Everyone wants -- or professes to want -- to "end welfare as we know it." Despite such lofty proclamations, welfare is still thriving. Last year, federal and state governments spent $411 billion on means-tested welfare programs that provide cash, food, housing, medical care, and social services to poor and low-income Americans. This greatly exceeded the $324 billion spent in 1993, the first year of the Clinton presidency.

At the core of America's vast, dysfunctional welfare system is Aid to Families with Dependent Children (AFDC). At present, nearly one out of seven children in the United States receives AFDC, residing with a mother married to a welfare check rather than a working husband. The typical family now on AFDC will spend nearly 13 years in the program.

"Ending welfare" must begin with reform of AFDC. Congress enacted major new legislation last summer that will start this process. The new law promises three major changes. First, it eliminates the entitlement system of AFDC funding, under which states that increased their AFDC caseloads received automatic increases in federal funding, while states that reduced dependence faced a fiscal penalty.

Second, the new law establishes performance standards that will require each state to reduce its AFDC caseload, or at least, if the caseload does not decline, require some recipients to work in return for their benefits.

Third, the law sets a new goal of reducing illegitimacy and will reward states that reduce out-of-wedlock births without increasing the number of abortions.

Although the new federal legislation sets the proper framework for reform among the states, the liberal welfare establishment and its allies in the media incessantly warn that reform will prove to be difficult, if not impossible. But one state has already proven the naysayers wrong: Wisconsin. Wisconsin's experience with welfare reform provides an unparalleled model for implementing reform that other states would be wise to follow.

In the last 10 years, while AFDC caseloads in the rest of the nation were rising steeply, the caseload in Wisconsin has dropped by half. In inner-city Milwaukee, the caseload has fallen by 25 percent, but in the rest of the state, caseloads have fallen by nearly 70 percent. In 28 of Wisconsin's 77 counties, the welfare rolls have already dropped by 80 percent or more.

And if all this weren't remarkable enough, the pace of Wisconsin's reduction in welfare dependency is accelerating. In Milwaukee, the AFDC caseload is now shrinking 2 percent per month; in the rest of the state, 5 percent. Wisconsin's achievements are utterly unprecedented in the history of AFDC. Liberal welfare experts used to insist that a successful work program might reduce welfare caseloads by 5 percent over five years; in much of Wisconsin, the number of people on welfare is steadily falling by that amount every 30 days.

Wisconsin has thus won more than half the battle against AFDC dependence and is proceeding with the other half with breathtaking speed. This victory is crucial, since welfare dependency severely hampers the healthy development of children. In the long term, the greatest beneficiaries of Wisconsin's dramatic achievements in reforming welfare will be the children themselves.

This remarkable story begins in 1987, when a major congressional debate on welfare culminated in the Family Support Act (FSA). Touted as yet another "end of welfare," the FSA was a complete bust. The Act did, however, generate the expectation among voters that welfare recipients would be required to work. In the same year, a second unheralded event occurred with far greater significance for the future of welfare: Tommy Thompson took office as governor of Wisconsin.

Following a gubernatorial campaign largely about welfare, Thompson entered office with a firm commitment to reform. The chart at right tells the rest of the story. Despite the rhetorical promises of the Family Support Act, the nationwide AFDC caseload remained constant in the late 1980s and then grew by more than a third between 1990 and 1994. The nationwide caseload has eased downward over the last two years, but the majority of states still suffer higher levels of welfare dependency than before the Family Support Act became law.

Wisconsin has been the only clear exception to this pattern. Upon taking office, Thompson initiated a series of reforms that cut welfare dependency during the late 1980s and blocked any resurgence during the 1990-93 recession. Starting in 1994, a second round of more sophisticated work-related reforms has caused the caseload to nosedive further. But the raw figures understate Thompson's achievements. As noted, welfare rolls across the country ballooned by some 35 percent during the early 1990s. There is every reason to believe that, without Thompson's reforms, Wisconsin would have followed this national trend. If it had, its AFDC caseload would have surged from around 100,000 recipients in 1987 to a peak of 135,000 in 1993. It is reasonable to conclude that Thompson has not merely cut his state's caseload in half (from 98,295 recipients to 48,451) but has reduced it by some two-thirds relative to the potential peak in dependence that Wisconsin would have experienced in the early 1990s in the absence of reform.

