The Democratic Party's presidential candidates all believe that health care reform will help them in 2008. So far, only John Edwards and Barack Obama have presented detailed proposals (and Obama's is not all that detailed when you look closely), but every Democrat in the race has plenty to say on the subject. In Mitt Romney, meanwhile, the Republicans have a candidate who helped enact a universal health care plan as governor of Massachusetts, and is proud to say so. Still wary of "socialized medicine," he and the rest of the GOP field see widespread concern about the rising cost of health care. Like the Democratic candidates, they all promise to do something about it if elected.

In short, the politics of health care is once again coming to the boil. Something may even happen.

The second Democratic presidential debate was revealing. Hillary Rodham Clinton has been tentative on the subject up to now, anxious not to inflame memories of the "Hillarycare" fiasco of 1994. She has not yet announced a plan, but she showed no embarrassment or hesitation last Sunday, no steering off the subject at all.

To the contrary, her command of the issue was on bold display, packaged in a tacit acknowledgment that the Clinton administration got the politics (not the substance) wrong and that she had learned lessons. Articles insisting that she was right on health care all along have been appearing for a while now, and the idea no longer gets a laugh. (The most recent and most impressively argued of these is a piece by Jonathan Cohn in The New Republic.) The mood has shifted. When it comes to working out the details, health care may still be a political minefield, but all of the presidential candidates are preparing to march in.

This is a good thing. The country's health care system is needlessly expensive and needlessly unfair. In terms of aggregate health outcomes, it is not particularly effective. It leaves tens of millions uninsured and makes many of those who are insured anxious about future loss of coverage and the financial catastrophe that might ensue. It locks people into jobs they might otherwise leave, puts an immense financial and administrative burden on companies, and saps the country's economic vitality. It is broken and needs fixing.

The two main challenges are easily stated: Curb rapidly accelerating costs and achieve universal coverage. But those goals push in opposite directions. How can you cut costs and also cover more people? A variety of sleights of hand are being proposed, with differences attuned to partisan prejudice, to give the impression that you can pretty easily do both things at once.

Republicans are fond of arguing that the key to universal coverage is to get the price of insurance down. If only coverage were more affordable, everybody would choose to buy it. Actually they wouldn't, as is clear from other insurance markets. Many of today's uninsured could afford a health plan, but choose not to buy one.

The converse Republican illusion is to believe that mandatory universal coverage (as in the Massachusetts plan) will make insurance cheaper by bringing "free riders" into the system. Well, if insurance was cheaper, more people (not everybody) would buy it; and if everybody was herded into the system, costs per consumer would fall (a bit). In other words, cost-control and broader coverage are indeed complementary, but only to a limited degree—enough to make the cost of achieving universal health care lower than a straightforward extrapolation would suggest, but certainly not lower (other things being equal) than the cost of the current system.

Democrats sometimes borrow those arguments. Obama's plan relies heavily on affordability to broaden coverage (leading his rivals in the party to criticize it for failing to make coverage truly universal). But the principal illusion on the Democratic side is that any increase in overall costs, such as it is, can be piled without ill effect on businesses rather than on taxpayers or health care consumers. As Democrats tell it, the profits of insurance and pharmaceutical companies serve no economic purpose. They are rents arising from rigged markets: All of that money can be clawed back and set against the cost of fixing the system. If taxes still need to go up after that, employers at large can pay. No problem, they also are making too much money.

The likelihood—indeed the certainty—that piling costs onto business will lead to higher prices, less innovation, and lower wages is never confronted. Returning the compliment, by the way, supposedly pro-enterprise Republicans are not above deploying this kind of argument. Romney's Massachusetts plan pays for near-universal coverage with new business taxes and obligations.

Those who argue for socialized medicine—for a single-payer, government-run system along British or Canadian lines—are at least being more honest and more coherent. Set the government up as a monopoly buyer of health care, and it can use its power to drive costs down. But patients have limited choice in those systems, and access is rationed. In this country, those facts cannot be waved away. Americans are unlikely to take to such restrictions—not if they are happy (as most people with coverage say they are) with their present arrangements.

The French system is an interesting hybrid, closer to that of the U.S. system than you might suppose—with patient autonomy and multiple insurance intermediaries, but mostly paid for by taxpayers and with heavy government regulation. The French system is cheaper than America's but more expensive than Britain's or Canada's. It gets good results, but helps keep France's taxes very high. And American doctors would have some reservations about adopting such a system: They are paid about three times what their French counterparts earn.

A quite different reform strategy—which I think is preferable on the merits, as well as politically more feasible—is to retain the distinctively American aspects of this system, notably its reliance on competing private providers, while in key respects strengthening, not attenuating, the power of market forces. The crux of this idea is to give consumers real choices. That in turn can happen only if employers are largely taken out of the health insurance decision.

Employers do not insure your house or your car; why should they insure your health? No reason, except that a huge tax subsidy encourages them to do so. With employers selecting plans on their behalf, most workers cannot decide for themselves how to trade benefits off against cost—whether to sign up with a health maintenance organization (with the restrictions that imposes), or to pay more for a preferred provider organization, or more still for no-restrictions coverage. If consumers lack the ability to make such choices (and, having made them, bear the consequences on cost), the health care market is hardly a market at all, and the downward pressure on costs that market discipline would otherwise supply is never going to kick in. This is to say nothing of the other great drawbacks of relying on employers (including the problem of "lock in").

The administration has proposed one useful step in this regard: extending tax relief to individually purchased plans. Much more needs to be done to push employers out of the health insurance market. Most of the reforms now being touted, by Democrats and Republicans alike, aim to do the opposite.

Using market forces and consumer autonomy to drive down costs is worth a try, in my view, but that by itself will not solve the coverage problem. I would do that by guaranteeing everybody, regardless of income or health history, a voucher that would buy a standard HMO-type plan. This could be done in a variety of ways. You could give everybody a voucher (which could be used as partial payment for a more expensive policy) and recover the cost from general taxation. Or you could give full vouchers only to people on low incomes, tapering the value to zero as incomes rose, again asking taxpayers to pick up the check—and in this case also mandating that everybody buy at least the basic plan. You could satisfy the "ignore pre-existing conditions" criterion either through regulation or by adjusting the value of vouchers according to health risk. The cost would depend on such details, and many more besides—but you can be sure it would not be small.

I can understand why no Republican contender for the presidency is proposing this. Only a very unusual Republican would suggest a significant rise in taxes to mainly benefit the least-well-off. But why is no Democrat proposing such a scheme? Because it is politically more appealing to pretend that this great and long-overdue social advance can be achieved mainly at the expense of greedy employers (and grasping insurance and pharmaceutical companies).

It just ain't so. The saddest part, meanwhile, is that keeping up this pretense entrenches employer-based insurance, the very aspect of the present system that does most to inflate costs.