« Deng Right? | Main | Paul Krugman: The Health Care Racket »

Friday, February 16, 2007

Daniel McFadden: An Evaluation of Medicare Part D

Daniel McFadden, winner of the Nobel Prize in Economics in 2000, evaluates the Medicare Part D prescription coverage program. He says that so far the results indicate the program has worked reasonably well, but "caution is advised" before implementing programs of this type in other segments of the health care industry:

A Dog's Breakfast, by Daniel L. McFadden, Commentary, WSJ: Last year, Medicare underwent a major expansion with the addition of Part D prescription drug coverage. A controversial feature of this new program was its organization as a market in which consumers could choose among various plans offered competitively by different insurers and HMOs, rather than the single-payer, single-product model…

Part D is a massive social experiment on the ability of a privatized market to deliver social services effectively. ...[M]y research group has monitored consumer choices and outcomes from the new Part D market. I will summarize our findings, but first I want to provide some perspective on the American health-care system...

Most Americans are aware that our health-care system is in deep trouble, a dog's breakfast of private providers and insurers that has weak and inconsistent incentives for quality control and cost containment. Many consumers cannot obtain health insurance at reasonable cost, and financing the system is stressing employers, pension funds, and the government's Medicare and Medicaid programs.

In terms of health delivered per dollar of cost, our system is grotesquely inefficient. ... Canada's single-payer system costs … about half our expenditure... The Canadian ... outcomes are typical of developed countries, where government managed and financed systems predominate. The extra cost of our system is not buying us better health. ...

In the future, things are going to get much worse. If current trends continue, health care in the U.S. as a proportion of GDP will rise to 40% by 2050, a level that will break the current system...

There are a number of reasons for this gathering storm. First, the U.S. population is getting older, and the old require more medical maintenance. Second, we are getting wealthier, and staying alive is the ultimate luxury good. Third, we demand expensive medical innovations... About a third of all medical costs are incurred in the last year of life, and are at best marginally effective. The incentives in the system do not force hard choices.

To deal with this future, three substantial reforms are needed. First, we need to wring out some of the inefficiencies. Something like 30% of our health costs come from administrative overhead, legal costs and defensive medicine. These could be largely eliminated...; we just need to emulate best practice in other developed countries.

Second, we need universal health insurance coverage, with active emphasis on preventive medicine... Perhaps this can be accomplished by cobbling together existing sources of finance, as in Gov. Arnold Schwarzenegger's current proposal for California. However, eventually we will have to go to a system that is not channeled through employers, something like a tax-financed medical voucher system.

Third, we need incentives that match choice of expensive treatments with consumers' willingness to pay for them, a benefit-cost analysis that places treatment choices and financial responsibility on the individual.

This brings us to Medicare Part D. This experiment in privatizing the prescription drug insurance market ... gives the individual the right, and responsibility, to make insurance plan choices that are in his or her self-interest. If consumers are up to this task, then their choices will ensure that the plans, and insurers, that succeed in the market are ones that meet their needs. However, if many are confused or confounded, the market will not get the signals it needs to work satisfactorily. ...

So, how well has the Part D market worked? ...

Consumers were quite confused about the Part D program before enrollment began... Start-up problems, widely publicized, further clouded the program's future. However, when the initial enrollment period closed, only 7.4% of the age 65-plus eligible population ... were not enrolled in Part D or comparable prescription drug coverage. It helped that monthly premiums, predicted to be about $37, were substantially lower... It helped that about 68% of the population was automatically enrolled... It also helped that Part D incorporates an annual subsidy of about $1,200 per enrollee, so that most potential enrollees benefit immediately.

Failure to enroll turned out to be a minor problem, but making Part D voluntary rather than mandatory did have some consequences. About 1.2 million seniors who use enough prescription drugs so that they would immediately benefit from enrolling failed to do so. Mostly, these were ... poorly educated people above the poverty line.

General opinions elicited after the close of the initial enrollment period give Part D a mixed report card. ...

However, when one looks at plan choices made, one finds that consumers were relatively consistent in recognizing their self-interest. Most consumers selected among the lowest-cost plans... There is some evidence that consumers did not fully appreciate the consequences of the gap, and enhanced policies that covered the gap were not heavily demanded...

The picture that emerges is that elderly consumers were mostly able to navigate the Part D market and reach reasonable choices, despite its novelty and complexity. Two concerns with insurance markets ... are adverse selection, in which only the sickest seek insurance, and insurers respond by ... refusing coverage for individuals and conditions likely to increase claims; and moral hazard, in which insurance coverage encourages additional treatments by reducing the incremental costs of these treatments.

Adverse selection has so far been avoided in the Part D market because the voluntary enrollment rate is high, including a high proportion of healthy seniors who currently are net contributors to the system. Further, adverse selection is not a problem for plan switchers because participating insurers must take all comers... There is moral hazard in the sense that prescription drug use is increasing for seniors newly insured under part D, to 4.4 from 3.3 prescriptions on average per month, according to our survey.

This increase must be assessed in terms of its health consequences. ... Anecdotal evidence indicates that Part D coverage will reduce medical problems and hospitalization costs enough to offset a significant portion of its cost. However, reduced adherence to therapies by consumers who hit the gap will probably have a significant adverse effect on health outcomes...

My overall conclusion is that, so far, the Part D program has succeeded in getting affordable prescription drugs to the senior population. Its privatized structure has not been a significant impediment to delivery of these services. ... We do not have an experiment in which we can determine whether a single-product system could have done as well, or better, along these dimensions, but I think it is reasonable to say that the Part D market has performed as well as its partisans hoped, and far better than its detractors expected.

Does this mean that Part D proves that privatization will be effective in other segments of the health-care market? Here, I think caution is advised. First, the success of Part D depends substantially on thoughtful and muscular management of the market. The former head of Medicare, Mark McClellan, and a dynamic, no-nonsense … government bureaucrat, Abby Block, bullied insurers to make sure there were, in her words, "no bad choices." It is unclear whether their successors will be as successful in standing toe-to-toe with the industry and making sure consumers' interests are protected.

A health insurance market like Part D probably requires this level of active management to work well... If privatization is going to work elsewhere in health care, active market management will be needed.

Finally, consumer-directed health care works only if consumers can understand the consequences of their choices. In much of medicine, providers are the agents that guide consumers through these choices. If consumer-directed health care is to be effective, these providers must give sound advice on both the health and financial consequences of alternative choices. This is possible if the incentives to providers and consumers are right, but the design of such markets should not be left to chance.

I agree caution is order and that substantial oversight will be required. I also think we should also wait until several years have passed before drawing any firm conclusions. This was a year with lot of publicity about the program, and with lots of watchful eyes and helpful hands attempting to guide market participants to the best outcome. The same amount of help may not be available in the future. In addition, it's also possible that problems with adverse selection will increase with time as healthy consumers learn more about the market and about alternative choices, and as insurers learn strategies to discourage people with potentially higher prescription costs from choosing their plans. It remains to be seen how much regulation and active oversight of these markets is needed as time goes on. McFadden, with his promotion of "active management" and not leaving things to "chance" leaves the impression that he believes it will be considerable.

    Posted by on Friday, February 16, 2007 at 12:06 AM in Economics, Health Care | Permalink  TrackBack (0)  Comments (26)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference Daniel McFadden: An Evaluation of Medicare Part D:


    Feed You can follow this conversation by subscribing to the comment feed for this post.