Almost everyone agrees that America's health care system has the incentives all wrong. Under the fee-for-service system, doctors and hospitals get paid for doing more, even if added tests, operations and procedures have little chance of improving patients' health. So what happens when someone proposes that we alter the incentives to reward better care, not more care? Well, Rep. Paul Ryan and Republicans found out. No surprise: Democrats slammed them for "ending Medicare as we know it."

This predictably partisan reaction -- preying upon the anxieties of retirees -- must depress anyone who cares about the country's future. Spiraling health spending is the crux of our federal budget problem. In 1965 -- the year Congress created Medicare and Medicaid -- health spending was 2.6 percent of the budget. In 2010, it was 26.5 percent. The Obama administration estimates it will be 30.3 percent in 2016.

Uncontrolled health spending isn't simply crowding out other government programs; it's also dampening overall living standards. Health economists Michael Chernew, Richard Hirth and David Cutler recently reported that higher health costs consumed 35.7 percent of the increase in per capita income from 1999 to 2007. They also project that, under reasonable assumptions, it could absorb half or more of the gain between now and 2083.

Ryan proposes to change that. Beginning in 2022, new (not existing) Medicare beneficiaries would receive a voucher, valued initially at about $8,000. Suddenly empowered, Medicare beneficiaries would shop for lowest-cost, highest-quality insurance plans providing a required package of benefits. The health care delivery system would be forced to restructure by reducing costs and improving quality.

It's shock therapy. Would it work? No one knows, but two things are clear.

First, as Medicare goes, so goes the entire health care system. Medicare is the nation's largest insurance program, with 48 million recipients and spending last year of $520 billion. About 75 percent of beneficiaries have fee-for-service coverage. If Medicare remains largely fee-for-service, the rest of the system will, too.

Second, few doubt that today's health care system has much waste: medical care that does no good; high overhead costs. In a paper, Cutler documented some evidence. In one survey, 20 percent of patients reported that doctors repeated tests because records were unavailable.

Fee-for-service is open-ended reimbursement; the government's main tool to control Medicare's costs is to hold down reimbursement rates. Doctors and hospitals respond by ordering more services to offset the rate limits. For all its flaws, say Ryan's critics, this system beats his. Indeed, the Congressional Budget Office has estimated that in 2022, Ryan's plan would be more than a third costlier than the status quo.

But the CBO may be wrong. When a voucher system was adopted for Medicare's new drug benefit, the CBO overestimated its costs by a third; the Centers for Medicare and Medicaid Services' overestimate was 42 percent.

Under Ryan's plan, incentives would shift. Medicare would no longer be an open ATM; the vouchers would limit total spending. Providers would face pressures to do more with less; there would certainly be charges that essential care was being denied. The Obama administration argues that better results can be achieved by modifying incentives within the existing system. Perhaps. But history suggests skepticism. Presidents since Jimmy Carter have made proposals to control spending, with meager results.

It's Ryan's radicalism versus President Obama's tinkering. Which is realistic and which is wishful thinking? Burdened by runaway spending, Medicare "as we know it" is going to end. The only questions are when and on whose terms?