Works and Welfare.

From 1960 to 1970, New York City's welfare caseload more than quadrupled, from about 50,000 cases to 220,000 cases. Spending went up even faster because grant levels were also increasing rapidly. The 1960-70 growth spurt also signaled a new disconnect between welfare and the real economy, for it came at a time when New York was enjoying unusually strong economic growth. Between 1965 and 1970, the city added almost two hundred thousand new jobs, and by 1968, New York City's unemployment rate, at only 3.1 percent, was the second lowest big-city rate in the country after Dallas.

Clearly, as Joseph Heller might have put it, "something happened." Whatever it was, about five years later it happened everywhere else in the country too. Welfare caseload growth in the United States between 1965 and 1975 almost exactly matched the 1960-70 growth spurt in New York.

What most people understand by "welfare" is the program formerly called "Aid to Families of Dependent Children," which the tougher rhetoric of welfare-form legislation has redubbed "Temporary Assistance to Needy Families." Surely, no other federal program generates more megawatts of controversy per dollar expended. The president's fiscal year 1999 budget projects $16.5 billion for welfare grants (states will roughly double that figure), or less than 10 percent of the total allocation for "income security," which includes unemployment compensation, food stamps, and the like, and only a miniscule fraction of the vast outlays for Social Security and medical assistance programs. Last year, the Congress forced the military to spend more than $16.5 billion on weapons the generals insisted that they didn't even want.

Robert Solow is a Nobel-laureate economist, with a long interest in welfare issues, and his slim little volume is an elegant attempt to discover why the program is so controversial, and to explore the policy implications of his conclusions. The book comprises two lectures given at Princeton University in the 1996-97 academic year, plus commentary by economists Glenn C. Loury and John Roemer, and by Gertrude Himmelfarb, the social historian, and Anthony Lewis, the political columnist.

Solow locates the welfare controversy at the nexus of two important social values, self-reliance and altruism. The traditional American ethos limits the proper objects of altruism to those who cannot be expected to be self-reliant. The original New Deal family-welfare legislation was inspired by the plight of "widows and orphans," victims of circumstances outside of their control - in short, the deserving poor.

The extraordinary 1960-75 welfare growth spurt came after social work professionals, lawyers, and other advocates expressly challenged the assumption that welfare eligibility should be linked to individual deserts. Welfare became an "entitlement," a form of "new property" in some formulations. The old system of caseworker supervision, with individually tailored grants and regular home visits, which in practice was often intrusive and racially discriminatory, was replaced by an impersonal grant-dispensing machinery which, in the early years at least, had almost no controls at all.

It is absurd, as some conservatives tend to do, to blame the demoralization of the inner cities on the easier availability of welfare. But liberals tend to fall into the opposite trap. In his commentary, Glenn Loury eloquently documents the behavioral pathology in some slum neighborhoods - the lack of stable attachments to marriage, school, or work, especially among young fatherless males - but insists that it has nothing to do with welfare, which seems equally absurd. Unmarried teen-age girls don't have babies to maximize welfare benefits, as a Charles Murray might argue, but it's hard to believe that the easy availability of welfare didn't make some contribution to making single mothers and fatherless children the norm in many neighborhoods.

Solow's solution, which he offered before the recent welfare-reform legislation was passed, might be termed "Clinton-plus" - to relieve the strains on the American store of altruism by requiring self-reliance on the part of the poor to the greatest extent decently possible. He would accomplish this by conditioning welfare benefits on the obligation to work or train for work. Solow is a director of the Manpower Development and Research Corporation (MDRC), the premier research organization dealing with welfare and work issues, and he supports his argument with extensive research data suggesting that the great majority of welfare recipients really do prefer work to welfare. Indeed, it has almost always been the case that most welfare recipients work; but the jobs are too unstable, irregular, or poorly paying to keep them off the rolls.

(Anthony Lewis is astonished at the revelation that most welfare recipients work. It is more astonishing that someone who has written on welfare issues as long as Lewis has didn't know that.)

The crux of Solow's argument, however, and the critical difference between his proposals and the recent "reform" legislation championed by the Republican Congress and neoconservative Democrats is his insistence that many, perhaps the vast majority, of welfare recipients will never be self-supporting, and may never be able to find jobs in the current economy. He makes the case for workfare in direct opposition to market fundamentalists who insist, in effect, that if the poor are forced to shift for themselves, they will.

MDRC has amassed impressive research on this question. For whatever reason, poor people tend to have very limited skills. Training, counseling, and other supports help some, but only a little. High-tech America does not place a high value on low-skilled people. When the time limits on the newly "reformed" welfare program start to run out in a couple of years, large numbers of people, deserving and undeserving alike, are going to be in very serious trouble.

The only way to protect both American traditions of altruism and self-reliance, Solow argues, is to ensure that jobs are available for all welfare recipients who must work. He makes no attempt to work out the mechanics of a program, but it is clear that it must be some combination of earned-income tax credits, cash grants, and supplemental medical and other assistance, training, and, above all, large-scale creation of subsidized public or private jobs for the less skilled. That will be very expensive, but there is a great deal at stake - not just a decent sufficiency for the least able poor, or the preservation of cherished American values, but something like our national integrity.

Charles R. Morris was the administrator of the New York City and Washington State welfare programs in the 1970s. He is the author of American Catholic (Times Books).