Welfare Reform as I Knew It: When Bad Things Happen to Good Policies

Senator Daniel Patrick Moynihan, a friend of some years, was the first to sound the warning. When I met with him shortly after arriving in Washington in February of 1993, he said, "So you've come to do welfare reform. . . . I'll look forward to reading your book about why it failed this time." Well, Senator, consider this the first installment.

In May 1994, a Times-Mirror poll asked the following question:

One proposal currently being discussed to reform welfare would require all able-bodied welfare recipients, including women with pre-school children, to go to school for two years to learn a skill while receiving benefits. After that, they would be required to either get a job or take a job the government would give them and their welfare benefits would be discontinued. Child care would be provided for the children of working mothers. Do you favor or oppose this proposal?

Ninety-one percent said they favored such an approach. The Times-Mirror question captured very closely the basics of the Clinton welfare reform plan. Yet in spite of such apparent overwhelming public support, nothing like this plan will become law before the 1996 election. Indeed, it may never happen at the federal level, which raises an obvious question: Why did the Clinton welfare reform plan fail? And what does this failure tell us about the Clinton presidency, congressional politics, and democratic institutions more broadly? As an assistant secretary of Health and Human Services (HHS), I was closely involved in the development of welfare reform. Along with Mary Jo Bane, also an assistant secretary, and Bruce Reed, a White House adviser, I cochaired President Clinton's working group on welfare reform and was one of the people who drafted the plan. I had devoted my life to studying poverty and welfare, and was flattered during the 1992 campaign when Bill Clinton had cited my work as an influence on his ideas. When the offer came for me to go to Washington and help to craft real welfare reform, I was thrilled.


Those of us who put together the policy started with a particular analysis of poverty and welfare. First, low-income working families get a particularly bad deal today. Many workers, especially those with less than a college degree, have seen their wages drop, and they often receive little or no health coverage and little help with child care or other costs. The deteriorating living standards of working families pose an enormous challenge if we genuinely want people to move off welfare and be able to live decently.

Second, the welfare system must be transformed. Everyone seems to agree that welfare should be "a hand up, not a handout." Yet welfare administration is mostly about eligibility and benefit determination—check writing. The welfare system sends the message in a myriad of ways that traditional employment is foolish. If we are serious about work and opportunity, we need to change the whole culture of welfare offices. From the moment someone walks through the door, every signal ought to be that work is the ultimate goal and expectation.

Next, even when parents live apart, both of them ought to have the responsibility and opportunity to nurture and provide for their children. Based on surveys of absent fathers, Elaine Sorenson of the Urban Institute estimates that more than $48 billion for children could be generated every year by a system of child support that found every absent parent and collected money according to the simple formula now in use in Wisconsin. Current collections are just $14 billion.

Finally, we need to reduce the large number of children born to unwed mothers, especially teen mothers. Child poverty will always be a problem if we fail to signal to prospective parents that they should not have children until they are prepared to nurture and provide for them.

By confronting these problems and designing a system oriented toward work and responsibility, we believed we could dramatically improve the well-being of children. The four key elements of the Clinton strategy reflected this analysis.


Make work pay. The strategy sought to support the efforts of all working families by dramatically expanding the earned income tax credit (EITC), ensuring universal health care coverage, and helping with child care costs of working families. The higher EITC and health coverage would amount to a substantial pay raise for millions of low-wage workers.

Two years and you work. The Clinton policy sought to transform the welfare system from one focused on eligibility and check writing to one designed to move people quickly to work. From their first contact with welfare, people would be expected to seek work or train for it. After two years, most healthy adults would be required to work, preferably in a regular private job, but if necessary in a subsidized private, nonprofit, or public-sector job.

Child support enforcement. The plan sought to improve child support enforcement through a series of measures: promoting the establishment of paternity at birth in the hospital, comparing reports on new hires with a new national registry of people ordered to pay child support, automatic wage withholding, improved interstate enforcement procedures, and even work requirements for absent parents who refuse to pay. The bill also included measures to improve opportunities for noncustodial parents to do more to nurture their children.

Fight teen pregnancy. The program offered grants to up to 1,000 high-risk schools that proposed innovative teen pregnancy initiatives, a nationwide clearinghouse of information on teen pregnancy, and a few intensive demonstration projects. The policy would also hold parents accountable for their children: Teen parents who were living with their child would be expected to stay in school, stay at home (or if the home was unsafe, in some other supervised setting), and once they completed school, go to work to support their child. Absent teen fathers would be held accountable by the child support system.


More has been accomplished in this agenda than many people realize. In particular, a large expansion of the EITC passed with Clinton's first budget, so that working families need not be poor. The story on child support enforcement is also positive. Virtually all of the important child support enforcement measures included in the administration's welfare plan are also included in the welfare proposals before Congress. If the EITC expansion survives and the child support measures pass, children will be much better off. And many states are seeking to adopt work-oriented welfare reforms by applying for waivers, a legal provision that allows the secretary of HHS to waive certain federal rules for states demonstrating alternative policies.

But much has not been accomplished. Health reform fell to defeat, and today there is no chance that this Congress will pass anything like the "two years and you work" elements of our plan. So what happened to the parts of the plan most directly related to changing the welfare system?

Subscribe to The American Prospect


To many people, the problem with "two years and you work" was what to do with people who reached two years without finding a job. There were good reasons to be concerned: More than two-thirds of recipients on the rolls at any one time have already been on welfare for more than two years (cumulated over one or multiple periods of eligibility). Some of these people are disabled; others are already working at least part time. Many would find jobs if they had the right set of supports and received a clear message that they had to work, but the government would still need to create a significant number of jobs.

Putting welfare recipients to work in subsidized jobs naturally raises anxieties among government workers that they might be displaced. In our first meeting, the president of the American Federation of State, County, and Municipal Employees, Jerry McEntee, said something like, "Near as I can tell you want to put more welfare recipients to work in public-service jobs than I have members." He found that "mildly" threatening.

Moving welfare recipients in large numbers from welfare to work had never been tried, nor had large-scale creation of subsidized jobs for welfare recipients. The closest cousin was the Carter administration's program of public-service jobs under the Comprehensive Employment and Training Act (CETA), which continues to have a bad odor in both Democratic and Republican circles.

Compounding these problems was the related issue of cost. Providing child care, training, and work is more expensive in the short run than simply writing checks—unless people move off welfare quickly. Senator Moynihan was emphatic that finding the money we needed was next to impossible.

Fortunately, poll data suggested a "two years and you work" plan would be popular with the public even if it required government jobs and additional money. For example, a November 1993 poll for U.S. News asked Americans whether they would favor a plan to "require job training for those on welfare and after 2 years require them to work." Ninety-three percent said yes. Even in a modified question that asked about requiring welfare recipients to work in "government jobs" if necessary, 82 percent favored the plan. An overwhelming majority expected reform to cost money in the short run, and they favored it nonetheless.

Still, poll data often can mislead, and the jobs and cost issue needed to be confronted. The Clinton welfare reform plan focused on ways to place people first in unsubsidized jobs, then in subsidized private and nonprofit jobs rather than in government or workfare jobs. It offered strong protections against displacement to public employee unions. And it started with only a third of welfare cases and gradually phased in the rest. Administrators and policymakers would thus get time to learn what worked and what didn't, and it would keep the cost down.

The slow phase-in did carry some political costs. Some critics cited it as evidence that we were not serious about work or time limits—a particularly galling criticism because in my view, the best evidence of our seriousness about reform is that we phased it in at a doable rate, rather than promising we could simply create a work-based system overnight. Starting with a small group of recipients and gradually expanding has been the hallmark of Republican governors who are credited with major welfare reforms.

Initially, the signs were positive in Congress. The bill introduced by House Republicans in the fall of 1993—six months before ours arrived—was unambiguously a "two years and you work" measure. Virtually every Republican was a cosponsor. Ironically, this initial Republican proposal would have created far more public jobs than we did and used the private sector far less. Moreover, the bill would have spent more on job creation and child care. Conservative Democrats developed their own plan, similar to Clinton's, but with a faster phase-in and more money for work slots and child care.

Still, no one wanted to raise taxes to finance welfare reform. Thus, all the bills made cuts in other programs to pay for the expansions in work and child care. Finding those cuts proved one of our most difficult tasks. Indeed, much of the struggle we faced within the administration and with Congress was the result of opposition not to the welfare reforms provisions, but to the other cuts used to finance welfare reform. Eventually, though, we found a financing package that most people could accept.

This history is rife with ironies. Since the Clinton bill was introduced in the summer of 1994, evaluations of state experiments have shown that work-oriented reforms in states with high benefits can push enough people to leave welfare quickly, or not apply at all, that welfare savings exceed job training and child care costs. Moreover, the kinds of cuts in social programs now being proposed even by Democrats are vastly greater than the modest reductions we proposed to finance welfare reform.

Another criticism of the Clinton bill comes from conservatives who argue that its biggest weakness was its failure to address out-of-wedlock childbearing. In their view, the welfare system has caused the dramatic changes in family structure of the past 30 years, and thus welfare reform ought to be about "illegitimacy," not work. It wasn't only the Clinton plan that upset them; these ultra conservatives also roundly attacked the 1993 House Republican bill for failing to attempt to reduce out-of-wedlock childbearing through massive cuts in benefits, especially for families with "illegitimate" children.

Long before almost anyone else talked about these issues, Senator Moynihan had warned about family changes. He now points out that by his projections, half of all children will be born out of wedlock by early in the next century. Yet he is especially critical of plans to cut benefits to "save" our children. No credible studies suggest that anything more than a tiny part of the changes in family structure can be traced to welfare benefits. Recently, New Jersey's denial of additional benefits to mothers on welfare who have additional children seems to have had no impact on out-of-wedlock births, according to data from a carefully controlled experiment.

The truth is that no available policies promise more than a very modest reduction in out-of-wedlock births and family breakups. That is why the Clinton reform plan focused on the area where the problem seemed most manageable—teen pregnancy—and included money for new demonstration projects and information sharing.

A final criticism that has emerged since the 1994 election is that federal leadership in welfare reform is a mistake. Most recent innovation has come from the states and, in principle, the states can design programs that make the most sense for their economy and population. Still, history is filled with examples of states choosing to ignore poor families or ignoring racial minorities, regions, or types of families. Moreover, if one state's rules differ markedly from those of another, there will be an incentive for migration. It is a lot easier to move poor people from welfare to the state border than from welfare to work. Needs and resources also differ widely across states. The states with the smallest tax base are usually the states with the greatest proportion of poor children and families. Fearful of becoming "welfare magnets," some states may cut benefits and impose more punitive measures than they would otherwise prefer.

Thus, we went with national rules regarding time limits and work. Beyond that states would have enormous flexibility for innovation. They could design virtually any welfare-to-work plan and subsidized work program that they wanted. But the basic "two years and you work" architecture would be national, and participation would be tracked nationally.

Looking back, I remain convinced the Clinton welfare reform plan made sense both practically and politically. The basic approach enjoyed strong public support, and while the dangers in steering legislation through Congress were real, they were not insurmountable. I am far less convinced, however, that the administration made the right strategic choices.


Conventional wisdom in Washington these days is summarized aptly by the title of a cover story in the New Republic shortly after the 1994 elections. "They Blew It," the article proclaimed, arguing that the administration basically got the policy right, but that it introduced welfare reform far too late (in particular, after health reform) and with far too little focused effort.

There was a moment early in the administration that might be said to have determined the fate of welfare reform. Until just days before it was sent to the printer, the first Clinton budget included several billion dollars for its future welfare reform proposal. Removing that money had almost no noticeable effect on the $1.4 trillion budget, but it greatly affected the timing and thus the fate of welfare reform. If that line had remained in the budget, the administration would have been forced to submit its welfare reform proposal by late spring 1993 rather than the summer of 1994. With the money taken care of, we could have avoided a search for dollars that ultimately consumed much of our energy.

The last-minute budgetary shift proved to be symptomatic of a deeper problem. Only a tiny handful of people in the White House really cared deeply about the issue. The president clearly did, along with a few others. A small group inside the White House and at HHS were left to develop the plan largely on our own. This seemed a blessing—until we needed to find money for welfare reform, or to get on the crowded legislative agenda, or to gain a presidential public appearance.

The decision to move health reform first was plausible at the time but in the end proved to be a mistake. The submission of the welfare reform legislation, now called the Work and Responsibility Act (WRA), was delayed until the summer of 1994 and then got relatively little fanfare. As health reform faltered and the crime bill struggled in Congress in August and September 1994, the administration did relatively little to push for passage of welfare reform before the elections. I doubt most Americans are even aware that we introduced a welfare reform bill.

Another strategic failing was rhetorical. Governing requires a powerful political message as well as good policy. Consider the phrase, "If you work, you shouldn't be poor," which Clinton used during the campaign. That simple but powerful concept compelled action when he became president. When I first arrived in Washington, advocates pointed out that in spite of the president's promise, the earned income tax credit that was about to be introduced in the budget was too low to raise the working poor out of poverty. As a result, we added more than $1 billion to the EITC in an afternoon. It was the last easy billion I found in Washington.

The president's famous promise to "end welfare as we know it" was the most potent sound bite on welfare. It came up so often that we referred to it as EWAKI. Yet while implying that welfare is a massive failure and conveying seriousness of purpose about reform, EWAKI only vaguely suggests that we can replace the current system with something better. Even more destructive was the phrase "two years and you're off." Our pollsters told us that "two" was the single most memorable number of the 1992 campaign. The problem, of course, is that "two years and you're off" seems to imply no help at all after two years. That is never what was intended. Nonetheless, this phrase gave real impetus to plans now before Congress and in the states that call for time limits followed by nothing—no welfare, no jobs, no support—even if the person is willing to work and genuinely cannot find any job. In my view, these measures are appalling.

A much better phrase, which more accurately conveys the Clinton welfare plan, was "two years and you work." Its virtue is that it conveys a promise (employment) as well as a threat. Polls and focus groups suggest "two years and you work" would have been just as popular as "two years and you're off," perhaps even more so. In the U.S. News poll cited earlier, when people were asked whether they favored a plan to limit welfare benefits to two years, not allowing beneficiaries back on welfare ever, only 22 percent were in favor.

Still, to attribute the collapse of positive welfare reform to the president's rhetoric would miss the larger reality of public opinion. Republican Governors Tommy Thomson and John Engler got enormous political mileage out of welfare reform in two relatively liberal states, Wisconsin and Michigan, long before Clinton ran for president.


Although the WRA came to Congress very late, there was a surprising amount of support for it when people finally saw the details. The Democratic caucuses in both chambers were far more supportive than many in the administration had anticipated, and even Republicans started out on a positive note. Senator Moynihan, who chaired the Senate Finance Committee, and Representative Sam Gibbons, who chaired the House Ways and Means Committee, were both supportive, and Gibbons started pushing hard for immediate action.

But the enthusiasm was not uniform. Several members of the House Ways and Means Subcommittee on Human Resources, where the legislation would originate, could barely disguise their frustration. We had satisfied organized labor; most liberal advocacy groups were not adamantly opposed to our bill, and some supported it. Nonetheless, these congressmen were convinced that any action on welfare reform would inevitably lead to a disastrously punitive bill, especially in an election year. Some suspected that the Clinton administration wanted welfare reform for the wrong reasons—that it was "boob bait for Bubbas," a fear Senator Moynihan had expressed months before the administration produced a bill. And though we had met with them repeatedly, they felt that in drafting the legislation, we did not pay attention to their concerns. Key subcommittee members argued vehemently that it made more sense to consider welfare reform in 1995, when election pressures would be reduced. They felt we had too little time left in 1994 to seriously consider the bill. And the administration chose not to push hard for reform in a summer when the crime and health bills were at the top of the agenda.

To many of us in the administration, some of the Democrats seemed unwilling to make serious changes in welfare and were out of touch with the public. They were extremely skeptical of time limits, even if followed by work. One member repeatedly asked me about how we would guarantee protection of people who refused to work. Meanwhile, I was fighting within the administration to assure work slots for people who were willing to work but unable to find any job. But whatever the reasons, action was put off in 1994, with the expectation that we would resume after the 1994 elections.

