m_topn picture
Atlantic Monthly Sidebar

M A R C H  1 9 9 7

The Worst Thing
Bill Clinton Has Done

A Clinton appointee who resigned in protest over
the new welfare law explains why it is so bad and suggests
how its worst effects could be mitigated

by Peter Edelman

Falling by the wayside I HATE welfare. To be more precise, I hate the welfare system we had until last August, when Bill Clinton signed a historic bill ending "welfare as we know it." It was a system that contributed to chronic dependency among large numbers of people who would be the first to say they would rather have a job than collect a welfare check every month -- and its benefits were never enough to lift people out of poverty. In April of 1967 I helped Robert Kennedy with a speech in which he called the welfare system bankrupt and said it was hated universally, by payers and recipients alike. Criticism of welfare for not helping people to become self-supporting is nothing new.

But the bill that President Clinton signed is not welfare reform. It does not promote work effectively, and it will hurt millions of poor children by the time it is fully implemented. What's more, it bars hundreds of thousands of legal immigrants -- including many who have worked in the United States for decades and paid a considerable amount in Social Security and income taxes -- from receiving disability and old-age assistance and food stamps, and reduces food-stamp assistance for millions of children in working families.

seahorse picture Web-Only: Where Do We Go From Here? We've gathered Peter Edelman and other experts for an Atlantic Unbound roundtable on the future of welfare. Join the debate in The Body Politic.

From the archives:

  • "The Best of Intentions, the Worst of Results" by Irving Kristol (1971). "One wonders how many white middle-class families would survive if mother and children were guaranteed the father's income (or more) without the father's presence? And how many white middle-class fathers would, under these circumstances, persist at their not-always-interesting jobs?"

  • "The Key to Welfare Reform" by David Whitman (1987). Whitman argued that that getting long-term recipients off the rolls is the only way to reduce public-assistance costs
  • When the President was campaigning for re-election last fall, he promised that if re-elected he would undertake to fix the flaws in the bill. We are now far enough into his second term to look at the validity of that promise, by assessing its initial credibility and examining what has happened since.

    I resigned as the assistant secretary for planning and evaluation at the Department of Health and Human Services last September, because of my profound disagreement with the welfare bill. At the time, I confined my public statement to two sentences, saying only that I had worked as hard as I could over the past thirty-plus years to reduce poverty and that in my opinion this bill moved in the opposite direction. My judgment was that it was important to make clear the reasons for my resignation but not helpful to politicize the issue further during an election campaign. And I did want to see President Clinton re-elected. Worse is not better, in my view, and Bob Dole would certainly have been worse on a wide range of issues, especially if coupled with a Republican Congress.

    I feel free to speak out in more detail now, not to tell tales out of school but to clarify some of the history and especially to underscore the damage the bill will do and explain why the bill will be hard to fix in any fundamental way for a long time to come. It is also important to understand what is being done and could be done to minimize the damage in the short run, and what would be required for a real "fix": a strategy to prevent poverty and thus reduce the need for welfare in the first place.

    Four questions are of interest now. Did the President have to sign the bill? How bad is it really, and how can the damage be minimized as the states move to implement it? Can it be fixed in this Congress? What would a real fix be, and what would it take to make that happen?


    WAS the President in a tight political box in late July, when he had to decide whether to sign or veto? At the time, there was polling data in front of him showing that very few people were likely to change their intended vote in either direction if he vetoed the bill. But even if he accurately foresaw a daily pounding from Bob Dole that would ultimately draw political blood, the real point is that the President's quandary was one of his own making. He had put himself there, quite deliberately and by a series of steps that he had taken over a long period of time.

    Governor Clinton campaigned in 1992 on the promise to "end welfare as we know it" and the companion phrase "Two years and you're off." He knew very well that a major piece of welfare-reform legislation, the Family Support Act, had already been passed, in 1988. As governor of Arkansas he had been deeply involved in the enactment of that law, which was based on extensive state experimentation with new welfare-to-work initiatives in the 1980s, especially GAIN in California. The 1988 law represented a major bipartisan compromise. The Democrats had given in on work requirements in return for Republican concessions on significant federal funding for job training, placement activities, and transitional child care and health coverage.

    The Family Support Act had not been fully implemented, partly because not enough time had passed and partly because in the recession of the Bush years the states had been unable to provide the matching funds necessary to draw down their full share of job-related federal money. Candidate Clinton ought responsibly to have said that the Family Support Act was a major piece of legislation that needed more time to be fully implemented before anyone could say whether it was a success or a failure.

