Working Papers
Geographic Variation in Healthcare Utilization: The Role of Physicians (with Amy Finkelstein, Matthew Gentzkow, Peter Hull, and Heidi Williams)
Abstract: We study the role of physician practice style in driving geographic variation in Medicare utilization. To do so, we estimate a model that allows for variation in patient demand, physician treatment intensity, and regional supply-side factors, as well as patient-physician sorting. The model is identified by quasi-experimental migration of Medicare patients and physicians, as well as their matching within regions. We find that physicians vary greatly in the intensity with which they treat otherwise similar patients. Our baseline decomposition suggests that about 30 percent of regional variation in health care utilization is explained by differences in average physician treatment intensity, 20 percent by other area supply factors, and 50 percent by differences in patient demand. An important margin of both patient demand and of other area supply factors is the number of different physicians seen. Within-region patient-physician sorting is relatively unimportant in explaining geographic variation.
Surgical Innovation and the Challenge of Changing Demographics
Abstract: I study the impact of market size on the development of novel surgeries, an important domain of medical innovation where many of the institutions traditionally considered essential to investment in research and development - intellectual property rights, approval regulation, and financial incentives play only a minor role. Using the codification of ICD9-CM procedure codes as a novel measure of new-surgery development, I ask two questions. First, is surgical innovation different from innovation in pharmaceuticals, a sector where traditional innovative institutions are salient? Comparing aggregate trends in pharmaceutical and surgical innovation in similar therapeutic markets, I find that the two processes follow different trends. Second, given this difference, does surgical innovation still respond to changes in patient demand? Using nationally representative inpatient and outpatient discharge data and quasi-exogenous changes in potential market size due to shifting US demographics, I find a positive and significant elasticity of surgical innovation with respect to potential market size. This elasticity is nonetheless lower than comparable estimates from the pharmaceutical literature. My results suggest that while surgical markets seem to respond to changes in medical need, thereby alleviating a first-order policy concern, the exact mechanism behind this responsiveness remains an open and important economic question.
Abstract: A large literature in healthcare economics connects the extraordinary rise in medical benefits and healthcare spending over the past century to the fast pace of innovation in the medical sector. In this study, I investigate the effect of age on a surgeon's propensity to learn new medical procedures. A major challenge in this type of analysis is identifying large numbers of medical technologies undergoing diffusion and a well-defined set of physicians at risk of adopting. I address this problem by exploiting the parent-descendant relationship among ICD9-CM inpatient procedure codes where each newly introduced code has a well-defined antecedent that was used to record both the novel procedure and its older alternative prior to introduction. Using the universe of new ICD9-CM procedure codes released between 2001 and 2013, I show that surgeons that are 10 years older at the time of new code approval are 1.8 percentage points (translating to sixteen percent relative to baseline) less likely to use a new code. The estimates imply that if all surgeons were to gain the adoption rate of the youngest cohort, procedure diffusion rate would increase by 19 percent. The difference comes from extensive margin adoption decisions and increases as the new code advances through its diffusion process. I show that the results are not affected by controlling for surgeon patient mix and a variety of surgeon observables. Evidence from analogous analyses in the diffusion of pharmaceuticals, diagnostic codes, and the subset of minimally invasive procedures suggests that this effect is likely to be driven by skill acquisition costs rather than information frictions.
Evidence on the Responsiveness of Export-Related VAT Evasion to VAT Rates in the EU (with Jon Bakija)
Abstract: In almost all countries that operate a value-added tax (VAT), the VAT “zero-rates” exports, meaning that exporting firms can claim credit for VAT on their inputs, but pay no VAT on sales of exports. This is necessary when the goal is to make the base of the tax domestic consumption. A consequence is that firms can reduce their tax burdens by misreporting some of their sales to domestic consumers as exports. In principle, reported exports from country i to country j should match up with reported imports into country j from country i, except for measurement errors, and costs of insurance and freight that are included in import value but not export value. VAT evasion can be another source of discrepancy which would tend to cause reported exports to exceed reported imports for trade flows in the same direction. We use data on such discrepancies in trade flows between pairs of European Union member countries during 1984 through 2011 to infer whether higher VAT rates are associated with greater over-reporting of exports. A difference-in-differences identification strategy, exploiting the fact that VAT rates changed in different ways over time in different countries, suggests that each percentage point increase in the exporting country’s standard VAT rate increases the discrepancy of exports over imports by about 1.1 percent of exports. For the typical EU-15 country, this implies that at the margin, about 15 percent of the static revenue gain from a VAT rate increase would be lost due to this particular channel for VAT evasion.