Dartmouth Debunked? Providers Don’t Drive Variation in Health Spending

Central planners love to cite the Dartmouth Atlas of Health Care. The Atlas is an impressive, decades-long effort to study geographic variance in health spending. The famous Atul Gawande, MD, is likely responsible for the fact that the Dartmouth results are better known among lay people than any other research in health economics.

The reason central planners love the Dartmouth results is that they easily feed into a narrative that goes like this: “Medicare spending in McAllen, Texas, is about twice as much as it is in El Paso, Texas, even though their populations are similar. The doctors in McAllen must be twice as greedy as the doctors in El Paso. So, we need to tighten the screws on Medicare payments until costs in McAllen are cut in half. If we do that nationwide, we solve Medicare’s fiscal crisis.”

One problem with this hypothesis is that it implicitly assumes that doctors in El Paso, despite living in easy reach of their colleagues in McAllen, have remained blissfully ignorant for years that they are leaving money on the table. If the Dartmouth hypothesis were true, the Hidalgo County Medical Society would set up a business to teach doctors in El Paso County, the rest of Texas, and eventually the rest of the United States how to bill Medicare like they do in McAllen. The cost differences would quickly be arbitraged away — upwards.

Our Greg Scandlen has previously discussed research that challenges the Dartmouth conclusions. Now, Louise Sheiner of the Brookings Institution has published research that further challenges the Dartmouth results.

This paper examines the geographic variation in Medicare and non-Medicare health spending and finds little support for the view that most of the variation is likely attributable to differences in practice styles. Instead, I find that socioeconomic factors that affect the need for medical care, as well as interactions between the Medicare system and other parts of the health system, can account, in an econometric sense, for most of the variation in Medicare health spending.

The paper also explores the econometric differences between controlling for health attributes at the state level (the method used in this paper) and controlling for them at the individual level (the approach used by the Dartmouth group.) I show that a state-level approach can explain more of the state-level variation associated with omitted health attributes than the individual-level approach, and argue that this econometric difference likely explains much of the difference between my results and those of the Dartmouth group.

In other words, demand rather than supply explains a lot of the cost differences. What is especially compelling about Sheiner’s analysis is that she looks at state-level variation. This is important because although Medicare is the federal government, providers are largely regulated by states. So, although this is a supply-side factor, it invites us to look at state regulations as a cost driver. (Until 2009, I used to write an annual report that ranked states by how they regulated providers, among other things.)

Print