A Success to Trumpet, and Protect.

THE results are in: Welfare reform is one of the greatest public-policy successes of modern times. There are now only half as many families on welfare as there were in 1994. Some states, led by innovative GOP governors, have shrunk their rolls by more than 80 percent. And almost everyone now agrees that the former welfare recipients have moved upward into the labor force, not downward into bleaker poverty. According to the New York Times, "the most comprehensive study of the new system says more and more [ex-welfare mothers] are going to work."

Conservatives should trumpet these accomplishments. We should say (adapting a phrase from James Carville) that we were right, and that our opponents were wrong.

How wrong? Recall the welfare-reform debate of 1995-96. Welfare advocates jeered at conservatives who were trying to fix the system. The Nation editorialized: "The welfare bill will destroy America's state of grace. In its place will come massive and deadly poverty, sickness and all manners of violence." That's not all. "People will die, businesses will close, infant mortality will soar." Rep. John Lewis of Georgia compared welfare reformers to Nazis, quoting Martin Niemoeller-"In Germany they came for the Jews and I didn't speak up because I wasn't a Jew. Then they came for the trade unionists and I didn't speak up because I wasn't a trade unionist"-and then bellowing: "Read the Republican Contract. They are coming for the children. They are coming for the poor. They are coming for the sick, the elderly, and the disabled." The anti-reform movement, led by groups like the Children's Defense Fund, forecast disaster-repeatedly and loudly.

But no disaster ever came. Incentive-based reform measures-such as work requirements, time limits for welfare, no extra payments for having more children, and aggressive job-placement efforts-appear to be having exactly the desired effect.

For almost three decades, the real-world impact of welfare confirmed what any sensible behavioral psychologist would have predicted from the outset: If you pay unwed teenage girls to have babies, they will have babies. If you say that a welfare check is an adequate substitute for a father's paycheck, then fathers will be expendable, and fewer fathers will stick around. And if you tell people that the checks that sustain them will keep arriving-unless and until someone in the household starts working and bringing in income-no one in the household will work.

Furthermore, even though welfare had always been sold to the public as a "temporary" safety net for the needy, by the end of the 1980s the average welfare recipient was staying on the rolls for twelve years.

In the early 1990s, some states launched elaborate systems of penalties and rewards in order to encourage welfare recipients to leave public assistance and engage in more productive activities-work, school, or training. Wisconsin governor Tommy Thompson mandated that all able-bodied welfare recipients work in order to receive benefits. In 1987, the number of welfare households in Wisconsin was about 100,000. Now it is below 10,000.

Many other states tried similar experiments. But it wasn't until the federal reform laws were enacted by the Republican Congress and signed into law by Bill Clinton (after two vetoes) that the national welfare system was reconstructed. This reform bill included all the provisions that the welfare lobby had vigorously opposed: work requirements, time limits, and a green light for state-level reforms. When Clinton signed the bill, Children's Defense Fund president Marian Wright Edelman castigated him for approving "the biggest betrayal of children and the poor since the CDF began."

For the first time in almost two generations, though, even most congressional Democrats had to concede that a major pillar of the New Deal/Great Society structure had failed. The initial success rate of welfare reform further exposes just how flawed the Great Society programs really were.

So where do conservatives go from here? First, we have to recognize that a lot of work remains to be done on welfare reform. Despite the encouraging trends, there are still 7 million people with an umbilical cord to the welfare state. How do we get this hard core off welfare and into the workforce?

The same crowd that opposed welfare reform in the first place (Bill Bradley is one of them) now argues that to end welfare for hard-core recipients, taxpayers will have to ante up a lot more money. The Democratic pro-welfare counteroffensive is based on a politically clever pro-children theme. Liberals are designing a new welfare state that looks suspiciously like the one Congress voted just a few years ago to get rid of; this time, their selling point is that welfare is not for the parents but for the kids-who, after all, can't possibly be held responsible for their parents' refusal or inability to work.

"Compassionate conservatives" will find it hard to resist this fluff. So we could end up with free medical care, free day care, free pre-school, and so on, for children. Bill Bradley's health program would cost at least $60 billion a year-a price even Al Gore attacks as excessive.

