The admonitions were unsurprising, for the Carter administration's failed stab at reforming welfare had left deep scars. Joseph Califano, a Carter-era cabinet secretary, dubbed welfare reform ``the Middle East of domestic politics'' - a world of warring tribes and insoluble differences. But the Clinton people were having none of it. America had changed, they argued. There was now broad agreement on the disastrous nature of welfare. Experts hated the system; so did the poor. And Bill Clinton, who had vowed to ``end welfare as we know it'', was no Jimmy Carter. ``When you have a chance to lead a revolution,'' said one official, ``you shouldn't settle for anything less.''
A year later, the revolution may be at hand. On June 14th Mr Clinton unveiled his long-awaited proposal to put a two-year time limit on welfare; beyond that point, recipients would have to work, in public-service jobs if necessary. It is one of the most radical welfare plans ever proposed by an American president. But it has not had an easy gestation; and though attitudes to welfare reform are changing mightily, nor will it have an easy birth. For, as an issue providing grist for politicians, welfare remains as ideologically volatile and racially loaded as ever. As a system providing work for grim bureaucrats and slim hope for the poor, it is deeply entrenched. Mr Clinton is not the first president to promise to slash the dole: Roosevelt, Johnson, Nixon, Reagan and Bush did so too. All failed.
When Americans talk about ``welfare'', they are referring to a single programme, Aid to Families with Dependent Children (AFDC). It was introduced by President Franklin Roosevelt in the same 1935 speech that he used to announce the ending of almost all cash assistance to the able-bodied. Today the federal government still pays out next to nothing to such people if they are childless. But if you are poor, unmarried and taking care of your child (and hence usually a woman), you qualify for AFDC.
After growing gradually at first, the number of people on AFDC more than tripled between 1960 and 1974, to 10.8m, as eligibility rules were loosened and the value of welfare rose with the addition of free health care through the Medicaid programme. Recession has caused the rolls to swell again in the past four years, to 14.2m. A stunning one in seven American children is now on AFDC. The programme's cost to the government, of which the states pay nearly half, is $23 billion. Medicaid and food stamps add another $40 billion.
These numbers are record highs. Even so, the sums are less worrying than what lies behind them: a doubling of single-parent families since 1965. This has been due largely to a soaring rate of illegitimacy. Fully 30% of children are now born out of wedlock. Experts disagree about how closely illegitimacy may be linked to other ills such as bad school performance, drug use or criminal behaviour. But one correlation is clear. The poverty rate for children in single-parent families (55%) is four times higher than the rate for those in two-parent families.
The role of welfare in all this is hotly debated. AFDC pays, on average, only $370 a month to mothers-of-two, not enough to lift them above the poverty line even if Medicaid is taken into account. Few academics believe that is enough in itself to be driving the decline in two-parent families. For a start, illegitimacy rates rose sharply in the 1970s despite the fact that the real value of welfare was falling (see chart 2). Research that compared states with different benefit levels found no relationship between higher pay-outs and childbearing.
Most scholars share, nonetheless, the view of Mickey Kaus, a writer on social policy, that AFDC makes out-of-wedlock births easier by providing an ``economic life support system'' for single mothers. Worse, welfare creates disincentives to work and marriage: do either and benefits stop. This mix, argues Mr Kaus, fosters a culture of dependency. While most single mothers who claim AFDC do so for only a short time, 30% stay on it for eight years or more. These long-term recipients make up about two-thirds of welfare rolls at any one time.
Nowhere is the single-parent culture more evident than in America's mainly black big-city ghettos. The illegitimacy rate among all blacks is 66%. In the inner cities, where teenage motherhood is common, the rate is higher still. John Kasarda, an economist at the University of North Carolina, reports that most urban black single mothers do not work, and that 80% of this jobless group gets AFDC. Mr Kasarda thinks welfare has served as an ``economic substitute for (the) traditional blue-collar jobs'' that have been shifting from city centres to suburbs since the 1960s. Sensible Americans have agreed for some time that the welfare system is hurting those it is supposed to help; that it is rife with perverse incentives; and that it is implicated in the tragedy of an ``underclass'' concentrated in isolated pockets of extreme ghetto poverty, characterised by idleness, illegitimacy, drugs and crime. But few could agree, until recently, what to do about it.
Only in the 1980s, after years of rows between the right (which wanted welfare mothers to work) and the left (which wanted to throw money at them), did a tentative consensus begin to emerge. It held that to break the cycle of dependence, AFDC would need to become a transitional programme aimed at making the poor self-sufficient, and one that sent signals discouraging rather than rewarding illegitimacy.