Many states brag about their recent declines in welfare dependency. In the past 24 months, for example, Indiana has cut its caseload by 32 percent, Oregon by 30 percent, Maryland by 29 percent, Massachusetts by 25 percent, Oklahoma by 24 percent, and Michigan by 22 percent. But, in almost every case, these successes merely represent a pruning back of the explosive surge in welfare dependency of the early 1990s. In reality, other reforming states lag a half-decade behind Wisconsin; they are only now engaging in the initial stages of dependency reduction that Wisconsin accomplished in the late 1980s. We might say that the reforms in most states have merely blown the foam off the top of the beer mug, while Wisconsin has already drained the mug halfway to the bottom.

Having cut its caseload in half, Wisconsin's reformers are now grappling with a more difficult, less employable group of welfare recipients. Skeptics argued that after Wisconsin weaned the most employable recipients off the rolls in the early stages, the decline in the caseload would slow and then stop. Yet the opposite has occurred. As new reforms have been implemented over the last three years, the decline has accelerated sharply. Wisconsin continues to reduce dependency at a rate surpassing all other states.

The general thrust of welfare reform in the Thompson administration has been to require reasonable behavior by recipients as a condition of receiving aid. An early example was Learnfare. Enacted in 1987, the Learnfare program required welfare recipients to ensure that their school-age children attended school regularly, and reduced welfare payments to families with truant children. Although Learnfare did not reduce the AFDC rolls directly, it did have a symbolic importance, sending a clear message to both the bureaucracy and the welfare clientele that, for the first time, the government seriously intended to demand constructive behavior of welfare recipients and to sanction those who were derelict.

The centerpiece of reform, however, was the requirement that a growing share of those on the AFDC rolls engage in employment-related activities such as training and tightly supervised job search. By the 1990s, many Wisconsin counties were operating sophisticated systems aimed at pushing welfare recipients quickly into the labor market. For example, a particularly effective program in Sheboygan County required most AFDC recipients to undertake closely supervised job search immediately after applying for benefits. Individuals who failed to find employment within a few weeks were required to perform community-service work until they could find a private-sector job.

Beginning in 1994, Thompson's staff initiated a more sophisticated and successful round of reforms. His administration instructed county welfare directors to de-emphasize education and job training in the classroom and to concentrate on activities leading to immediate work. The state created new incentives to guide the welfare bureaucracies: Counties would no longer be simply allocated work, training, and day-care funds but would be required to earn those funds by increasing the number of recipients placed in jobs or community-service work.

The governor's staff also understood that many applicants entering the welfare system had other options. The easiest way to break this group of the debilitating habit of dependency was to prevent it from forming in the first place, by reducing the number of new AFDC enrollments. The state began a pilot program based on this principle in 18 counties in 1994 (gradually expanded to cover 60 by early 1996). The program, Work First, provided new applicants with counseling on the negative effects of dependence, offered short-term aid (such as car repairs) that might eliminate the need to enroll in AFDC, and required most new applicants to begin working in private-sector jobs or community-service jobs almost immediately after enrolling in welfare. (Once a recipient starts a full-time job in the private sector, AFDC benefits are eliminated; however, AFDC benefits generally continue at a reduced level while the recipient works part time. By contrast, a recipient performing community-service work continues to receive full AFDC payments but must work for the benefits obtained.)

Another pilot program called Work Not Welfare (WNW) began in January 1995 in two counties, Fond du Lac and Pierce. WNW placed an absolute time limit of 24 months on receiving AFDC. Although this program obviously had no direct impact on dependency in most of the state, it did send a strong symbolic message to welfare recipients: Long-term dependence would no longer be tolerated. (Still, caseloads in those two counties have fallen only slightly more rapidly than in most other counties.)

In December 1995, the governor's staff instituted a radical new system for rewarding good performance in county welfare offices. The new system employed competitive bidding for the management of each county's welfare system. Governmental or private organizations currently running AFDC in each county were thus threatened with competition and could be replaced by outside organizations. The current welfare organizations, however, could escape the competitive bidding process if they fulfilled new performance criteria specified by the state. Chief among these criteria was a requirement that each county reduce its AFDC caseload by roughly 15 to 25 percent (requirements varied by county) over the subsequent 12 months.