    Inadequate safety net Instead Clinton promised to end welfare as we know it and to institute what sounded like a two-year time limit. This was bumper-sticker politics -- oversimplification to win votes. Polls during the campaign showed that it was very popular, and a salient item in garnering votes. Clinton's slogans were also cleverly ambiguous. On the one hand, as President, Clinton could take a relatively liberal path that was nonetheless consistent with his campaign rhetoric. In 1994 he proposed legislation that required everyone to be working by the time he or she had been on the rolls for two years. But it also said, more or less in the fine print, that people who played by the rules and couldn't find work could continue to get benefits within the same federal-state framework that had existed since 1935. The President didn't say so, but he was building -- quite incrementally and on the whole responsibly -- on the framework of the Family Support Act. On the other hand, candidate Clinton had let his listeners infer that he intended radical reform with real fall-off-the-cliff time limits. He never said so explicitly, though, so his liberal flank had nothing definitive to criticize. President Clinton's actual 1994 proposal was based on a responsible interpretation of what candidate Clinton had said.

    Candidate Clinton, however, had let a powerful genie out of the bottle. During his first two years it mattered only insofar as his rhetoric promised far more than his legislative proposal actually offered. When the Republicans gained control of Congress in 1994, the bumper-sticker rhetoric began to matter. So you want time limits? the Republicans said in 1995. Good idea. We'll give you some serious time limits. We now propose an absolute lifetime limit of five years, cumulatively, that a family can be on welfare. End welfare as we know it? You bet. From now on we will have block grants. And what does that mean? First, that there will be no federal definition of who is eligible and therefore no guarantee of assistance to anyone; each state can decide whom to exclude in any way it wants, as long as it doesn't violate the Constitution (not much of a limitation when one reads the Supreme Court decisions on this subject). And second, that each state will get a fixed sum of federal money each year, even if a recession or a local calamity causes a state to run out of federal funds before the end of the year.

    This was a truly radical proposal. For sixty years Aid to Families with Dependent Children had been premised on the idea of entitlement. "Entitlement" has become a dirty word, but it is actually a term of art. It meant two things in the AFDC program: a federally defined guarantee of assistance to families with children who met the statutory definition of need and complied with the other conditions of the law; and a federal guarantee to the states of a matching share of the money needed to help everyone in the state who qualified for help. (AFDC was never a guarantor of income at any particular level. States chose their own benefit levels, and no state's AFDC benefits, even when coupled with food stamps, currently lift families out of poverty.) The block grants will end the entitlement in both respects, and in addition the time limits say that federally supported help will end even if a family has done everything that was asked of it and even if it is still needy.

    In 1995 the President had a new decision to make. What should he say about the Republican proposal? The Republicans started considering the issue in the House in the heady post-election period, when it seemed not at all dissonant for them to talk of reviving orphanages and turning the school-lunch program into block grants. The Administration concentrated its fire on these exponentially extreme measures and said nothing about time limits and the destruction of the entitlement. The President won the public argument about orphanages and school lunches, but his silence on the rest of the bill made it more difficult to oppose the time limits and the ending of the entitlement. For months, while the Republican bill was going through the House and the Senate, the President said nothing further. He might have said, "This isn't what I meant in my campaign rhetoric of 1992." He might have said, "This is totally inconsistent with the bill that I sent up to the Hill last year." He might have sent up a new bill that clearly outlined his position. He might have insisted that the waivers he was giving the states so that they could experiment with reform under the existing law were a strategy superior to the Republican proposals. He did none of these things, despite importuning from Hill Democrats, outside advocates, and people within the Administration.

    The House Democrats had remained remarkably unified in opposition to the House Republicans' bill, which gave new meaning to the word "draconian." But when Democratic senators were deciding how to vote on the more moderate Senate bill, which nonetheless contained the entitlement-ending block grants and the absolute time limit, they looked to the President for a signal. Had he signaled that he remained firm in opposing block grants and the arbitrary time limit, there is every reason to believe that all but a handful of Democratic senators would have stayed with him. The opposite signal left them with no presidential cover for a vote against the Senate bill. It invited them to vote for the bill.

    Prior to the Senate vote on September 19, 1995, the President sent the signal that he could sign the Senate bill (but warned that he would veto a bill that was too much like the House version). The Senate Democrats collapsed and the Senate passed its version of the bill by a vote of 87 to 12. To make matters worse, the President had been presented with an analysis showing that the Senate bill would push more than a million children into poverty. The analysis had been commissioned from the Urban Institute by Secretary of Health and Human Services Donna Shalala's staff (specifically Wendell Primus, the deputy assistant secretary for human-services policy), and Shalala had personally handed it to the President on September 15.


