Proponents of the new Republican budget are trying to cite one of Bill Clinton’s signature achievements as proof that their proposal can work. It’s not a very convincing defense.

The budget House Republicans approved last week would transform Medicaid from entitlement into a block grant. Instead of maintaining an open-ended federal commitment to the program, financing benefits for anybody that qualifies, the budget would simply hand state lawmakers a sum of money, set by a pre-determined formula, and let them decide how best to use it.

Liberals (like me) have warned that such a change will leave the program grossly underfunded, forcing states to cut back on eligibility, benefits, or both--and leaving poor people in desperate need of help. But proponents dismiss that argument, noting that critics said the very same thing about Clinton’s 1996 bipartisan welfare reform, which transformed the program into a block grant and never produced the predicted suffering.

Here, for example, are Peter Ferrera and Phil Kerpen defending the Republican budget on the Wall Street Journal editorial page:

The reform was wildly successful. The welfare rolls were reduced by two-thirds nationwide, and even more in states that pushed mandatory work most aggressively. As a result, in real dollars total federal and state spending on TANF by 2006 was down 31% from AFDC spending in 1995.
At the same time, because former welfare dependents were put to work, child poverty declined every year. Ron Haskins of the Brookings Institution reports that by 2000 the poverty rate for black children was "the lowest it had ever been," and because of their renewed work effort, low-income families formerly on welfare saw their total income increase by about 25%.
Block-granting Medicaid, as Mr. Ryan has proposed, would build on that success. ...

It’s true that welfare reform defied the dire predictions of critics at the time and that, by many standards, it’s been a success. But does that really tell us much about what a Medicaid block grant would do? Does it tell us anything at all? I don’t think so.

The most obvious difference between welfare and Medicaid is that the welfare is a program of cash assistance designed to help purchase basic goods and services--clothing, shelter, and so on. Medicaid is a program of subsidized health insurance designed to purchase medical care. The cost of medical care consistently and predictably rises more quickly than costs of other goods and services. But, under the Republican budget, the value of the block grant would rise only as quickly as the price of other goods and services.

That formula would mean the federal government spends far less money on Medicare, relative to current levels, over the next ten years--$771 billion, according to the official estimates. Supporters of the Republican budget, like Ferrera and Kerpen, claim states will find enough new efficiencies in the Medicaid system to offset that withdrawal of cash from the program. But with such a dramatic shortfall to overcome, these supposed efficiencies would have to be enormous--and that seems, well, impossible.

Medicaid is already much cheaper than either Medicare or private insurance. (See the graph below.) As my colleague Harold Pollack has noted here at TNR, "There’s no way to spend substantially less on it unless you want to serve far fewer people, provide recipients with far less financial protection, or underpay medical providers even more aggressively than Medicaid does now." And this is not merely his opinion or mine. Virtually every respectable authority I've read or seen, including the Congressional Budget Office, agrees.

But there’s another issue to consider. The theoretical underpinning of welfare reform was that the provision of cash assistance to the jobless was fostering a culture of dependency, discouraging people from taking jobs. Take away the welfare check, the theory went, and people would find work to replace their lost income. The government would save money and former recipients of welfare would be better off.

Maybe that theory was correct and maybe it wasn’t. (I'm willing to believe it had a lot of truth.) But it doesn’t really say much about Medicaid and its population. About a quarter of Medicaid beneficiaries are elderly, disabled, or both. Few of them can even even work, let alone find health insurance. The remainder of the Medicaid program consists of low-income adults and their children, which is the traditional welfare population. But the jobs available to these people are, typically, the very same ones that don’t offer health benefits. (And it's not like there are so many jobs available anyway.)

In other words, these people can’t do what welfare recipients of the 1990s did: They can't trade in government-issued benefits for an employer-issued alternative. Take away Medicaid and they’d be left with no insurance at all. And while it's fashionable in some circles to suggest no insurance is preferable to Medicaid--yes, Ferrara and Kerpen make this argument too--that argument also defies credulity, for reasons many of us have explained before.

By the way, if Ferrara's name sounds familiar, that's because he's one of the original, best known promoters of Social Security privatization--another program that, according to its supporters, promised vast government savings with no ill effect on beneficiaries. Make of that what you will.