                                                    Decisive New Policies

In April 1996, Wisconsin unveiled two more decisive reforms. Work First (renamed Self-Sufficiency First) went into effect in all counties including Milwaukee, and a new Pay for Performance (PFP) system was implemented statewide. For decades politicians have talked about making welfare recipients work while creating regulations that made this impossible. With Pay for Performance, Wisconsin closed the loopholes, creating for the first time a real work requirement for AFDC recipients. Prior to PFP, a recipient who failed to obtain a private-sector job might be required to perform community service. If the recipient failed to actually perform this service work, however, the state could only cut AFDC benefits slightly. Under PFP, a recipient would see his welfare check reduced in direct proportion to the number of hours of community work he fails to perform. An individual who performs no work would receive no AFDC or food stamps. Critically, the PFP principle was also applied to other constructive activities such as class attendance or supervised job search.

PFP effectively eliminated the freedom of most Wisconsin AFDC recipients to receive welfare without working. Caseloads, already in rapid decline, began to plummet. In the first seven months after the implementation of Pay for Performance and Self-Sufficiency First, the AFDC caseload dropped 14 percent in Milwaukee and 33 percent in the rest of the state. If the trends continue, the welfare rolls statewide will drop by an additional one-third over the next 12 months.

Apologists for big welfare have naturally sought some pretext for ignoring or trivializing Wisconsin's achievement. The most common ploy is to attribute the decline in AFDC caseload to a "good economy." Although Wisconsin has enjoyed low unemployment and healthy growth in jobs, it is ridiculous to claim that this has brought about a dramatic reduction in caseload. After all, over the past 40 years states have often experienced robust economic growth without any significant drop in the welfare rolls, let alone a drop of 50 percent.

The limited role of economics in Thompson's victory over dependence can be seen by comparing Wisconsin with the 13 states that have experienced lower levels of unemployment than Wisconsin over the past decade. In these states, the welfare rolls, on average, actually increased by some 20 percent. None produced a substantial decline in welfare caseloads. Although a robust economy has undoubtedly helped Wisconsin to reduce welfare dependency, it is far from the principal cause.

Another common dodge of the skeptics is to claim that reform has raised welfare costs. This charge is no surprise; defenders of the status quo have always claimed that taxpayers must "invest" more funds in order to "end welfare." But Wisconsin's reforms did not result in increased spending in either the short or long term. Although the state has increased its outlays on welfare administration, job training, and day care, these expenditures have been more than offset by the rapidly shrinking caseload. Wisconsin spends more per family on welfare now than in 1987, but it has half the number of families on welfare. The greatest expenditure increase has been for welfare administration; although day-care costs have also increased modestly, they still constitute only 6 percent of the total. Today Wisconsin's aggregate spending on AFDC benefits, administration, training, and day care, in current dollars, is some 5 to 10 percent lower than in 1986, the last year before Thompson became governor. But in the rest of the nation, similar expenditures have nearly doubled in the same period: Clearly, Wisconsin's reforms have produced huge de facto savings, not higher costs. In inflation-adjusted terms, Wisconsin's spending is actually down by a third since 1986.

The Wisconsin experience provides a cornucopia of lessons for the rest of the nation. In reforming welfare, Wisconsin has rediscovered a philanthropic philosophy once ubiquitous in American charities, but largely abandoned over the past 40 years. This philosophy regards dependence and idleness as harmful to the welfare recipient and insists that he perform useful labor in exchange for benefits he receives.

Thus a serious work requirement provides not only a sound moral foundation, but also performs a crucial gatekeeping function. A key problem for rational charity is separating those who truly need aid from those who do not but are willing to take a free handout if one is offered. Work requirements serve that purpose. For example, in the 19th century, religious organizations throughout the United States provided food and shelter to persons who would today be called "homeless." Before providing a free meal and a bed, however, the shelter would require the man seeking aid to perform some useful chore such as chopping firewood. Charity workers had discovered that such a requirement greatly reduced the numbers seeking aid. This "work test" winnowed out those who did not need aid and allowed the philanthropists to focus limited resources on the truly needy.