In spite of the strategic and policy problems we faced, my view going into the November elections was that we were eventually going to get good legislation. I expected the bill to move left in subcommittee and then right on the floor of the House. In the Senate, things seemed to be in reasonably good shape, though many battles remained to be fought.

Then came November 1994.


After the election, welfare reform went on a remarkable political journey. It's not just that Republican proposals replaced Democratic bills on the congressional agenda. Few people realize how radically the Republican welfare plans have changed. The bill that emerged from Congress late last year bore only a passing resemblance to the original Contract with America. Far from being a coherent expression of a more conservative philosophy, the legislation was an uneasy compromise between competing positions advanced by various factions in the Republican Party.

First came the work-oriented reformers. They believe people should be expected to work or in some cases to train for jobs, even if that requires more money. Next, the social policy critics argued that misguided government support is the root of social evil. For them, less is better, and all aid should come with the strictest possible rules regarding behavior from work to school attendance. The devolvers emerged as a third group. They have been led by Republican governors who want Washington to limit its role to providing resources and to defer to the states on the substance of welfare policy. Finally, there were the budget cutters, who do not really care much about welfare policy one way or another. Their top priority is to cut the budget, and they will go as far to cut social programs as the political realities allow.

These positions are incompatible, but that has not prevented Republicans from embracing them. Most Republican members of Congress, like most Democrats, have little knowledge about welfare and much dislike of it. Thus, no gravitational force of shared conviction prevents the policy from oscillating from one position to another as influence shifts among coteries of activists. Work-oriented reformers created the 1993 House Republican bill—and all of the Republicans signed on. In the Contract with America, the ideological policy critics withdrew money for jobs and child care and added strict work rules, cold-turkey time limits, and harsh sanctions for unmarried parents and their children. Again, virtually all House Republicans initially signed on.

Then Republican governors rebelled at such "conservative micro management," complaining that they were left with a nightmare: less money and less flexibility to make their own policies. Under their influence, the chairman's mark—the bill considered and marked up in the key House subcommittee—was almost a pure block grant, with feeble and meaningless work requirements. Democrats pounced on the bill as "weak on work." As a result, the work requirements were strengthened to the point of being almost unachievable. Yet no more money was added to enable states to carry them out; indeed, the dollars kept shrinking. The final House bill actually cut further than the Contract with America did. The Senate softened things up somewhat. It added money for child care, required states to maintain most of their current spending, eliminated some of the worst restrictions such as those on benefits for legal immigrants, and adopted more practical work requirements. But in adopting the block grant approach, the Senate also eliminated the national entitlement for cash assistance, severely reduced federal spending, and failed to make the work requirements feasible.

Though I was long since out of office, I was deeply disappointed when the president implicitly endorsed the Senate measure. It was not a good bill and would have significantly increased child poverty. In January, fortunately, the president vetoed the bill that emerged from House and Senate conference.

As I write, there is yet another flurry of activity surrounding welfare. Given the pending bills, my hope is that nothing passes in 1996. If what emerges is close to the previous conference bill, I fear for our children. Few people realize just how small the block grants are and how much they vary by state. Arkansas will have less than $600 per poor child per year in federal dollars for cash support, work and training, and child care! That is less than $12 per week. But somehow, the state is going to place tens of thousands of mothers in jobs. In contrast, many of the wealthier northeastern states will get more than $2,000 per poor child per year. That still amounts to just $40 per week per poor child. Simultaneously, states will have to cope with dramatic cuts in support for disabled children, immigrants, and Medicaid, not to mention the impact of any recession.

States cannot and will not do the impossible, but they will do the possible. The possible is to cut people off, to offer less service, and to provide less child care for the working poor not on welfare. Because the block grant will reduce federally required state spending and eliminate federal laws regarding eligibility, some states will find it much easier to cut people off than to move them to work. And so the race to the bottom will begin. Even governors and legislators who want to focus on work-based reform may find it too costly if nearby states threaten to dump their poor by simply cutting benefits.

