The expansion of the government's Medicare health-insurance program for the elderly to cover prescription drugs is more than a long-overdue acknowledgment of the growing importance of drugs in health care. It is a massive experiment in harnessing market forces -- consumer choice, competition, private insurers, private drug makers -- to improve a big government-run, taxpayer-subsidized program.

Market forces have a mixed history in shaping health care, and the debate over their efficacy continues. One camp argues that making consumers bear more of the cost, linking pay to performance and other market approaches are the only wise, fair mechanisms for allocating health care. The other camp considers that unworkable and unlikely to provide quality care to all, and argues that only a well-designed, government-organized system makes sense.

So what have we learned so far from the Medicare experiment?

Well, the market does sometimes work its magic. Eighteen months ago, there was speculation -- in this column and elsewhere -- that companies wouldn't offer a stand-alone insurance policy for prescription drugs. Today, there's a plethora of providers. The Medicare Web site offers my parents in New Haven, Conn., 45 options from 20 different vendors, several much cheaper than the projected $37 monthly premium.

Back in 2003, much of the focus also was on the peculiar structure of the benefit outlined in the statute: Beneficiaries pay the first chunk of their drug bills, then split the cost with the insurer for the next chunk, then pay all the rest (the so-called donut hole) until their annual tab hits $5,100. Today, with companies allowed to offer actuarially equivalent packages, all sorts of policies are available, some of which shrink that donut hole.

"If Congress had listened to the independent experts about what the benefit would cost and how to design it, we'd have people a lot less happy," says Mark McClellan, who oversees Medicare. "The drug benefit is going to cost a lot less" -- about 15% less than projected -- "and people are going to get coverage a lot more like they want, because they have choice."

So the pill-popping elderly, beneficiaries of billions in taxpayer subsidies, are thrilled, right?

Not quite. A Wall Street Journal/NBC News poll finds 40% of Americans 65 and over have an unfavorable opinion of the benefit and 23% a favorable one, with a sizable 37% undecided. Three-quarters deemed the plan "too complicated and confusing."

And, the government says that of the roughly 17 million Americans for whom the stand-alone drug insurance makes sense, only about one million have signed up. (Another 20 million, many of whom already had drug insurance, will get government-subsidized coverage through employers, health-maintenance organizations or programs for the poor.)

This market-friendly expansion of Medicare doesn't seem to be paying political dividends or attracting much consumer interest so far. "This is a test case," says Jonathan Oberlander, a University of North Carolina political scientist. "It's a theory in search of population, and they've found the wrong population. A lot of what you're seeing is what happens when you try to turn seniors and people with disabilities into computer-linked consumers. It's insanity."

Much of the problem may be that the drug benefit reflects political compromises that overlook lessons learned elsewhere about how consumers cope with an overwhelming array of choices.

Take Sweden's experience with mandatory individual retirement accounts. In 2000, when the program started, about four million workers of all ages were supposed to choose from 465 mutual funds in which to invest their savings; 67% of those eligible actually made a choice. The government assigned the rest to a fund designed for those who didn't choose. Now, newcomers to the job market have even more options -- 727 funds. Last year, only 8% made a choice.

"The more choices people have beyond a very limited number, the more likely they are to go to the default," the choice for people who don't make any choice, says R. Kent Weaver, a Georgetown University political scientist who has studied Sweden. That tendency is behind the move in the U.S. to automatically enroll workers in workplace retirement-saving plans, with the right to opt out, rather than expecting them to go through the hassle of enrolling. But because Congress insisted on making the Medicare drug benefit voluntary, a person who doesn't choose gets no coverage. (Read a summary of Mr. Weaver's work on Sweden.)

Mr. Weaver says the relevant lessons are: (1) When options are overwhelming, people are likely not to choose; (2) When everyone in the society, not just one segment, is offered a choice, and there is a lot of media attention and advice, more will choose; and (3) The government must devise a sound option for people who can't make up their minds. To which I'd add (4) Don't make the choice look even more complicated than it truly is.

All this is worth remembering the next time politicians revisit Medicare and Social Security, which they must.

Write to David Wessel at capital@wsj.com.