Why better incentives won't lead to better health care.

Why better incentives won't lead to better health care.

Why better incentives won't lead to better health care.

How to fix health policy.
Aug. 4 2009 1:09 PM

The Wrong Cure

Better incentives, for doctors or insurers, won't lead to better health care.

Patient and doctor. Click image to expand.
A patient and doctor

The problem with America's health care system is not primarily one of costs, though reform should contain them, or one of care, though reform should improve it. It is one of incentives. The question is this: How do you incentivize better medicine?

The problem with incentives, as 35 years of research in psychology shows, is that you get what you pay for. Teachers incentivized to produce higher test scores get higher test scores but not better-educated students. CEOs incentivized to improve the performance of the company's shares improve the performance of the company's shares but not the performance of the company. The common dynamic at work is simple: You incentivize something that is both a good index of the thing you really care about and is easily measured. Standardized test scores are an index of learning; share price is an index of company performance.


Before long, however, people subjected to these incentives find ways to improve their standing on the index while also changing the index so it is no longer a reliable measure of the thing it was created to evaluate. Teachers teach to the test, and CEOs operate with very short time horizons. Then, when institutions malfunction, our first impulse is to blame not the people themselves but the "dumb" incentives. This kind of criticism was common in response to the near collapse of our financial system. And it is common in response to the high cost of our system of health care.

Take the example of health care reform in Massachusetts, where health insurance is mandatory and is costing more than the state can afford. Massachusetts currently has a fee-for-service system, which gives doctors an incentive to order more tests, which cost more money. So a state panel has recommended that Massachusetts move to a per-capita system, where patients pay a set fee. If their medical costs turn out to be less than their fees, the provider network would keep what remained as profit. Insurers would then have an incentive to keep costs down.

If this scheme looks familiar, it should. It amounts to a so-called "capitation" plan. This approach to financing health care became popular about 40 years ago with the rise of HMOs. As sociologist Paul Starr points out in The Social Transformation of American Medicine, HMOs started out as nonprofits inspired by a noble vision of medical care. The idea was to charge patients (or their insurers) annual fees and cover everything—with an emphasis on prevention and well-patient care. Do more doctoring when it's cheap so that there will be less to do when it's expensive. What quickly happened is that idealistic, nonprofit HMOs were driven into the margins of medical practice by the for-profit variety. In nonprofit HMOs with a fee-for-service plan, doctors have an incentive to do more. In for-profit HMOs with a per-capita plan, doctors have an incentive to do less.

Given this history, one wonders why the Massachusetts commission expects a different result this time around.

Massachusetts' attempt to create incentives that are "smarter" is probably doomed. It's relatively easy to eliminate incentives that are dumb, but it's amazingly difficult to come up with incentives that are smart enough to do the job. After all, the incentives that contributed to our financial fiasco were hailed, just a generation ago, as a brilliant feat of human engineering. Yes, in principle, share price can be a reflection of the general health of a company. But once you incentivize share price, smart people can find a way to move it up without really improvingor sometimes even jeopardizing—the overall health of the firm.