One of the most forceful advocates of this new consensus was a Harvard academic named David Ellwood. He pointed out that many welfare recipients would be worse-off in minimum-wage jobs without health and child care than they were when receiving AFDC and Medicaid. He was also an early critic of the current child-support regime which allows many absent fathers to avoid helping their children financially. If welfare were to be a truly transitional programme, government would need to ``make work pay'' by guaranteeing health insurance and increasing the Earned Income Tax Credit (EITC), a subsidy for low-income workers. Next would come a system for withholding the wages of ``deadbeat dads''. Education and training would have to be beefed up.
Thus did Mr Ellwood's vision of welfare reform include lots of carrots. But once they were dangling, he was ready to wield a big stick: a strict time limit on AFDC, with last-resort public-sector jobs for those unable to find work in the private sector. That a liberal professor was willing to advance such an idea was a sign of how much the politics of welfare had changed. A still more striking one came when Mr Clinton, citing Mr Ellwood, made the idea a central part of his bid for the presidency.
Mr Clinton did more than campaign on the Ellwood plan. He made a$21 billion increase in the EITC part of last year's big budget deal. He launched a health-care reform crusade with the emphasis on universal coverage. He appointed Mr Ellwood and a Harvard colleague, Mary Jo Bane, to the Department of Health and Human Services (HHS). And he put them, along with Bruce Reed, a White House domestic-policy guru, in charge of the welfare taskforce.
The taskforce's plan, as now set out by Mr Clinton, contains few surprises. The new AFDC regime will offer expanded training, job placement and child care; it will crack down on absent fathers who try to avoid paying child support; and it will impose a lifetime limit (with some qualifications) of 24 months of cash support. If a welfare mother is offered a job, she must take it. If she is still unemployed after two years, she will be offered a publicly subsidised job at the minimum wage. The new system would be phased in, affecting the youngest (and hardest to employ) claimants first.
The Clinton plan is quite different from ``workfare'', a form of programme with which some states have been experimenting for a decade or so. With workfare, a mother who does not work can be ``sanctioned'' and so lose part of her AFDC cheque. With the Clinton plan, after two years there are no more AFDC cheques - only pay cheques. If you do not work, you get nothing.
The Clinton plan was not easily forged. The taskforce's members had their differences and their doubts. Mr Ellwood, having been called ``the father of time-limited welfare'', quipped early on that he wanted ``a paternity test''. Some talked of sending a message to unwed mothers; others feared that millions of children would be hurt. These sharp, agonised debates were held in secret, but their content was often leaked to a New York Times reporter, Jason DeParle.
That crafting the Clinton plan was such an ordeal is telling. There may be a new consensus over the framework for reform: but in welfare policy the details are all, and on these no agreement exists. In Congress, even first principles are in dispute. The left hates time limits; some on the right talk of ending not just welfare ``as we know it'' but of ending welfare altogether. Half a dozen competing proposals have been introduced in the past year. Bitter rows loom.
One big row will come over money. Voters may think that putting time limits on welfare will cut costs. Not so: sending cheques is cheap, when compared with offering training, child care and public jobs. But in last year's scramble for deficit reduction, the money for welfare reform was left out of the five-year budget. This, says a taskforce member, was ``the killer blunder''. It obliged the taskforce to propose new spending cuts or tax increases to raise the $9.3 billion (over five years) that its plan will cost.
The search was an anguished one. Opposition from liberals killed some options (such as taxing food stamps). Furious lobbying killed others (such as taxing the gambling industry). In the end the cash was found by diverting money from an environmental clean-up fund, cutting disability payments to drug addicts and reducing benefits to elderly legal immigrants.
Mr Ellwood thinks it ``amazing that the controversy so far has been all about financing and not policy''. True; but the former has huge implications for the latter. Originally, the taskforce had pushed for a more ambitious plan costing $16 billion over five years - the ``Cadillac plan'', as some called it. But because of Mr Clinton's unwillingness to offend any organised interest, the plan was scaled back to $9.3 billion - the ``Volkswagen plan''.
What was lost in the paring-back? Mainly, billions in new spending on child care for low-income working families. In private, Mr Ellwood and others argued passionately that the provision of child care would help prevent such families from falling on to the welfare rolls; and that giving it to welfare recipients but not to the working poor would create a new perverse incentive in the system. Without this spending, says one cabinet secretary, the whole welfare plan is ``seriously compromised''. Congress has other ideas about financing. Two of the main rival welfare bills - one sponsored by 160 House Republicans, the other by agroup of moderate Democrats - share much with the Clinton plan. But the Republican bill has a public-jobs programme twice the size of Mr Clinton's; and the Democratic one has much more cash for child care for the working poor. Both propose spending more than the administration's bill would do. Both propose a harsh innovation to raise the money: cutting off nearly all benefits to legal immigrants.
This last proposal puts the administration in an awkward position: it would be happy to see something closer to its Cadillac plan enacted, but it is opposed to a crackdown on immigrants who are playing by the rules. Officials say this would be unfair and might provoke a sharp reaction. Anti-immigrant sentiment may be running high, but cutting off all benefits would just shift more costs to the states - especially to politically crucial, immigrant-laden California.
On this - and perhaps only this - the administration can count on support from the left. Led by Bob Matsui, a Californian congressman, liberals will oppose such controversial measures as allowing states to deny additional benefits to women who have a second child while on AFDC. They are likely to rally round a bill seeking to expand the current welfare-to-work system embodied by the Family Support Act of 1988. This chronically underfunded law was supposed to turn the welfare system on its head. But it had no work requirement or time limit - which is precisely why the left likes it.
Then there is the right, which draws increasingly on the work of Charles Murray, a social theorist. Mr Murray's famously draconian idea is to abolish welfare. Non-work, he argues, is not the root of the underclass - illegitimacy is. Marginal reforms will not alter childbearing behaviour - only ``pulling the plug'' will do so. The sway of Mr Murray, who once dared offer his ideas only as ``thought experiments'', is the most vivid evidence of how far the welfare debate has shifted towards the right.
A few congressional Republicans support getting rid of welfare, and others want to halt AFDC to all women under 21. But many more cherish Mr Murray because he supplies them with intellectual justifications for opposing Mr Clinton. Never mind that the president is proposing the sort of time limits and work requirements for welfare that conservatives have long championed. Empower America, a group headed by prominent right-wingers, urges Republicans to ``discredit (Mr Clinton's) moderate pretensions'' and so seize ``an opportunity in the realm of politics''.
With health care occupying Congress's every waking moment and the liberal House leadership less than keen on welfare reform anyway, the fate of the Clinton plan will probably not be clear until next year. The administration is right to insist that there is broad acceptance in Congress, and in the country, of a two-year time limit. But it is also plain that the consensus may be too fragile to support radical change.
Assume, for the sake of argument, that the consensus holds; assume that the Clinton plan is enacted in toto. As Ms Bane rightly points out, passing a law is only 10% of the battle. Reform will succeed only if it transforms welfare from a programme that can be a way of life into one that has clear limits, and as such sends a strong signal to young women who are thinking of having babies out of wedlock. This will depend largely on whether the two-year time limit is made to stick; which in turn poses challenges that make passing the bill look a cakewalk.
Start with an uncomfortable fact about many welfare recipients, especially the long-term ones at whom reform is aimed primarily. Often ill-educated and ill-trained, they are among the least employable people in the labour force. Donna Shalala, the secretary of HHS, reckons that fewer than 20% of those on the rolls are ``unemployable''. But in state after state, programmes designed to get recipients into private-sector jobs have performed dismally.
Programmes stressing education and training have rarely reduced the rolls by more than 10%. Those focusing on job placement have proved a bit more promising. One such programme in the county of Riverside, California, has raised wages and cut welfare costs, and is often cited as a national model. Yet according to a new study by the Manpower Demonstration Research Corporation (MDRC), a welfare-research group, fully 50% of Riverside's welfare recipients remained on welfare rolls after two years.
Paul Offner, an influential aide to Senator Daniel Patrick Moynihan, believes that imposing a serious time limit on AFDC would require the creation of 1.5m public jobs. The administration hopes more jobs can be found in the private sector: it intends creating only about 400,000 public jobs by 2004, a point at which about 75% of welfare claimants would be subject to the two-year limit. It worries, too, that union members hate the idea of big public-jobs schemes; so do conservatives. And the public might come to see too big a jobs programme as welfare by another name.
The administration hesitates to put a time limit on how long anyone can stay in a public job, but Mr Reed argues that welfare reform would be failing if it turned into a ``giant public-works programme''. So if the time limit is to be a serious one and yet a lid is to be kept on public jobs, then the system will have to become much better at moving welfare mothers into the private sector.
That means turning a system that revolves around cheque-writing into one that revolves around job training and placement. Today's welfare administrators are often not much sparkier than their clients. They excel at handing out reams of forms to determine whether people are eligible for AFDC, but not at helping them find work. ``When we talk about changing the culture of welfare,'' says congressman Dave McCurdy, chief sponsor of the moderate Democratic welfare bill, ``we're not talking about welfare recipients; we're talking about welfare bureaucrats.''
Good ideas for how to do this abound, but are largely unproven. Indeed, the truth is that everything about time-limited welfare is unproven. In that sense, Mr Clinton's plan constitutes an enormous risk. If too many people reach the two-year cliff and fall off, the public-jobs programme will be overwhelmed. Some people will refuse to work at all. What will happen to them? And to their children?
Such are the questions thrown up when policy theorists begin to think of themselves as revolutionaries. Typically, the answers are anti-climactic. The revolution stalls; nothing much changes. Yet that, too, would imply a huge risk, because America's welfare system does have to change - not for the sake of voters or politicians, but for the sake of the poor. If Mr Clinton fails, the next band of zealots to storm the barricades of policy on poverty may have the poor's interests less clearly in mind.