bq. Stand-alone outpatient prescription drug plans (PDPs), introduced in January 2006, have become the most popular source for coverage of outpatient prescription drugs under Medicare relative to other available Medicare plan types…Using county-level enrollment figures from the Centers for Medicare & Medicaid Services linked to other public sources, we study attribute substitution in beneficiary decision-making with respect to PDP enrollment. To do so, we relate county-level PDP market share to county-level political support for the administration implementing the new benefit (the Bush Administration), controlling for socio-demographic and market characteristics. We find statistically significant evidence that greater support for the Bush administration is associated with increased PDP market share.
This is a fascinating finding, one worth elaborating by quoting more from the paper:
bq. …the elasticity of PDP enrollment with respect to the Bush vote is 0.14. To obtain a sense of the relative importance of this effect, we calculated the change in PDP enrollment due to a one-standard deviation change in each of the independent variables separately. By this measure, the effect of the Bush vote is larger than the effect of all other independent variables except ‘Pct. eld. in poverty’ (the percent of elderly in county below poverty level) and ‘Lag pct. MA’ (percent target population in county enrolled in an MA plan in the prior year). Put another way, the effect of the Bush vote is larger than the effect of other variables that are generally accepted to be important and relevant factors associated with enrollment decisions: premium, educational attainment, urban/rural status, provider density, income, and risk score.
When Austin sent me this paper, he was curious about what might underlie this finding. Here are some various interpretations, working under the assumption that the finding is real (i.e., not spurious, etc. — which I have no reason to think it would be).
One possibility is that Republicans are simply taking cues from the Bush Administration’s championing of this initiative. Here are Frakt and Pizer:
bq. Our interpretation…is that the Bush vote variable measures the receptiveness of beneficiaries to the administration’s recommendations to enroll in the new drug plans. Beneficiaries who were receptive to the administration’s recommendations were more likely to substitute a favorable recommendation for the actual financial attributes of their options and enroll.
A second possibility — complementary to the first — is that how other politicians, e.g., state leaders, talk about the plan matters. That is what Austin suggests in his blog post. He is writing about the ACA, but it could be applicable to this finding regarding PDPs:
bq. If state leaders continue to describe that reform, or the mandate, in negative terms, I would expect much less enthusiasm for participation. On the other hand, if leaders talk about reform more positively, the population may more readily accept it, including the mandate. Since we’ll likely see, as we have, state variation in how local leadership talks about reform, that will play a role in the variation of acceptance of it and compliance with the mandate.
I’ll add a third possibility, thinking about Jowei Chen’s research on disaster relief. Chen finds that the Bush administration directed more FEMA aid to areas that tended to vote Republican. PDPs are not FEMA aid, obviously. But I wonder if the receptiveness of Republican-leaning counties to the Medicare PDP was due to direct efforts by the Bush administration to promote the PDP in those areas. In other words, the relationship that Frakt and Pizer found might not be the result of general partisan cues, passively received by fellow partisans, but of a more explicit effort to sell the policy in these localities. That’s just speculation, of course. I have no idea if this sort of salesmanship was happening on the ground.
I welcome other thoughts in comments.