    THIS was the major milestone in the political race to the bottom. The President had said he was willing to sign legislation that would end a sixty-year commitment to provide assistance to all needy families with children who met the federal eligibility requirements. In the floor debate Senator Edward Kennedy, who voted against the bill, described it as "legislative child abuse." In late 1995 and early 1996 the Republicans saved the President from having to make good on his willingness to sign a welfare block-grant bill by sending him versions of the bill that contained horrible provisions concerning food stamps, disabled children, and foster care, which he vetoed. The Republican strategy at the time was to run against the President as a hypocrite who talked welfare reform but wouldn't deliver when he had the chance.

    But President Clinton was not finished. Perhaps he saw some threat to himself in the Republican strategy. Perhaps he did not see the entitlement as being quite so meaningful as others did. It is important to remember that he is not only a former governor but the former governor of Arkansas. AFDC benefits in Arkansas were so low that he might not have seen the entitlement as meaning what it does in higher-benefit states. He might have thought that as governor of Arkansas he would have been able to design a better program if he had received the federal money in the form of a block grant, without the restrictions, limited as they were, that were imposed by the federal AFDC program. And many people have remarked that he seems never to have met a governor he didn't like -- an observation that appeared valid even after the 1994 elections reduced the number of Democrats in the gubernatorial ranks.

    Whatever the reason, when the governors came to town for their winter meetings early last year, the President invited them to draft and submit new proposals on welfare and, for that matter, Medicaid. For a time it seemed to some observers that the President might even be willing to consider block grants for Medicaid, but it quickly became apparent that Medicaid block grants would have negative consequences for a much larger slice of the electorate than would welfare block grants. Large numbers of middle-income people had elderly parents in nursing homes whose bills were paid by Medicaid -- to say nothing of the potential impact on hospitals, physicians, and the nursing homes themselves, all of which groups have substantial political clout. Welfare had no politically powerful constituency that would be hurt by conversion to block grants.

    Hill Republicans, still pursuing the strategy of giving the President only bills that he could not sign, tied the governors' welfare and Medicaid proposals into a single bill. It was clear that the President would veto the combined bill, because by spring he had come out firmly against block grants for Medicaid.

    As of late spring it looked as if a stalemate had been reached, and that 1996 might pass without enactment of a welfare bill. Behind the scenes, however, White House political people -- Rahm Emanuel and Bruce Reed, in particular -- were telling Hill Republicans almost daily that if they separated the welfare and Medicaid bills, they could get a bill that the President would sign. In early summer a new dynamic arose on the Hill. House Republicans, especially freshmen, began to worry that they were vulnerable to defeat on the basis that they had accomplished so little of what they had come to Washington to do. Thinking that Bob Dole was a sure loser anyway, they decided to save their own skins even though it would be to the detriment of the Dole candidacy. The Republicans decided to separate welfare and Medicaid, and began to move a freestanding welfare bill through Congress. The Senate and House bills were each roughly comparable to the respective Senate and House bills passed in 1995, but this time the conference outcome was very different: the conference produced a bill that was fairly close to what the Senate had passed. This time the Hill Republicans wanted the President to sign it.

    The game was over. Now no one could ever say again with any credibility that this President is an old liberal.


    Up in a puff of smoke BEFORE I begin my critique, I need to say something about the motivations of those who genuinely support this new approach. Some of them, anyway, had in my estimation gotten impatient with the chronicity of a significant part of the welfare caseload and the apparent intractability of the problem. I believe they had essentially decided that handing everything over to the states was the only thing left to try that didn't cost a huge amount of money. They may well understand that there will be a certain amount of suffering, and may believe that the bucket of ice-cold water being thrown on poor people now will result in a future generation that will take much more personal responsibility for itself and its children. I think they have made a terrible mistake, as I will try to show, but I respect the frustration that motivated at least some of them. How bad, then, is it? Very bad. The story has never been fully told, because so many of those who would have shouted their opposition from the rooftops if a Republican President had done this were boxed in by their desire to see the President re-elected and in some cases by their own votes for the bill (of which, many in the Senate had been foreordained by the President's squeeze play in September of 1995).

    The same de facto conspiracy of silence has enveloped the issue of whether the bill can be easily fixed. The President got a free ride through the elections on that point because no one on his side, myself included, wanted to call him on it. He even made a campaign issue of it, saying that one reason he should be re-elected was that only he could be trusted to fix the flaws in the legislation. David Broder wrote in The Washington Post in late August that re-electing the President in response to this plea would be like giving Jack the Ripper a scholarship to medical school.

    Why is the new law so bad? To begin with, it turned out that after all the noise and heat over the past two years about balancing the budget, the only deep, multi-year budget cuts actually enacted were those in this bill, affecting low-income people.

    The magnitude of the impact is stunning. Its dimensions were estimated by the Urban Institute, using the same model that produced the Department of Health and Human Services study a year earlier. To ensure credibility for the study, its authors made optimistic assumptions: two thirds of long-term recipients would find jobs, and all states would maintain their current levels of financial support for the benefit structure. Nonetheless, the study showed, the bill would move 2.6 million people, including 1.1 million children, into poverty. It also predicted some powerful effects not contained in the previous year's analysis, which had been constrained in what it could cover because it had been sponsored by the Administration. The new study showed that a total of 11 million families -- 10 percent of all American families -- would lose income under the bill. This included more than eight million families with children, many of them working families affected by the food-stamp cuts, which would lose an average of about $1,300 per family. Many working families with income a little above what we call the poverty line (right now $12,158 for a family of three) would lose income without being made officially poor, and many families already poor would be made poorer.

    The view expressed by the White House and by Hill Democrats, who wanted to put their votes for the bill in the best light, was that the parts of the bill affecting immigrants and food stamps were awful (and would be re-addressed in the future) but that the welfare-reform part of the bill was basically all right. The immigrant and food-stamp parts of the bill are awful, but so is the welfare part.

    The immigrant provisions are strong stuff. Most legal immigrants currently in the country and nearly all future legal immigrants are to be denied Supplemental Security Income and food stamps. States have the option of denying them Medicaid and welfare as well. New immigrants will be excluded from most federal means-tested programs, including Medicaid, for the first five years they are in the country. All of this will save about $22 billion over the next six years -- about 40 percent of the savings in the bill. The SSI cuts are the worst. Almost 800,000 legal immigrants receive SSI, and most of these will be cut off. Many elderly and disabled noncitizens who have been in the United States for a long time and lack the mental capacity to do what is necessary to become citizens will be thrown out of their homes or out of nursing homes or other group residential settings that are no longer reimbursed for their care.

    The food-stamp cuts are very troubling too. Exclusive of the food-stamp cuts for immigrants, they involve savings of about $24 billion. Almost half of that is in across-the-board cuts in the way benefits are calculated. About two thirds of the benefit reductions will be borne by families with children, many of them working families (thus reflecting a policy outcome wildly inconsistent with the stated purposes of the overall bill). Perhaps the most troubling cut is the one limiting food stamps to three months out of every three years for unemployed adults under age fifty who are not raising children. The Center on Budget and Policy Priorities describes this as "probably the single harshest provision written into a major safety net program in at least 30 years" -- although it turns out that more states than the drafters anticipated can ask for an exception that was written to accommodate places with disproportionate unemployment. One of the great strengths of food stamps until now has been that it was the one major program for the poor in which help was based only on need, with no reference to family status or age. It was the safety net under the safety net. That principle of pure need-based eligibility has now been breached.

    Neither the cuts for immigrants nor the food-stamp cuts have anything to do with welfare reform. Many of them are just mean, with no good policy justification. The bill also contains other budget and benefit reductions unrelated to welfare. The definition of SSI eligibility for disabled children has been narrowed, which will result in removal from the rolls of 100,000 to 200,000 of the 965,000 children who currently receive SSI. Although there was broad agreement that some tightening in eligibility was warranted, the changes actually made will result in the loss of coverage for some children who if they were adults would be considered disabled. Particularly affected are children with multiple impairments no one of which is severe enough to meet the new, more stringent criteria. Child-nutrition programs have also been cut, by nearly $3 billion over six years, affecting meals for children in family day care and in the summer food program. Federal funding for social services has been cut by a six-year total of $2.5 billion. This is a 15 percent cut in an important area, and will hamper the states in providing exactly the kind of counseling and support that families often need if a parent is going to succeed in the workplace.

    So this is hardly just a welfare bill. In fact, most of its budget reductions come in programs for the poor other than welfare, and many of them affect working families. Many of them are just cuts, not reform. (The bill also contains an elaborate reform of federal child-support laws, which had broad bipartisan support and could easily have been enacted as separate legislation.)

    THIS brings us to welfare itself. Basically, the block grants mean that the states can now do almost anything they want -- even provide no cash benefits at all. There is no requirement in the new law that the assistance provided to needy families be in the form of cash. States may contract out any or all of what they do to charitable, religious, or private organizations, and provide certificates or vouchers to recipients of assistance which can be redeemed with a contract organization. So the whole system could be run by a corporation or a religious organization if a state so chooses (although the latter could raise constitutional questions, depending on how the arrangement is configured). Or a state could delegate everything to the counties, since the law explicitly says that the program need not be run "in a uniform manner" throughout a state, and the counties could have varying benefit and program frameworks. For good or for ill, the states are in the process of working their way through an enormous -- indeed, a bewildering -- array of choices, which many of them are ill equipped to make, and which outside advocates are working hard to help them make well.