And what of the federal commitments? Who will defend cuts in the welfare block grant versus reductions in Medicare or farm programs or tax cuts? Will a block grant long endure with a funding formula yielding payments per poor child as wildly divergent across states as this one? Many of those on the right privately admit the real goal is to end federal spending on welfare entirely and that this is the first step on that slippery slope. That is certainly where I would predict we will end up.


As I reflect on the experience, I continue to believe that we came very close to success in spite of our many mistakes. If it had not been for the 1994 elections, we could have had thoughtful and progressive reform legislation. So maybe the outcome is simply idiosyncratic—bad timing. We got hit by a freight train, in part, of course, because our own train moved too sluggishly. And yet the failure to achieve meaningful reform poses a larger challenge. Is there any way to avoid Senator Moynihan's implicit claim that welfare reformers are doomed to write their own obituaries?

While many people complain about welfare, most are poorly informed and uninterested in learning or hearing much more. The press will do little to illuminate the real policy alternatives. There will always be an inside-the-Beltway dialogue carried out in the press with the premier reporters. But few papers will cover the issue, and few people will read what they publish. Unless the issue is number one or two on the national agenda, the public will remain skeptical and alienated. Worse yet, the issues of race and class lie just below the surface, occasionally producing ugly stereotypes, often clouding the political dialogue. Nowhere in domestic policy is the us-versus-them mentality worse. One can rail against the darker and faceless underclass, assigning blame and denying responsibility.

Many members of Congress are also poorly informed. The few who do care deeply enough to follow it day to day are more knowledgeable, but they also tend to have strong ideological views and gravitate toward a few choke positions on committees, where they can wield considerable power. That's why dramatic change is so difficult when power has been stably distributed for a long time. And that is why the policy changes can be so breathtakingly large when a new group comes into power.

After such a litany, one might be tempted to give up. And yet there remains a fundamental reality: Americans are afraid for their future and genuinely do want to help those who would help themselves. I believe the anger and the ignorance are not born primarily of selfishness or bigotry. The harshest critics of welfare are the recipients themselves. Rather, it comes from a sense that core American values are being undermined, not reinforced, by the welfare system.

So the fight for reform will go on. But to those who fight the next battle, let me offer a few words of advice. First, you need to make a fundamental strategy call: Do you go for a radical change with high political visibility or more modest changes and lesser attention? Dramatic changes can occur in only two ways with a hot-button issue like welfare: a revolution where a new party seizes control or a high-profile effort that captures the imagination of the public. The latter probably demands intensive executive leadership or a major social movement.

The alternative is quieter reform that requires settling for less dramatic changes and working in a bipartisan way. Bipartisanship is much harder to achieve on highly charged issues. But if the spotlight is elsewhere, quiet diplomacy can do a great deal. We are close to achieving a remarkable success on child support enforcement that is the result of intensive bipartisan effort. It has gotten almost no attention because it creates little controversy. Quiet diplomacy, however, has a cost in boldness. In child support enforcement, we lost the most innovative and exciting new idea: demonstrations of an insured child support system that would have guaranteed custodial parents at least some child support money each month. The Family Support Act of 1988 falls somewhere in between these two strategies, but it too was a bipartisan effort that gathered relatively limited public attention.

Articulating your core values is critical. Social policy directly and indirectly sends some of society's most powerful messages about what we respect or condemn. In deciding on those values, you have to listen closely to what the public believes and expects. In my experience, even those who don't agree will hear you out if you are honest about your values.

Next, recognizing that the public will hear and comprehend only a limited set of messages, keep your ideas and message simple and clear. Political language often obscures more than it clarifies; rarely does it have real policy content. I'll take "make work pay" over "end welfare as we know it" any day.

Cultivate Congress and the press. It is so easy and tempting to see them as your natural enemy. Members of Congress have strong ideas of their own; the press often seems to want to simplify, exaggerate, and inflame. But they are the fundamental instruments of our democracy, and virtually all genuinely want to serve the public interest and believe they are doing so.

Finally, introduce your bill early and, if you want it to pass, stay out of the way of freight trains.

You may also like