Peter Orszag’s physician access argument

In a Bloomberg column yesterday, Peter Orszag made an argument against Medicare premium support I had not seen before. He cites research showing that as traditional Medicare’s market share grows, so does physicians’ willingness to accept Medicare patients. Reversing this, he argues that if premium support shifts patients out of traditional Medicare, physician access will shrink.

How important is Medicare’s market share in influencing physician participation? The evidence is limited, but the best study to date suggests it is significant. In the 1990s, Peter Damiano, Elizabeth Momany, Jean Willard and Gerald Jogerst, all associated with the University of Iowa, surveyed Iowa physicians and examined variation among counties. They found that for each percentage-point increase in the share of Medicare beneficiaries in a county’s population, doctors were 16 percent more likely to accept patients on Medicare. The only other study I know of on this topic, an unpublished analysis by Matthew Eisenberg of Carnegie Mellon University, also found an effect from Medicare’s market share, albeit one that was substantially smaller than the one Damiano and his colleagues found.

About 10 percent of the U.S. population is now enrolled in traditional Medicare, and an additional 5 percent has private Medicare plans. Let’s assume, for the sake of argument, that the Ryan [premium support] plan would cause another 5 percent of the population to shift, and to be conservative let’s cut in half the Damiano estimate of the impact from that reduction in Medicare’s market share. Then the chance that a doctor is willing to see traditional Medicare patients would be expected to decline by a whopping 40 percent. The share of doctors accepting Medicare would fall from about 90 percent to 54 percent.

I have neither read the studies nor ever considered this angle before. Are there good counterarguments? One might be that Orszag is extrapolating out of sample. Not having read the studies, I can’t be sure. Any others? I don’t know (yet).


Hidden information below


Email Address*