As proposals for health care reform begin to emerge from various Congressional committees, the attacks are also beginning. There is certainly room for debate, but we need to keep the debate from jeopardizing our chances for real reform.

[IMGCAP(1)]Two major objectives need to be achieved. We must ensure that all Americans have coverage for at least major health care expenditures. We must bring down the rate of increase in health care expenditures. The first is a moral imperative, the second an economic necessity.

The Democratic left is arguing for a national public plan based on Medicare. With enough subsidy, it could provide for universal coverage. Had we established such a plan when Medicare was enacted, we might not have the cost crisis we now face. Forty-four years later, however, we have a huge imbalance between primary and specialty care, entrenched special interest groups and a dysfunctional system. The transition to an effective and responsive health care system will not be achieved by legislative mandate.

The Democratic center and the Republicans oppose such a national plan — but without offering a viable alternative. President Barack Obama is rightly making clear that doing nothing is not an option. Here is a proposal to move us forward.

The proposal by Sen. Edward Kennedy (D-Mass.) includes an individual mandate and state-based exchanges to facilitate choice of coverage. But any employer contribution, no matter how small, should be eligible for tax exclusion — thereby benefiting low-wage and part-time workers. Such contributions need not be mandated. If caps are necessary to pay for subsidies, they should be set to reduce the tax exclusion for upper-income families.

The mandated coverage should focus on hospitalization and chronic illness — the major expenses that force the uninsured into bankruptcy. Specifying the complete minimum benefit package service by service will result in a Christmas tree bill laden to an unaffordable level by special interests. Far better are income-based subsidies lowering the costs for the poor, letting them decide what they need, with partial financial responsibility for the coverage that best suits them.

True cost containment, however, will come not from penalizing patients, but from realigning incentives for physicians and hospitals. This can be done voluntarily by offering bundled payments to those providers willing to accept responsibility for hospital episodes or for managing chronic illness. The focus on volunteerism is not just a nod to the right to facilitate compromise. It rewards those already doing a good job and encourages others to change. Those continuing their high-cost and low-value ways will have much less sway with the market than with the Congress.

A national public plan will be politically unable to select the “good— players like this. Instead, we should establish a publicly chartered risk pool for hospitalization and chronic illness. Private and state-based public plans can offer varying levels of comprehensive benefit packages through the types of exchanges Kennedy envisions, with mandates and subsidies to ensure everyone is covered. These plans will buy coverage from the major risk pool, reducing their cherry-picking efforts and lowering insurance costs. The voluntary risk pool will spread costs broadly across enrollees, and by not being a federal agency, will be able to offer new ways of payment that will lower the rate of cost growth.

The left is appropriately arguing for universal coverage, as it has for decades. But we now need a modern day approach to cost containment. A single-payer system cannot achieve that at this stage of the game. We can, however, marry universal coverage with market-based pressures to restructure the system. Doing so is not just good politics, it is good economics.

Harold S. Luft, Ph.D., is professor emeritus in health policy at the University of California-San Francisco. He is the author of “Total Cure: The Antidote to the Health Care Crisis— (Harvard University Press, 2008). More information is available at haroldluft.com.