America's dysfunctional healthcare financing system needs to be reformed. But the goal should not be universal coverage. Reform should simply aim to make health insurance more affordable and portable.

Universal coverage has so dominated the healthcare discussion that even some Republicans have tried to devise market-friendly ways to achieve it. The case for doing so is presented in practical, moral and political terms.

The practical case is that uninsured people raise premiums for everyone else. But such cost shifting raises premiums by 1.7 percent at most, according to a 2008 study published in the journal Health Affairs. Reforms that increase the number of people with health insurance, while stopping short of universal coverage, would presumably make that small percentage even smaller.

Efforts to eliminate this relatively tiny expense, on the other hand, would surely generate new costs. To mandate that everyone purchase health insurance, as many have suggested, would require that the government specify what constitutes adequate coverage — in other words, what health conditions an insurance policy would need to cover. Every provider group with a lobbyist, from massage therapists to fertility specialists, would want in. The result would be expensive insurance policies and costly government subsidies to help people buy them. Young and healthy people, especially, would be forced to overpay. So we would end up with more cost-shifting, and no savings.

This is another way of saying that universal coverage cannot be achieved using free-market methods — a point that many liberals correctly make. A bipartisan bill in the Senate introduced by Ron Wyden, an Oregon Democrat, purports to use the market to provide universal coverage. It would keep insurance companies in business, but only by converting them into regulated and subsidized public utilities, eliminating most existing insurance plans and expanding the IRS by a quarter. It owes more to Rube Goldberg than Milton Friedman.

The moral case for universal coverage is that we have an obligation to see to it that the poor and the near-poor have access to good healthcare. But universal coverage is only one way of realizing that goal, and not necessarily the best one. For people with pre-existing health problems, for example, direct subsidies would probably be more efficient than rigging insurance markets to make sure they are covered. As Michael Cannon, a health policy analyst at the Cato Institute, has written, "There is no evidence that a dollar spent on universal coverage will save more lives than a dollar spent on clinics, or reducing medical errors, or nutrition, or fighting poverty, or even improving education." And if universal coverage generally reduces the quality of care or retards medical innovation, it could end up being bad for everyone, including the poor.

The political case for universal coverage is based on the assumption that voters want it. But people's preference for universal coverage is not as great as their desire to reduce healthcare costs, a Kaiser Family Foundation poll found in late 2007. So it's not clear that people would accept higher taxes, mandates or the prospect of rationing healthcare one day just to make sure that every individual is covered. During the Democratic presidential primaries, Hillary Clinton repeatedly attacked Barack Obama's healthcare plan for not covering everyone — and as you may have noticed, he survived. If Democratic primary voters are not wedded to universality, the larger public surely is not.

An alternative approach would be to make it easier for people to buy insurance that isn't tied to their employment. The existing tax break for employer-provided insurance could be replaced with a tax credit that applies to insurance purchased either inside or outside the workplace. At the same time, state mandates that require insurers to cover certain conditions, which make it expensive to offer individual policies, could be removed.

These two reforms would address most people's anxieties about the healthcare system. Insurance would be more affordable, especially for people who cannot get it through an employer, so the number of people with insurance would rise. Indeed, this would enable more than 20 million more Americans to get insurance, according to a model created by Steve Parente, a health economist at the University of Minnesota.

More important, people would own their insurance policies and thus be able to take them from job to job. They would no longer need to worry about losing their job and their insurance at the same time, or feel they need to stay with a job they dislike because they need the benefits.

Critics of this free-market solution have argued that it would cause the current employer-based health insurance system to unravel. But that system is already unraveling, and if public health plans are created, as called for under Obama's proposal, they would unravel further as employers dumped their workers onto the public plans. A second argument is that people with pre-existing conditions would find it hard to get coverage. In fact, in the long run, the option to buy renewable policies that people could take from job to job would keep most people from needing to face this problem. Direct government subsidies could help the remainder.

The third complaint against free-market health insurance is that it wouldn't cover absolutely everyone, because it would neither force people to buy insurance nor require the government to provide it. Pharmaceutical companies and other provider groups would make a bit less money than they would if there were universal coverage — although they would probably be better off than they are now.

For most people, though, especially those in the middle class, it would mean paying less for health insurance. Some people, of course, would still choose to go without it. But that would be their call, as it should be in a free country.

Ramesh Ponnuru is a senior editor at National Review. This commentary was written for The New York Times.