With one out of seven children enrolled in AFDC, the current system is so large that serious reform will be impossible. The initial task in transforming welfare is to shrink the AFDC caseload to manageable proportions. Culling those who do not truly need aid from the welfare rolls will allow the system to focus its efforts on those who have the most difficulty becoming self-sufficient, and will free up resources and energy needed to deal with the underlying problems, such as educational failure and illegitimacy, that promote future dependence.

Wisconsin's example provides nine clear rules on how to sharply reduce dependency:

1. Set the right goal. If the ultimate aim of reform is to reduce dependence, the official goal must be to reduce the welfare caseload. A large drop in caseload entails dramatic administrative change and threatens the financial self-interest of the welfare industry. Ingenious welfare bureaucrats will thus propose other performance criteria that allow them to claim success in reducing dependence while caseloads continue to rise. Such ersatz benchmarks generally include: the length of time spent on welfare; the number of recipients in training, part-time employment, or make work-jobs; or the number who leave welfare. Decisionmakers should not be fooled: It is the size of the caseload that matters.

2. Focus on the size of the caseload, not welfare exits. Measuring the number of recipients who leave welfare -- or "exits" -- is misleading. Large numbers of "exits" from welfare will occur even when welfare caseloads are rising. States with liberal welfare systems may have larger numbers of exits because they encourage highly employable persons to enroll in welfare. By contrast, a serious work requirement may actually reduce welfare exits since it will discourage the most employable persons from enrolling in welfare in the first place.

3. Avoid education and training. Government training and remedial education programs in general do not increase recipients' wage rates and do little to reduce dependence. A recent Labor Department study of the government's largest training program, the Job Training Partnership Act (JTPA), found that the program had little or no effect on the wages of trainees: The average hourly wage rate of female trainees rose 3.4 percent, while the hourly wages of males did not increase at all.

4. Use work requirements to reduce welfare applications. The most important effect of a work requirement is to reduce dramatically the number of persons who apply for welfare. This is called the "dissuasion" effect of work requirements. By operating programs such as Self-Sufficiency First and by requiring most new applicants to find private-sector employment or perform community-service work shortly after enrolling in welfare, Wisconsin has cut the number of new AFDC entrants almost in half over the last two years.

 

5. Require continuous activity. In the private sector, employees are expected to work continuously, not intermittently. This principle must be duplicated in welfare. Once a recipient begins supervised job search, training, or work, some activity should be required without interruption or lessened intensity until the recipient leaves AFDC. In order to reduce welfare recidivism, the work obligation should resume as soon as a former welfare recipient returns to the AFDC rolls.

6. Establish a pay-after-performance benefits system. Welfare should be based on "pay-after-performance": Recipients will not receive the welfare check until after they have performed work or other required activity. If they fail to perform the required number of hours of activity, the welfare check must be reduced on a pro-rata basis.

7. Use community-service "workfare" as an enforcement mechanism. Upon applying for welfare, employable recipients should be required to begin a supervised search for employment. If they have not found a private-sector job within six weeks, they should be required to perform community-service work. Of course, the real goal of reform is to see that recipients obtain private-sector employment, not to push them into make-work jobs. But in a conventional welfare system, large numbers of recipients will claim they cannot find private-sector jobs. If such "unsuccessful job seekers" are permitted to remain idly on the rolls, reform will fail. Instead, all individuals who fail to obtain private-sector jobs should be placed immediately in community service slots on a pay-for-performance basis.

This effectively eliminates any recipient's chance of receiving a welfare income without working, and pushes recipients into private-sector jobs while dissuading other individuals from entering welfare. Mandatory community service is thus the crucial backstop to a serious work requirement. Of course, this does not mean that large numbers of recipients will end up in make-work community service. In Wisconsin few do, but the threat of community work is the key to propelling recipients into the private sector.

8. Impose work requirements on the most employable recipients first. The initial goal of welfare reform should be to restrict welfare to those who truly need it and to eliminate from the rolls those who do not. In order to accomplish this goal and to shrink welfare caseloads, work requirements should be focused on the most employable welfare recipients first. These would include two-parent families (10 percent of the caseload in a typical state) and mothers who do not have preschool children (typically 50 percent).

