Arnold Kling  

Orszag gives my talk

PRINT
Subprime Daily Briefing, Dec. ... Behavioral Philosophy?...

Peter Orszag writes,


A congressional hearing tomorrow will focus on new long-term budget projections from the Congressional Budget Office. CBO projects that under current law, federal spending on Medicare and Medicaid measured as a percentage of gross domestic product will rise to 12% in 2050 and almost 20% around 2080 from 4% today. The bulk of that projected increase arises from steadily growing health-care costs per beneficiary.

...High-spending regions do not generate better health outcomes, on average, than the lower-spending ones. When health care at some of the nation's leading medical facilities costs half as much as care at other top-rated facilities for the same types of patients, something must be wrong with the system. Some academic research suggests that national costs for health care can be reduced by perhaps 30% without harming quality.

Understanding the reasons for such differences and finding effective ways to reduce them while ensuring high-quality care will not be easy. Potentially promising approaches include generating more information about the relative effectiveness of medical treatments, and enhancing the incentives for providers to supply, and consumers to demand, better care, rather than just more care.

Moving the nation toward a more efficient health system inevitably will be a process in which policy steps are tried, evaluated and maybe reconsidered. Beginning that arduous process now is essential to securing the nation's long-term economic future.


Orszag is Director of the Congressional Budget Office. He has a blog, which is pretty cool.

I am going to give a talk at noon today on Capitol Hill. The overlap with what Orszag wrote is fairly large.

When he says that steps need to be "tried, evaluated and maybe reconsidered," I agree. However, as a Masonomist, I think this makes government a poor candidate for leading health care reform. The government is not very good at trial-and-error learning.

For example, suppose we want doctors to have incentive pay. That means that we want them to change their behavior in response to bonus criteria. An effective system would create a large change in behavior at relatively low cost.

Politically, however, the survival of a system is likely to be related inversely to its effectiveness. If a system does not cost much and yet it pressures doctors to change their behavior, they will lobby to kill it. On the other hand, in the UK, a new incentive system generates large pay increases for doctors with little change in behavior. In that case, doctors will lobby to keep it.

With markets, "trial and error" works, because error gets killed off. With politics, error often survives.


Comments and Sharing






COMMENTS (2 to date)
Buzzcut writes:

Medicare costs are going from 4% to 20% because of per-user costs, not because of an expansion of the program as the population ages? That's the first time I've heard that.

Seems like we have a simple solution for that fact. Means test the program. Drop the number of beneficiaries. To do that, my personal preference would be to count medicare benefits as taxable income. While you're at it, make Socialist Insecurity benefits 100% taxable as well.

BGC writes:

Killing-off error is anathema to bureaucracies, who benefit from permanent efforts to reverse failure.

Here in the UK public sector bad schools ('failing schools') are never closed; which would be an easy, cheap and effective way of improving the educational system (together with expanding the good schools).

Instead, vast expense and effort are spent in 'turning around' failing schools. This generates jobs and builds empires in the bureacracy.

Also, it hardly-ever works - leading to permanent jobs, plus a new academic field of research of how to turn around failing schools.

No - the state can't do trail and error. Failure is too beneficial to too many people.

Comments for this entry have been closed
Return to top