The Never-Ending Welfare Debate

Is PRWORA Really Revolutionary?

Norman Barry, a contributing editor of Ideas on Liberty, is professor of social and political theory at the University of Buckingham in the UK. He is the author of An Introduction to Modern Political Theory (St. Martin’s Press).

After a long struggle, a “revolutionary” welfare reform bill, the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), finally became law in 1996. It must be some sort of success since former President Clinton has boasted of his solving of a problem that had bedeviled American public life for at least 30 years. But what kind of reform is it and what ideological (theoretical) principles underlie it?

The answer to the last question is crucially important since any future interventions will be governed by some sort of theoretical understanding of the nature and causation of welfare problems and an appreciation of the legitimate role of the federal government and the states in their resolution. Interestingly, the left contributed little to the debate that began in the late 1970s (although William Julius Wilson’s The Truly Disadvantaged might be an exception). While Michael Harrington’s The Other America might have been the unacknowledged inspiration for Lyndon Johnson’s Great Society, it was Charles Murray’s Losing Ground and two books by Lawrence Mead, Beyond Entitlement and The New Politics of Poverty, that provided the theoretical and statistical weaponry for a sustained assault on that program, leading to the 1996 reform.

But formidable though the artillery is, it was not always aimed at the same target, and the solutions suggested were by no means the same. Charles Murray’s critique of welfare (and also the work of Michael Tanner) is in the classical-liberal tradition, which regards any state involvement in welfare as counterproductive (even malign) in creating a whole generation that had never worked or participated fully in civil society, while Mead believes that government has a role in preparing and training persons for life in that society. For him, there can be no spontaneous regeneration of people already debilitated by decades of welfare. In effect, they must be gently coerced into virtue: “The solution to the work problem lies not in freedom but in governance.”1

The first distinction to be made about welfare (and it is rarely made by Americans) is between the institutional and the residual welfare states.2 The former refers to a wide range of collectively provided and universal services (financed by social insurance) covering poor relief, health, pensions, education, and more, while the latter is limited to cash payments exclusively targeted at the poor. America’s welfare system, with the exception of Social Security and Medicare, is residual while Europe’s is mainly institutional.

The problem with the institutional welfare state is its excessive cost and the reduction in liberty it produces, while the residual generates behavioral disorder. In America the problem of welfare was never its cost (the costs of aid to the poor rarely exceeded 2 percent of public spending and only hit 4 percent if Medicaid is included). Old people on zero-priced health care and unfunded old-age pensions don’t destroy their lives on sex, drugs, and rock ‘n’ roll, but in America the problem with primarily cash welfare was that it undermined the work ethic and produced a generation of people who had never been employed, encouraged (if not exactly caused) unmarried motherhood, wrecked families (or made it irrational to start them), and reduced the inner cities to Third World status.

How did this happen? The major cause was the vast expansion of the American welfare state (which began with the New Deal) under Johnson’s Great Society of the 1960s. The problem of the Great Depression was the working poor and the vast increase in the unemployed. The problems then were not behavioral but structural: people were in distress through forces beyond their control (I leave aside the “macroeconomic” causes of the Depression and the inappropriate methods used to deal with it). But the welfare issues of the 1970s had little to do with economics and all to do with behavioral traits induced by social policy. Ironically the single policy held to be most responsible for social decay, Aid to Families with Dependent Children (AFDC), was originally a New Deal program, Aid to Dependent Children (ADC), contained in the Social Security Act (1935).

The differences are instructive. ADC was aimed at unfortunate women (widows or deserted wives) who were in no way welfare parasites, but rather were victims of unpropitious circumstances. The program was actually derived from “widows’ pensions,” which had been introduced by many states early in the twentieth century. It was based on the ideas that mothers should be encouraged to stay at home, bring up children, and relieve the labor market. But AFDC (1961) became something entirely different. It encouraged welfare dependency, poverty, unmarried motherhood, and nonwork; this time round people thought staying at home was a bad thing. Although it certainly was not generous, changes in the law (notably the abolition of the “man in the house rule” by a Supreme Court decision, King v. Smith, 1968) made access to it easy.

Although arithmetically most AFDC recipients were white, it had a disproportionate effect on the black community, where it almost destroyed the family. Between 1962 and 1972 the number of AFDC families tripled from 1 million to 3 million. By 1994, 66 percent of AFDC families were headed by never-married mothers; in 1975 the figure was only 33 percent. In certain inner-city areas the illegitimacy rate for black births was 80 percent (in 1950 the figure for the black community overall was 20 percent). Thirty percent of all welfare recipients go on welfare because of unwed motherhood. This was moral hazard on the grand scale. Of course, there were many other counterproductive schemes produced by the Great Society, but the expanded AFDC was the most notorious.

What was particularly distressing from a conservative and moralistic point of view was that there was little or no aid for the poor of intact families. A welfare culture was being created that discouraged the maintenance of the traditional American values of work and personal responsibility. In fact, the creation of the National Welfare Rights Organization (1970) did precisely the opposite; it sedulously promoted welfarism and sought out welfare claimants.

Optimists versus Pessimists

The critical response to this was divided between classical liberals who thought that the withdrawal of all welfare would spontaneously generate the right values and pessimistic conservatives who saw a moral role for the state. Some libertarians clung to Milton Friedman’s idea of a Negative Income Tax, despite the many flaws in his system,3 and a somewhat diluted version of this, the Family Assistance Plan (FAP), was proposed by President Nixon in 1970, though it was defeated in Congress. Charles Murray actually produced some empirical evidence against it; it seemed to be particularly destructive of stable families and the work ethic, and it breached that reciprocity between donor and recipient that many feel should underlie welfare. But at least FAP would have abolished AFDC as an entitlement and returned welfare to the states.

Mead’s response to classical liberalism is instructive. He did not see the welfare problem as explicable in terms of simple cause and effect: that an increase in welfare automatically brought about more dependency and its complete removal would just as quickly eliminate the problem.4 He showed that although the value of welfare began to decline in the 1970s, the numbers on welfare increased. Certainly the rise in illegitimacy seemed impervious to any laws of causality. But the oft-quoted figure of a 10 percent increase in AFDC correlating with an 8 percent decrease in marriage, and the massive increase in welfare dependency between 1989 and 1993, must have something to do with causality.

Mead argued that the immediate ending of welfare would produce vast numbers of impoverished people driven to crime, drugs, and prostitution; they would pose a threat to social order. He contended that the experience of welfare had produced people who fell outside the normal class structure; they were socially incompetent, lacked any kind of work skills, and were capable only of reproducing themselves, thus generating “cycles of poverty.” Mead has always been a firm supporter of “workfare,” under which welfare is dependent on recipients’ fulfilling work requirements. Classical liberals thought this was just a waste of money.

It is clear that by the end of the 1980s the country was ready for substantial changes. Although surveys reveal that only 10 percent would have abolished welfare entirely, there was a general revulsion against the dependency that the current system was producing (not to mention crime and other social disorder). Most significant was the desire for something like Mead’s work enforcement. The 1988 Family Support Act made a tentative move toward workfare, but it lacked proper enforcement power and was ineffective.

Much more significant were moves at the state level and via the governors’ association to change the system in the localities. By the mid-1990s, 43 states had secured waivers (granted by the Department of Health and Human Services) from AFDC; this enabled them, for example, to compel unwed teenaged mothers to live with their parents and to deny welfare if an unmarried mother had more than one child. The prime mover in all this was Tommy Thompson, governor of Wisconsin: between 1987 and 1995 the state’s AFDC rolls went down by a third. And in the country as a whole, families on welfare dropped by 14 percent. Between 1993 and 1997 caseloads fell by 26 percent. (In some states the drop was even bigger.)

The country was prepared, then, for more substantial reform. Clinton had promised on his election in 1992 to “end welfare as we know it” and had given to Congress in 1994 a rather tame bill (although even this would have put a time limit of two years on the receipt of AFDC). The bill that eventually passed (PRWORA) was signed by Clinton in 1996 after two earlier vetoes. The presidential election was only months away, and he knew that welfare reform was extremely popular and had secured considerable bipartisan support in Congress. But it owed much more to Mead’s thinking on welfare than to classical liberalism.

Although superficially PRWORA is quite radical, in many ways it continues and formalizes at the federal level things that were already happening at state level. It sweeps away the major entitlement program, AFDC, and returns the bulk of welfare decision-making, and some financing, to the states. The states receive block grants, fixed until 2002, and are free to do what they like about welfare within certain important parameters set by Congress. AFDC has been replaced by Temporary Assistance to Needy Families (TANF), but this includes strict work requirements and time limits (two years in the first instance and no more than five years over a lifetime): the aim was to get the rolls reduced by half by this year. There are minor adjustments to things such as Medicaid and food stamps. But it certainly doesn’t abolish welfare (though the statute envisages a reduction in spending of $155 billion over five years), and the aim of running it down over time has yet to be tested.

Has It Worked?

At first glance the new reform looks like a great success. The welfare rolls fell by 53 percent between 1993 and 1997, and 10 percent more people than those on the old AFDC have found unsubsidized jobs in the private sector. It is difficult to say whether the new figures are a result of the economic prosperity of recent years or a product of the rigor of the new welfare regime. Perhaps the new and superficially encouraging employment figures could be seen as an example of the “creaming” of the labor market; that is, already competent people on welfare are being recruited by astute employers to new and better paid (than welfare) positions.

Indeed, that was always the classical-liberal objection to workfare; jobs are taken up by those who would have worked anyway. It might be the case, then, that these workers are quickly attracted to employment, leaving Mead’s “incompetents” to sink even lower. But the sum total of the improvements might be highly vulnerable to a downturn in the economy. Already observers have detected since 1998 a slowdown in the retreat from welfare. Still, some evidence for the success of the new welfare order can be gauged from the fact that during the early Reagan years 20 million new jobs were created, but the number of people on welfare rose by 13 percent. Nothing similar has happened in the last five years.

A scholarly, yet cautious and skeptical, analysis of the new system has been provided by the Cato Institute.5 Lisa E. Oliphant doubts that the major aims of the reform have been achieved. There are still 3 million people on the welfare rolls, and of the people who have come off welfare (because of the time limit) between 30 and 40 percent have not entered regular employment. Just as in the old regime, there is a group of people who live on the margin of civil society, and without the support of AFDC or TANF it is not exactly clear what is happening to them now. One can only assume that they constitute the same underclass as before, surviving in a world of drugs, petty crime, and prostitution.

Also, these people probably find other forms of welfare. It seems that the 1996 reform has not produced a generation of self-sufficient, responsible agents able to take their place in market society; those who get jobs do not get very good ones, and fewer than 25 percent of former recipients stay in their jobs for more than a year. There has been a subsidized work program (worth $9 billion in the first four years of the reform), but it seems to have gone the way of failed initiatives in the past. Furthermore, far too many ex-welfare clients still cannot survive without the crutch of the state.

As Oliphant notes, the people who have left welfare are still living on state (and federal) benefits. The welfare state is not exhausted by AFDC and now TANF; there is still a whole network of state props (often cash). These include Medicaid, Food Stamps, housing subsidies, and child-care subsidies (though Mead wouldn’t care about the cost of this since it does enable women to get back to work) and Supplemental Security Income. A little-publicized subsidy is the Earned Income Tax Credit: it not only accounts for half the income of welfare leavers but also acts as a powerful disincentive to do extra work because it leads to very high marginal tax rates. Oliphant concludes that there has been little progress toward true self-sufficiency since 1996. Indeed, that old bugbear of welfarism, unmarried motherhood, continues in its debilitating way, especially among teenagers.

Classical-liberal critiques of PRWORA remain as potent as when they were applied to the previous welfare system. It is not a root-and-branch reform but deals (admittedly radically) with the after-effects of welfare. It might be that all the states will use their new freedom to initiate truly innovative measures but this is unlikely. What is really required is the prevention of the descent into dependency, and this means making efforts to abolish the whole welfare system.

In a bold move, Michael Tanner suggests that it should be announced that from nine months and one day onward there will be no aid for mothers of newly born babies.6 That’s a decisive way of dealing with moral hazard. It does not at all follow that the eventual elimination of welfare will lead to widespread suffering. America has a long and honorable tradition of charity, which was badly damaged by the rise of compulsory state welfare. It should be restored.

I should like to conclude by referring to a distinction alluded to earlier—that between the institutional and the residual welfare states. Even if the prevailing welfare system is abolished, a serious problem will remain: the elderly. Americans think that Social Security is not part of welfare; it is not seen as a state payment but something Americans have earned through a self-financing social insurance system (the institutional welfare state). But they haven’t (although they could have if their Social Security tax had been invested in the stock market). At present Social Security looks a bit like a residual system: it is a redistributive, pay-as-you-go scheme that transfers income from the young to the old.7 It is vulnerable to demographic changes. There is no genuine “trust fund.” The anomalies of the institutional and residual welfare states come together in a predictably bizarre way when it is realized that poverty in America is a feature of the very young, not the old.


  1. Lawrence Mead, The New Politics of Poverty (New York: Basic Books, 1992), p. 181.
  2. See Norman Barry, Welfare, 2nd ed. (Buckingham, U.K.: Open University Press), ch. 8, and “Conservative Thought and the Welfare State,” Political Studies, vol. 45, 1997, pp. 342-45.
  3. See Norman Barry, “Friedman,” in V. George and R. Page, Modern Thinkers on Welfare (London: Harvester Wheatsheaf, 1995), ch. 2.
  4. The New Politics of Poverty, ch. 5.
  5. Lisa Oliphant, “Four Years of Welfare Reform: A Progress Report,” Policy Analysis No. 378, Cato Institute, August 2000.
  6. Michael Tanner, The End of Welfare (Washington, D.C., Cato Institute, 1996), p. 131.
  7. See Norman Barry, “The State, Pensions and the Philosophy of Welfare,” Journal of Social Policy, vol. 3, 1985, pp. 468-90.