10,000 companies to hire people moving off welfare. Intermediary organizations connect these firms with pre-screened, work-ready candidates who can be brought up to speed quickly.

As a result, many companies report hiring welfare recipients as a positive experience and good for the bottom line. They note that retention rates for workers hired from welfare are often much higher than rates for other new hires holding similar jobs. United Airlines, for instance, retained 70 percent of the 760 welfare recipients it hired in 1997, compared to 40 percent of other workers it hired for similar jobs. Likewise, Giant Food, a supermarket chain based in the mid-Atlantic region, reports that 79 percent of ex-welfare recipients are still on the job three months after being hired versus 50 percent of all other comparable new workers.

There is no "one-size-fits-all" program for moving people from welfare to work. Some need more help than others in finding and keeping a job. Part of the challenge is steering ex-recipients to the job-readiness program most likely to work for them. STRIVE, for example, uses both a "boot camp" approach and long-term follow-up to help people who have been out of the labor force for a long time, including those coming out of prison or over-coming an addiction. Chicago's Project Match frequently provides up to two years of employment and related support services to chronically unemployed residents of the Cabrini-Green housing project. America Works, meanwhile, achieves remarkable results with a one-week "work readiness" course and four months of follow- up after placing ex-recipients in transitional jobs. A New York state study showed that 82 percent of individuals who found jobs through America Works were not on public assistance 14 months after job placement.

Competing Incentives

The 1996 welfare reform bill helped both to drive case-loads down and, to a lesser extent, to reorient welfare toward work.

Welfare is now time-limited and individuals receiving it must comply with various "work-related activity" requirements. States, meanwhile, can suffer financial penalties or reap rewards based on their performance against the law's minimum rates of caseload decline. States must move a rising percentage of their caseloads (25 percent in 1997, rising to 50 percent in 2002 and beyond) into approved "work activities," including subsidized and unsubsidized employment, community service, work experience, and limited job-search and job-readiness training. States missing their caseload reduction targets can lose from between 5 percent and 21 percent of their new federal welfare block grants, while states managing to cut their caseloads below their 1995 levels are rewarded with bonuses.

As Milwaukee Mayor John Norquist has observed, the law gives states more incentives to drive down the rolls than it does to move people into private sector jobs. Norquist favors performance-based contracting that makes agencies accountable for moving "participants out of poverty and into full-time private-sector jobs."

The states' welfare block grants are based on 1995 caseloads, when welfare rolls were considerably higher than at present. Today's smaller caseloads have yielded a combined state welfare surplus of more than $3 billion. States in the black can either spend more per recipient than in the past, use the extra funds to finance policy innovations, or set up reserves for future economic downturns. Many state officials, like Iowa's welfare director Douglas E. Howard, worry that politicians in Washington will decide that "if you are not spending the money, you must not need it."

The Best Indicators of Success

President Clinton and other politicians around the country proudly display graphs illustrating steep declines in caseloads. But where are the graphs showing the number of jobs former welfare recipients have found and their average wages? Such numbers, which are far more difficult to track, are probably the best indicators of success.

We know from indirect evidence that homelessness and hunger have risen modestly since welfare reform began. But we have no way of knowing what actually happens to families dissuaded from signing up for welfare and left to their own devices. Do the heads of households find jobs? Do they end up on welfare later? Do they receive other forms of public assistance or resources from their communities and extended families? In other words, do they become self-sufficient? We don't yet know the answers to such questions because virtually all studies of welfare reform's effects track only those who were on welfare and left. We know of no studies tracking the "would-have-beens": people diverted from the rolls in the first place.

There are many people now on welfare who will not be able to find or keep a job. Drug addiction, alcoholism, mental illness, physical disability, and other conditions will prove to be insurmountable obstacles for an uncertain proportion of the welfare population. The welfare reform law assumes that 20 percent of those on the rolls can't work and so exempts that proportion of the case-load from the five-year time limit. But many people in the welfare field think this figure is overly optimistic. In the future, welfare workers will have to be sensitive toward people with real limitations even as they confront the old, self-fulfilling excuses of others who've lost their work-related connections and confidence or have found other ways to avoid getting a job.

Final Questions

As time limits start to come due and more of the "hard cases" are left on the rolls, what perils lie ahead? Three questions bear close watching: