In January, Congress tried and failed for the second time to override President Bush’s veto of a Democratic proposal to expand the State Children’s Health Insurance Program (SCHIP), a 10-year-old initiative that aimed to extend coverage to low-income kids. Congressional leaders promise another attempt in the coming weeks. The rhetoric has been hot: Republicans claim that the Democratic proposal would lead to socialized medicine by creating a middle-class entitlement; Democrats claim that their proposal is vital to poor kids. Unfortunately, this ongoing Washington power struggle will do little to strengthen America’s torn health-care safety net.

Despite all the debate, members of both parties speak glowingly of SCHIP’s success in insuring children. Bipartisan support exists for expanding the program—the fight is about how much. But SCHIP is in fact deeply flawed: in Minnesota and five other states, SCHIP dollars largely get spent on adults, not kids, and across America, many of the children in SCHIP already had private insurance before signing up.

SCHIP’s weaknesses are no surprise. Historically, similar federal programs have encouraged people to replace their private insurance with government coverage. Calculating the degree of such “crowding out,” to use the technical term, MIT’s Jonathan Gruber estimates that 60 percent of the new beneficiaries of government efforts to reduce the number of uninsured already had private insurance (though, for the record, the former Clinton advisor backs the Democratic position on SCHIP). In addition to shifting costs to taxpayers, public insurance programs aren’t particularly effective. Medicaid, for example, among the fastest-growing government programs, is rife with “waste, fraud, and abuse,” in the words of the New York Times. The administration of the program is so bad that, for several years, Medicaid in New York was covering Viagra for convicted sex offenders serving time. Meanwhile, the quality of care can sometimes be atrocious, with recipients unable to access specialists and facing caps on how many prescriptions they can receive each year. It’s not surprising that Stanford researchers found dramatic mortality differences between AIDS patients on Medicaid and those covered by private insurance.

Other federally funded public health programs, including those that serve the uninsured—like the Disproportionate Share Hospital program (DSH)—consume billions of dollars, which should be enough to cover those working poor who lack insurance year after year. But they don’t, in part because of poor design and bad administration. The result: millions remain without insurance, and millions more struggle to pay for their coverage.

A major reason that programs like SCHIP, Medicaid, and DSH are so inefficient is that, under the current funding regime, the more a state spends, the more federal matching dollars it receives. “Crowding out” occurs because governors get paid, in effect, to expand Medicaid without concern for focus or efficiency. What’s to be done? Washington has debated remedies for years—the Reagan administration wanted to hand Medicaid over to the states, while President Clinton favored a cap on federal Medicaid contributions. The most practical solution is to turn all federal health funding for poor and low-income citizens over to the states in the form of block grants that don’t rise as state spending does. This would make one level of government accountable for helping people get health care and coverage, and reduce bureaucratic waste.

State control would also encourage innovation. States have little interest in rethinking expensive programs today, since spending is largely on Washington’s dime: between 1980 and 2005, Medicaid spending quintupled, adjusting for inflation. Even when budgets are tight, states don’t reform their programs; they impose wage and price controls on doctors and pharmaceuticals. Thus, many of America’s poorest citizens have coverage on paper, but can’t find medical specialists willing to see them. Instituting block grants—with Washington giving up its role as master but allocating a fixed amount to states year after year—would create 50 laboratories of experimentation. Armed with health dollars, states would innovate.

Frustrated by mediocrity, a few states are already rethinking their programs, and these experiments are the most exciting developments in health care. Massachusetts is implementing a plan to cover everyone in the Bay State: it requires citizens to get insurance but provides numerous subsidies. Florida has turned over Medicaid to competing private insurance plans in two counties. South Carolina has redrafted its Medicaid program, giving recipients more ownership of their coverage. Bypassing bureaucrats, Colorado directly funds disabled Medicaid recipients to cover basic health needs. But in a nation that spends nearly a third of a trillion dollars on the health-care safety net, these efforts remain the exception, not the rule. We badly need experimentation. We don’t know what the exact prescription is for expanding coverage and reducing cost. Some states would implement free-market reforms; others would opt for bigger state-managed plans. My belief is that the former would work better—but no one knows for sure until we test more approaches.

Progress will come when a bipartisan majority embraces a smaller role for Washington. In today’s environment—where Republicans and Democrats busily compete to outbid each other on SCHIP funding—that idea seems farfetched. But 12 years ago, a bipartisan majority in Congress ended welfare as we knew it, sending poverty rates falling to the lowest levels in decades. The basic principle of that effort—state innovation—worked for welfare. It is also the key to health-care reform.

Dr. David Gratzer, a physician, is a senior fellow at the Manhattan Institute. He is the author of The Cure: How Capitalism Can Save American Health Care.