    The change in the structure is total. Previously there was a national definition of eligibility. With some limitations regarding two-parent families, any needy family with children could get help. There were rules about participation in work and training, but anybody who played by the rules could continue to get assistance. If people were thrown off the rolls without justification, they could get a hearing to set things right, and could go to court if necessary. The system will no longer work that way.

    The other major structural change is that federal money is now capped. The block grants total $16.4 billion annually for the country, with no new funding for jobs and training and placement efforts, which are in fact very expensive activities to carry out. For the first couple of years most of the states will get a little more money than they have been getting, because the formula gives them what they were spending a couple of years ago, and welfare rolls have actually decreased somewhat almost everywhere (a fact frequently touted by the President, although one might wonder why the new law was so urgently needed if the rolls had gone down by more than two million people without it).

    Many governors are currently crowing about this "windfall" of new federal money. But what they are not telling their voters is that the federal funding will stay the same for the next six years, with no adjustment for inflation or population growth, so by 2002 states will have considerably less federal money to spend than they would have had under AFDC. The states will soon have to choose between benefits and job-related activities, with the very real possibility that they will run out of federal money before the end of a given year. A small contingency fund exists for recessions, and an even smaller fund to compensate for disproportionate population increases, but it is easy to foresee a time when states will have to either tell applicants to wait for the next fiscal year or spend their own money to keep benefits flowing.

    The bill closes its eyes to all the facts and complexities of the real world and essentially says to recipients, Find a job. That has a nice bumper-sticker ring to it. But as a one-size-fits-all recipe it is totally unrealistic.

    Total cutoffs of help will be felt right away only by immigrants and disabled children -- not insignificant exceptions. The big hit, which could be very big, will come when the time limits go into effect -- in five years, or less if the state so chooses -- or when a recession hits. State treasuries are relatively flush at the moment, with the nation in the midst of a modest boom period. When the time limits first take effect, a large group of people in each state will fall into the abyss all at once. Otherwise the effects will be fairly gradual. Calcutta will not break out instantly on American streets.

    To the extent that there are any constraints on the states in the new law, they are negative. The two largest -- and they are very large -- are the time limit and the work-participation requirements.

    There is a cumulative lifetime limit of five years on benefits paid for with federal money, and states are free to impose shorter time limits if they like. One exception is permitted, to be applied at the state's discretion: as much as 20 percent of the caseload at any particular time may be people who have already received assistance for five years. This sounds promising until one understands that about half the current caseload is composed of people who have been on the rolls longer than five years. A recent study sponsored by the Kaiser Foundation found that 30 percent of the caseload is composed of women who are caring for disabled children or are disabled themselves. The time limits will be especially tough in states that have large areas in chronic recession -- for example, the coal-mining areas of Appalachia. And they will be even tougher when the country as a whole sinks into recession. It will make no difference if a recipient has played by all the rules and sought work faithfully, as required. When the limit is reached and the state is unable or unwilling to grant an exception, welfare will be over for that family forever.

    Under the work-participation requirements, 25 percent of the caseload must be working or in training this year, and 50 percent by 2002. For two-parent families 75 percent of the caseload must be working or in training, and the number goes up to 90 percent in two years. The Congressional Budget Office estimates that the bill falls $12 billion short of providing enough funding over the next six years for the states to meet the work requirements. Even the highly advertised increased child-care funding falls more than $1 billion short of providing enough funding for all who would have to work in order for the work requirements to be satisfied. States that fail to meet the work requirements lose increasing percentages of their block grants.

    The states are given a rather Machiavellian out. The law in effect assumes that any reduction in the rolls reflects people who have gone to work. So states have a de facto incentive to get people off the rolls in any way they can, not necessarily by getting them into work activities.

    The states can shift a big chunk of their own money out of the program if they want to. There is no matching requirement for the states, only a maintenance-of-effort requirement that each state keep spending at least 80 percent of what it was previously contributing. This will allow as much as $40 billion nationally to be withheld from paying benefits over the next six years, on top of the $55 billion cut by the bill itself. Moreover, the 80 percent requirement is a static number, so the funding base will immediately start being eroded by inflation.

    Besides being able to transfer some of their own money out, the states are allowed to transfer up to 30 percent of their federal block grants to spending on child care or other social services. Among other things, this will encourage them to adopt time limits shorter than five years, because this would save federal money that could then be devoted to child care and other help that families need in order to be able to go to work. Hobson's choice will flourish.

    The contingency fund to cushion against the impact of recessions or local economic crises is wholly inadequate -- $2 billion over five years. Welfare costs rose by $6 billion in three years during the recession of the early nineties.

    The federal AFDC law required the states to make decisions on applications within forty-five days and to pay, retroactively if necessary, from the thirtieth day after the application was put in. There is no such requirement in the new law. All we know from the new law is that the state has to tell the Secretary of Health and Human Services what its "objective criteria" will be for "the delivery of benefits," and how it will accord "fair and equitable treatment" to recipients, including how it will give "adversely affected" recipients an opportunity to be heard. This is weak, to say the least.


