For at least 20 years, commentators have bewailed the lack of adequate health insurance among growing numbers of Americans. For an even longer time, analysts and experts have warned that health care costs were rising unsustainably. Yet no consensus has formed on what to do about these twin adversities. Conservatives propose encouraging individuals to buy private health insurance and placing more reliance on market forces. Liberals continue a struggle initiated during the New Deal to provide publicly financed health coverage. Neither has persuaded the other.

The result: a political standoff that has blocked an honest evaluation of any major approach and has left everyone frustrated. Meanwhile, the ranks of the uninsured swell, health costs soar, and states and businesses cut benefits.

It seems clear to us that the political logjam in Washington will continue unless we take a different approach. We believe that approach involves exploiting the unique strength of American federalism. Specifically, we urge Congress to authorize individual states to carry out any of a broad range of strategies to reduce the number of uninsured people within their borders. And Congress ought to reward states with grants sized according to their progress toward agreed targets. The State Children's Health Insurance Program, enacted in 1996, is a close cousin to this approach. So too is the process of state innovation that led to the welfare reform legislation passed that year.

The state-federal approach that we propose would work like this:

First, Congress would specify goals, which would include reducing the number of uninsured people over several years. Defining coverage targets would be difficult, because of states' widely varying percentages of uninsured people.

Setting goals does mean Congress would have to define a minimum level of coverage to classify a person as "insured." We believe that for both political and policy reasons, any such standards should sustain the Medicaid entitlement, financed by an uncapped federal matching grant. At the same time, Congress should encourage states to explore better ways of delivering services. Defining what constitutes insurance for others -- those not entitled to Medicaid and lacking private or public health coverage -- would be controversial and difficult, because reasonable people disagree passionately on whether good insurance must provide comprehensive coverage or simply less-expensive catastrophic protection, with individuals being subject to price incentives for other costs.

Conservatives also fear that statutory minimum benefits would lead to standardized and expensive benefits, as provider groups lobbied to add requirements while thwarting attempts to pare outdated services. Liberals fear that the benefit standard would be set so low that people would suffer ruinous costs or be forced to forgo needed care. One way of avoiding a political impasse on this issue might be to allow some variation in basic benefits among states, provided the plans are of equivalent value.

Under the second element in our strategy, Congress would pass a "federal policy toolbox" of options that would be available to states. This menu would include tax credits to individuals for the purchase of group insurance or to businesses for mandatory employment-based coverage; Medicaid or other public program expansions; individual mandates; plans that focus on children or the near-elderly; so-called association plans; or an expansion of the Federal Employees Health Benefits Program to nonfederal workers.

Third, states willing to participate could choose one or more items from this menu or propose their own strategies that they found suitable to the needs and preferences of state voters, within the minimum federal standards. Single-payer plans could be tested on a limited scale, avoiding a clash over the law governing large corporate plans. There might be negotiation over the fine print of the goals or the benefit package, but other than that, the state would normally get a green light. Approval of state plans might be handled best by a bipartisan commission.

The final element -- necessary to make the effort attractive and affordable to the states -- would be financial support. Congress would commit in advance to provide grants to each participating state in addition to current federal funding, based on the state's progress toward the agreed goals. Congress should also require and pay for careful studies of progress and of the ways in which health care delivery, costs and use respond to each approach.

The bottom line for states would be straightforward: Reduce the ranks of the uninsured and you qualify for additional federal aid. How you do it is largely up to you; if you make no progress, you get nothing.

We believe it is time for those who disagree about how best to extend health insurance -- as we do -- to recognize that narrowed coverage is increasingly dangerous to the nation's social, economic and physical health. To break out of today's ideological impasse, we must agree that while some basic federal standards are needed, states as diverse as Oregon, Utah, Texas and Massachusetts must be free to adopt diverse strategies to reach those who are uninsured. In addition to making substantive progress on insurance coverage, this process of policy discovery would provide information that could spur future changes in health policy. If Congress creates the authority and offers the financial support we recommend, we do not pretend to know whether state health policy will diverge or converge over time or whether Congress will use the knowledge generated by state experiments to fashion a national health care policy. But we are certain that health policy cannot improve unless it moves off dead center, where it has been stuck for a generation as more Americans have lost coverage and costs have exploded.

Henry J. Aaron is senior fellow in economic studies at the Brookings Institution. Stuart M. Butler is vice president for domestic policy studies at the Heritage Foundation.