Under this new welfare regime, the mother would be responsible for taking care of herself, but the federal government would bear the cost of just about

every expense of raising her children. And this would put us right back where we started: with fathers being superfluous. It is important to remember that almost no two-parent families in America need government- subsidized health care, day care, pre-school, etc. The Census Bureau data are so clear on this point that it is beyond dispute: The best antidote to poverty is to keep father and mother together. But these new federal programs are de facto designed for families without fathers.

Charles Murray, whose 1984 classic Losing Ground inspired so much of the GOP welfare-reform effort, has correctly noted that ultimately, the test of welfare reform's success is whether the illegitimacy rate falls. If the percentage of children raised in a healthy family environment rises, we will have succeeded; if not, the cost savings will be mostly irrelevant, since the tragic social costs of the welfare state-largely borne by children-will remain. It's still too early to tell, but the initial signs are modestly encouraging-especially for teens. Out-of- wedlock birth rates have fallen slightly in recent years, with big declines among African-American women. But surely, creating a vast arsenal of means-tested children's programs would reintroduce the harmful incentives we have been trying to eradicate.

The best evidence we have suggests that the chief impediment to pushing hard-core welfare recipients into the workforce is that the value of benefits still remains very generous, relative to the economic return from working. That is to say, many of those on welfare today are simply making a rational economic decision: It's more financially rewarding not to get a job.

The perverse behavioral incentives erected by the welfare state remain much more of an obstacle to work than most analysts acknowledge. In 1995, my Cato Institute colleague Michael Tanner and I documented that in most states the full array of welfare programs-AFDC, food stamps, Medicaid, public housing, etc.-paid better than the after-tax payment from working in most starter jobs. In high-benefit states like New York, the value of the full package of welfare benefits can still reach $12 an hour, more than many entry-level salaries.

Our latest analysis finds that the higher the value of welfare in a state, the lower the declines in welfare caseloads. Hawaii, which offers the most generous welfare benefits of any state, totaling over $30,000 a year, has recorded the smallest caseload reduction in the nation. Conversely, Mississippi-whose welfare package comes to less than $11,000 a year-reduced its welfare rolls by 70 percent.

Fortunately, welfare-reform measures enacted at the federal and state levels have made work more financially rewarding. According to Ron Haskins of the House Ways and Means Committee, in the 1980s a typical welfare mother who went to work full-time figured to gain only about $3,000 in extra income from her efforts. Today, that same mother would gain $9,000.

The second round of welfare reform should achieve two goals: an increase in the rewards for work, and a decrease in the rewards of welfare. Work is the key. For all its faults, the one income-support program that has the fewest perverse effects is the Earned Income Tax Credit. The EITC provides working families with a livable income, by supplementing wages with a government subsidy. American taxpayers hate welfare, but they do want those who work 40 hours a week to have an income that provides them with all the necessities of life: health care, food, housing, etc.

The EITC is designed roughly along the lines of Milton Friedman's negative income tax. It has two big advantages over every other welfare program: First, the recipient gets a payment only by working. Second, there is no need for a vast and self-serving welfare bureaucracy (though we need to do a better job of weeding out rampant EITC fraud). The EITC is also far superior to the minimum wage, because the EITC does not increase unemployment; EITC payments above the market wage are made by taxpayers, not employers.

Republicans suffered a bout of temporary insanity last year in trying to scale back the EITC to save money. But the ideal welfare reform would dismantle all welfare programs except the EITC, and devote the cost savings to making the EITC somewhat more generous. Great care must be taken to reconfigure the tax system so that working families don't face insurmountable tax rates as they leave the EITC range.

The main virtue of this plan is that it would turn the work-disincentive problem of welfare on its head. It would also discourage illegitimacy as a means by which young girls can gain financial independence. Because it guarantees that working families will have the income needed to care for their children, this proposal is the best, and maybe the only, antidote to the expansions of government that Democrats are currently proposing. The plan represents truly constructive-and compassionate-conservatism.

In 1935, the founder of the modern welfare state declared: "Continued dependence on relief induces a spiritual and moral disintegration fundamentally destructive to the national fiber. To dole out relief in this way is to administer a narcotic, a subtle destroyer of the human spirit." The speaker was, of course, Franklin Roosevelt. The tragedy is that it took some 60 years for policymakers in Washington to conclude that he was right.

Republicans deserve high praise for the welfare reform of the 1990s. We may be winning, but we haven't won yet-and if we lose Round II, we will find ourselves right back where we started.

Mr. Moore is director of fiscal policy studies at the Cato Institute.