Are There Differences in Medicare Spending Across States?

September 9, 2017

Research by the Dartmouth Atlas team has indicated that Medicare spending is concentrated regionally.  States such as Florida spend much more per beneficiary than do doctors who treat similar patients in low cost states like Minnesota.

A recent paper by Reschovsky and co-authors, however, has determined that variation in supply-induced demand is not a major driver in regional variation in health care costs once one controls for health status at the beneficiary level.  Read more to find out how the authors came to this conclusion.


Medicare spending is highly concentrate in a few number of high-cost beneficiaries.  ”The 5 percent of Medicare beneficiaries with the highest spending account for 43 percent of total program spending, while the top 25 percent of spenders, who often have multiple chronic conditions, account for 85 percent of total spending.

To determine if regional variation in practice patterns explains much of the reason for this skewed distribution of Medicare medical spending, the paper examines the impact of demand (e.g., patient health status and other characteristics, income) and supply (e.g., area level variables describing provider supply and market structure) side factors on regional variation in spending.  The data used include physician information from the Community Tracking Study matched to Medicare claims data.  The analysis is conducted at the beneficiary level rather than using area averages as other research has noted problems with the area level approach.


The research finds the following:

…a key finding from this work is that key conclusions from prior small area analyses (e.g., Fisher et al. 2003; Center for the Evaluative Clinical Services 2007) that much of the variation in cost of treating Medicare beneficiaries is driven by supply-induced demand (e.g., ‘‘supply-sensitive care’’) cannot be supported when one comprehensively controls for health status and conducts analysis at the beneficiary level

Section 1109 of the Health Care and Education Reconciliation Act of 2010, which amended PPACA, allocates 400 million dollars over 2 years to hospitals in the lowest cost quartile of counties (based on Part A and B spending adjusted for age, sex, and race). Apart from equity concerns about rewarding hospitals for costs largely attributable to other area providers, our results suggest that local health care costs are largely driven by disease burdens that go well beyond that which can be accounted for by demographic characteristics. The payments also fail to account for differences in outcomes across areas or hospitals.

Another finding of the paper is that the elderly are not sedentary.  In fact, one in five Medicare beneficiaries receive medical treatment in more than one census division.  These beneficiaries could either be snowbirds or the medical care they require is not available near where they live.  According to the provisions of health reform (i.e., PPACA), “Beneficiaries are to be attributed, ex post, to ACOs rather than prospectively assigned to ACOs. If care is not consistently concentrated within a narrow number of providers, especially among high-cost beneficiaries, ACO success in improving care and lowering cost may be limited.