    GIVEN this framework, what can we predict will happen? No state will want to be a magnet for people from other states by virtue of a relatively generous benefit structure. This is common sense, unfortunately. As states seek to ensure that they are not more generous than their neighbors, they will try to make their benefit structures less, not more, attractive. If states delegate decisions about benefit levels to their counties, the race to the bottom will develop within states as well.

    I do not wish to imply that all states, or even most states, are going to take the opportunity to engage in punitive policy behavior. There will be a political dynamic in the process whereby each state implements the law. Advocates can organize and express themselves to good effect, and legislatures can frustrate or soften governors' intentions. There is another important ameliorating factor: many welfare administrators are concerned about the dangers that lie in the new law and will seek to implement it as constructively as they can, working to avoid some of the more radical negative possibilities.

    Citizens can make a difference in what happens in their state. They can push to make sure that it doesn't adopt a time limit shorter than five years, doesn't reduce its own investment of funds, doesn't cut benefits, doesn't transfer money out of the block grant, doesn't dismantle procedural protections, and doesn't create bureaucratic hurdles that will discourage recipients. They can press for state and local funds to help legal immigrants who have been cut off from SSI or food stamps and children who have been victimized by the time limits. They can advocate an energetic and realistic jobs and training strategy, with maximum involvement by the private sector. And they can begin organizing and putting together the elements of a real fix, which I will lay out shortly.


    EVEN given effective advocacy, relatively responsive legislatures and welfare administrators, and serious efforts to find private-sector jobs, the deck is stacked against success, especially in states that have high concentrations of poverty and large welfare caseloads. The basic issue is jobs. There simply are not enough jobs now. Four million adults are receiving Aid to Families with Dependent Children. Half of them are long-term recipients. In city after city around America the number of people who will have to find jobs will quickly dwarf the number of new jobs created in recent years. Many cities have actually lost jobs over the past five to ten years. New York City, for example, has lost 227,000 jobs since 1990, and the New York metropolitan area overall has lost 260,000 over the same period. New York City had more than 300,000 adults in the AFDC caseload in 1995, to say nothing of the adults without dependent children who are receiving general assistance. Statistics aside, all one has to do is go to Chicago, or to Youngstown, Ohio, or to Newark, or peruse William Julius Wilson's powerful new book, When Work Disappears, to get the point. The fact is that there are not enough appropriate private-sector jobs in appropriate locations even now, when unemployment is about as low as it ever gets in this country.

    For some people, staying on welfare was dictated by economics, because it involved a choice between the "poor support" of welfare, to use the Harvard professor David Ellwood's term, and the even worse situation of a low-wage job, with its take-home pay reduced by the out-of-pocket costs of commuting and day care, and the potentially incalculable effects of losing health coverage. With time limits these people will no longer have that choice, unappetizing as it was, and will be forced to take a job that leaves them even deeper in poverty. How many people will be able to get and keep a job, even a lousy job, is impossible to say, but it is far from all of those who have been on welfare for an extended period of time.

    The labor market, even in its current relatively heated state, is not friendly to people with little education and few marketable skills, poor work habits, and various personal and family problems that interfere with regular and punctual attendance. People spend long spells on welfare or are headed in that direction for reasons other than economic choice or, for that matter, laziness. If we are going to put long-term welfare recipients to work -- and we should make every effort to do so -- it will be difficult and it will cost money to train people, to place them, and to provide continuing support so that they can keep a job once they get it. If they are to have child care and health coverage, that will cost still more. Many of the jobs that people will get will not offer health coverage, so transitional Medicaid for a year or two will not suffice. People who have been on welfare for a long time will too often not make it in their first job and will need continuing help toward and into a second job. Both because the private sector may well not produce enough jobs right away and because not all welfare recipients will be ready for immediate placement in a private-sector job, it will be appropriate also to use public jobs or jobs with nonprofit organizations at least as a transition if not as permanent positions. All of this costs real money.

    For a lot of people it will not work at all. Kansas City's experience is sadly instructive here. In the past two years, in a very well-designed and well-implemented effort, a local program was able to put 1,409 out of 15,562 welfare recipients to work. As of last December only 730 were still at work. The efforts of Toby Herr and Project Match in Chicago's Cabrini-Green public-housing project are another case in point. Working individually and intensively with women and supporting them through successive jobs until they found one they were able to keep, Herr had managed to place 54 percent of her clients in year-round jobs at the end of five years. This is a remarkable (and unusual) success rate, but it also shows how unrealistic is a structure that offers only a 20 percent exception to the five-year time limit.