This strategy may seem counterintuitive, but it is essential to reducing dependence. The number of crucial community-service work slots (where the recipient is required to work for benefits) in the first phases of reform will be quite small in relation to the overall caseload. If the least employable recipients occupy these slots, they will remain there for long periods, clogging up the system. When highly employable recipients, by contrast, are faced with the prospect of performing community-service work, most will respond by quickly leaving AFDC, freeing the work slots for others, who will in turn leave the rolls. Through this revolving process, the caseload will begin to shrink quickly. (A variant of this principle is to focus work requirements on recent applicants who are, in general, more employable than the rest of the caseload.)

9. Establish bureaucratic incentives and competition. Throughout the United States, most of the welfare industry is liberal, regards welfare recipients as victims of social injustice, and is threatened by reforms which will sharply reduce its welfare clientele. In order to ensure the faithful and efficient implementation of conservative reforms, decisionmakers must establish precise performance criteria linked to rewards and sanctions for the welfare bureaucracies. In Wisconsin, welfare offices were forced to compete with one another to earn funding, and ultimately each county office faced the threat of elimination if it failed to meet high performance standards set by the governor.

 

The lessons from Wisconsin greatly influenced the national welfare reform enacted in Washington last year. The new federal law encourages states to pursue work policies similar to Wisconsin's. Among the specific features of the new federal law drawn from Wisconsin are the federal performance standards based on caseload reduction, the use of workfare to "dissuade" new applicants, and a requirement that states set up pay-for-performance systems.

Tommy Thompson has shown that state governments can overcome AFDC dependency. The greatest benefit will accrue to children. In the past, liberals have been mesmerized by the belief that "poverty" somehow harms children and that welfare, by "combating poverty," is therefore good for kids. Hence they are timid, if not outright adversarial, in their attitude toward serious efforts to reduce welfare dependency. But studies that compare children on welfare with poor children not on welfare show that it is actually welfare dependency, not poverty, that harms children. A childhood of welfare dependency lowers children's IQs, increases their likelihood of academic failure, and diminishes their future earnings as an adult. Welfare is a system of child abuse; by radically reducing dependence, Wisconsin's reforms will improve the future well-being of children.

There is, however, one very important shortcoming to Wisconsin's welfare achievement: The current reforms have not cut the state's illegitimate birth rate. Illegitimacy does much more harm to children's development than does welfare dependency. The ultimate goal of reform must be not only to reduce dependency but to rebuild marriage. One can only hope that over the next decade, Wisconsin's reformers will tackle the problem of out-of-wedlock births with the ingenuity and diligence they have already applied to the question of dependence.

Throughout his tenure as governor, Tommy Thompson has routinely accomplished what the welfare industry declared impossible. He has demolished many of the fables buttressing the welfare status quo. Among the venerable myths debunked by Wisconsin are the following: Recipients really want to work but jobs are not available; the lack of day care makes employment impossible; education and training are the key to reducing dependence; and it costs more to reform than to continue the status quo. Thompson has not only rewritten the rule book on fighting dependence; he has invented a new language in which future rules will be written.

Perhaps the most surprising aspect of Wisconsin's story the extraordinary outcomes produced by mundane policies. There is nothing radical about initiatives like Self-Sufficiency First or Pay for Performance. Indeed, these policies are pretty much what most voters have in mind when they hear talk of making welfare recipients work. It is true that in the fall of 1997, Thompson will inaugurate a new set for reforms termed Wisconsin Works, or W2. This will abolish AFDC entirely and replace it with a pure employment-based system of assistance. Although great things are expected of W2, its arrival should not overshadow the fact that the current reforms will have already eliminated a vast portion of the Wisconsin AFDC caseload before W2 even begins.

Other states that are in early stages of reform need not leap as far as W2. Any state that will enact Pay for Performance and Self-Sufficiency First, and follow the nine principles outlined above, will dramatically reduce the welfare dependency of its citizens. Wisconsin has shown the way; it is now up to rest of the nation to apply the lessons learned.

Robert Rector is a research fellow at The Heritage Foundation.