Arnold Kling

David Cutler on Health Care

PRINT
Ancient Technology and the Wea... What Austrian Brain Drain?...

I went to Cato to hear leading health care economist David Cutler and others discuss Pay for Performance (if you go to the link in a couple of days, you should be able to view the event). Beforehand, Robin Hanson and I got a chance, along with some Cato folks, to discuss the issue of widespread findings in cross-sectional studies of zero benefits from health care.

Cutler takes the view that health care on average is beneficial, but on the margin brings benefits close to zero. That is, we do a lot of stuff that has little benefit. So it seems that he and I would agree that Americans make extravagant use of medical procedures that have high costs and low benefits.

Nonetheless, Cutler argues that having consumers pay more for health care would lead to worse outcomes. He believes that "supply-side" policies, such as limiting availability of services or changing the way providers are compensated (the latter being pay-for-performance or P4P), can reduce ineffective care without reducing effective care. On the other hand, if you expose consumers to health care prices, they will cut back proportionately on all forms of care. Thus, if you try to reduce third party payments and shift to out-of-pocket spending (a central proposal in Crisis of Abundance), you will also reduce the use of valuable medical care, ultimately leading to worse health outcomes.

Ultimately, I think Cutler is taking the position that in health care consumer decisions are irrational and government intervention leads to better outcomes. My own view is that consumer decisions are irrational in all areas, not just health care. Still, government experts tend to do worse, as the experience with socialism demonstrates.

Even assuming that government experts have knowledge that doctors and consumers lack, it seems to me that the role of government ought to be limited to publishing its information. I just cannot get comfortable with government officials enacting taxes and subsidies in order to offset what they claim are individual biases in decision-making.

As for P4P, I think it is an intellectual swindle. Its advocates say that;

1. The correlation between pay and performance in medicine is currently low.

2. Quality in medicine is, in Cutler's words, "hit or miss." Many wasteful procedures are done. Some good procedures are foregone.

3. Therefore, P4P will improve quality.

I am confident that P4P can lead to increases in the use of some procedures and decreases in the use of others. The swindle is the implicit assumption this will raise the overall correlation between compensation and quality from approximately zero to approximately one. I think it will stay at approximately zero.

During the Q&A at the event, I compared government trying to implement P4P in health care to trying to implement government P4P for middle management. After all, middle management in America's big corporations and other organizations is also "hit or miss." Yet nobody thinks that a big project to have government pay bonuses to good middle managers at Intel or General Motors would solve the problem. Even government itself does not attempt to determine the pay of individual managers in such a centralized fashion. We don't think that people in Washington know more about your performance than the people who work most closely with you. My guess is that, using a program designed and implemented in Washington, we have about the same ability to affect the correlation between compensation and quality in medicine as we would in middle management.


Comments and Sharing





COMMENTS (2 to date)
Mike writes:
On the other hand, if you expose consumers to health care prices, they will cut back proportionately on all forms of care.

This is a testable hypothesis, no? How much faith do folks put into the results of the RAND health experiment for 25 years ago? I recall that consumer health outcomes were not adversely affected when given more responsibility for paying the costs of care.

Proposing another large scale experiment might be the way to go before advocating one way or the other here. But, do we have time to wait 10, 20 or more years for the results of such a study? I don't think anyone has the patience for that - and instead we'll have to do with some other way of evaluating such a claim.

-Mike

MQ writes:

There is a government-regulated system to provide incentives to the managers of private firms based on their outcomes. It's called the stock market.

Most of the stuff in this post just seems like assertions without much evidence. Medicine is different than other consumer goods.

MQ

Comments for this entry have been closed
Return to top