    I want to be very clear: I am not questioning the willingness of long-term welfare recipients to work. Their unemployment is significantly related to their capacity to work, whether for personal or family reasons, far more than to their willingness to work. Many long-term welfare recipients are functionally disabled even if they are not disabled in a legal sense. News coverage of what the new law will mean has been replete with heartbreaking stories of women who desperately want to work but have severe trouble learning how to operate a cash register or can't remember basic things they need to master. A study in the state of Washington shows that 36 percent of the caseload have learning disabilities that have never been remediated. Many others have disabled children or parents for whom they are the primary caretakers. Large numbers are victims of domestic violence and risk physical retaliation if they enter the workplace. These personal and family problems make such people poor candidates for work in the best of circumstances. Arbitrary time limits on their benefits will not make them likelier to gain and hold employment. When unemployment goes back up to six or seven or eight percent nationally, as it will at some point, the idea that the private sector will employ and continue to employ those who are the hardest to employ will be even more fanciful than it is at the current, relatively propitious moment.

    When the time limits take effect, the realities occasioned by the meeting of a bottom-line-based labor market with so many of our society's last hired and first fired will come into focus. Of course, a considerable number will not fall off the cliff. An increased number will have obtained jobs along the way. The time limits will help some people to discipline themselves and ration their years of available assistance. Some will move in with family or friends when their benefits are exhausted. The 20 percent exception will help as well.

    But there will be suffering. Some of the damage will be obvious -- more homelessness, for example, with more demand on already strapped shelters and soup kitchens. The ensuing problems will also appear as increases in the incidence of other problems, directly but perhaps not provably owing to the impact of the welfare bill. There will be more malnutrition and more crime, increased infant mortality, and increased drug and alcohol abuse. There will be increased family violence and abuse against children and women, and a consequent significant spillover of the problem into the already overloaded child-welfare system and battered-women's shelters.


    I AM amazed by the number of people who have bought the line that the bill was some little set of adjustments that could easily be done away with. Congress and the President have dynamited a structure that was in place for six decades. A solid bipartisan majority of Congress and the President himself have a stake in what they have already done. Fundamental change in the bill is therefore not possible this year. So the answer to the question is no, not in any fundamental way. One possible area for adjustment is in the immigrant and food-stamp provisions. These occasioned the most hand-wringing from the President and some of the people who voted for the bill. They could be changed without redoing everything. The President has made some proposals for limited change on these items.

    The bigger question is welfare. If there is going to be a short-term fix of the new law, it will be not in the fundamentals of the new structure but rather in some of the details. It might possibly include the following, although I hasten to say that even this list stretches credulity.

  • Jobs. Congress could make extra funds available to the states for job creation, wage subsidies, training, placement, support and retention services, and so on. The President has proposed a fund of $3 billion over three years for this kind of activity, saying it would result in a million new jobs. As campaign rhetoric, this was pure spin. It amounts to $3,000 per job. There is simply no way in which $3,000 per job will get a million jobs for people who have been on the welfare rolls for extended periods of time. The President has also proposed a modest additional tax credit for hiring welfare recipients. This, too, will have little practical effect.

  • Time limits. The Democrats tried very hard to create a voucher covering basic necessities for children in families that had run up against the time limit. The idea failed by a narrow margin in the Senate, and is worth pursuing. Another item worth advocating would be raising the 20 percent exception to the time limit to 25 or even 30 percent.

  • Work requirements. The states are chafing under the requirements about the percentage of the caseload that has to be participating in work or related activities. It would help a little if people were permitted to receive vocational training for longer than the twelve months the law allows.

  • Limits on state flexibility in the use of funds. The law is excessively flexible on what the states can do with the block-grant funds. A number of possible changes would be helpful: reducing the percentage that can be transferred out of the block; raising the requirement for states' contributions of their own funds; requiring states to comply with the plans they adopt; requiring states to process applications for assistance expeditiously; and clarifying the procedural protections for people denied or cut off from assistance.

  • Data. It is vitally important that adequate data be gathered and reported on what happens under the new legislation. The new law contains some funding for research and some instructions about data to be gathered, but additional funds and specification would be helpful.

    If reliable and affordable health care and child care were added to this list, and were available beyond a transitional period, it would help a lot. However, my crystal ball tells me that whatever is enacted in these areas will be modest at best, and the new structure will remain substantially in place. And of course not even these adjustments would solve the fundamental problems created when the previous structure was dynamited: the disappearance of the national definition of eligibility and of the guarantee that federal funds will be available for all eligible children.


    A REAL fix would involve, first, jobs, jobs, jobs -- preferably and as a first priority in the private sector, but also in the public sector, where there is real work to be done. And then everything that enables people to be productive citizens. Schools that teach every child as well as they teach every other child. Safe neighborhoods. Healthy communities. Continuing health-care and day-care coverage, so that people can not only go to work but also keep on working. Ending the racial and ethnic discrimination that plagues too many young people who try to enter the job market for the first time.

    When we discuss jobs, we need to be talking about opportunities for men and women both. That may seem obvious, but the welfare bill skews our focus. By allocating to long-term welfare recipients such a large share of the limited resources available for jobs and training, we may be draining funds and attention from others who deserve to be a higher priority. Inner-city young men come particularly to mind. We need to be promoting responsible fatherhood, marriage, and two-parent families. If young men cannot find work, they are far less likely to marry. They may have children, but economics and low self-esteem may defeat responsibility. Tough child-support enforcement is part of the solution, but genuine opportunity and clear pathways to opportunity are vital.

    The outside world tends to believe that the inner city is hopeless. (I do not mean to neglect strategies to reduce rural poverty.) That is not the case. In the toughest neighborhoods, with all the dangers and pitfalls of street life, there are young people who beat the odds, stay in school and graduate, and go to college or get a job. These young people have exceptional strength and resiliency. But there are many more who could make it with a little extra support and attention. It is enormously important that we increase the number of young people who make it. We give a lot of lip service to prevention, whether of crime or drug abuse or teen pregnancy. But we will never prevent these negative outcomes as well as we could until we pursue a general strategy of creating opportunity and clear pathways to opportunity -- a positive youth-development strategy.

    Many of the jobs that welfare recipients and other low-income people get do not pay enough to pull them out of poverty. Continuing attention to the minimum wage and the Earned Income Tax Credit will be necessary. States should insist, as the city of Baltimore has, that all their contractors pay all their workers a sufficient wage to keep them out of poverty (or at least approximately enough to keep a family of four out of poverty), and should fund their contracts accordingly. Current child-care and health-care policies are insufficient to allow low-wage workers to stay out of poverty even if transitional subsidies let them escape temporarily when they leave the welfare rolls. Federal and state child-care subsidies should help all workers who would otherwise be poor, not just those who have recently left the welfare rolls. And at the end of the day we still have 40 million Americans, including 10 million children, who do not have health coverage. We still have to deal with that as part of a real antipoverty strategy.

    We have been reduced to the politics of the waitress mom. She says, all too legitimately, "I bust my tail. I don't have decent child care. I don't have health coverage. Why should 'these people' get what I don't have?" We started to bring greater equity to the working poor but, except for the recent minimum-wage increase, progress was halted by the 1994 congressional elections. A real fix would help the waitress mom as well as those a rung below her on the income ladder.

    We are not just talking policy; we are talking values. We are talking people, especially young people growing up, who understand that they have to take responsibility for themselves, both as earners and as parents.

    Personal responsibility and community responsibility need to intersect. The community has a responsibility to help instill and nurture values. The community has a responsibility to offer support, especially to children and youths, so that everyone has an opportunity to acquire the tools necessary to achieve the personal responsibility that is such a vital element in the equation. The community has a responsibility to help parents do their job. And community means something different from programs, something larger, although programs are part of the equation. Liberals have tended to think in terms of programs. The community's taking responsibility is a much larger idea. But communities cannot succeed in isolation. National leadership and policy are essential as well.

    Welfare is what we do when everything else fails. It is what we do for people who can't make it after a genuine attempt has been mounted to help the maximum possible number of people to make it. In fact, much of what we do in the name of welfare is more appropriately a subject for disability policy. The debate over welfare misses the point when all it seeks to do is tinker with welfare eligibility, requirements, and sanctions. The 1996 welfare law misses the point.

    To do what needs to be done is going to take a lot of work -- organizing, engaging in public education, broadening the base of people who believe that real action to reduce poverty and promote self-sufficiency in America is important and possible. We need to watch very carefully, and we need to document and publicize, the impact of the 1996 welfare legislation on children and families across America. We need to do everything we can to influence the choices the states have to make under the new law. We can ultimately come out in a better place. We should not want to go back to what we had. It was not good social policy. We want people to be able to hold up their heads and raise their children in dignity. The best that can be said about this terrible legislation is that perhaps we will learn from it and eventually arrive at a better approach. I am afraid, though, that along the way we will do some serious injury to American children, who should not have had to suffer from our national backlash.

    Illustrations by Robert Goldstrom

    Copyright © 1997 by The Atlantic Monthly Company. All rights reserved.
    The Atlantic Monthly; March 1997; The Worst Thing Bill Clinton Has Done; Volume 279, No. 3; pages 43-58.

  • m_nv_cv picture m_nv_un picture m_nv_am picture m_nv_pr picture m_nv_as picture m_nv_se picture