Content area
Abstract
The first chapter of this dissertation explores how the regulatory approval process affects innovation incentives in medical technologies. While prior studies of medical innovation under regulation have found an early mover regulatory advantage for drugs, I find the opposite to be true for medical devices. Using detailed data on over three decades of high-risk medical device approval times in the United States, I show pioneer entrants spend approximately 34 percent (7.2 months) longer in the approval process than the first follow-on innovator. Back-of-the-envelope calculations suggest that the opportunity cost of capital of a delay of this length is upwards of 7 percent of the total cost of bringing a new device to market. I consider how different types of regulatory uncertainty affect approval times and find that a product's technological novelty is largely unrelated to time spent under review. In contrast, uncertainty about application content and format appears to play a large role: when objective guidelines for evaluation are published, approval times quicken for subsequent entrants. Finally, I consider how the regulatory process affects firms’ market entry strategies and find that financially constrained firms are less likely to enter new device markets as pioneers.
The second chapter considers the voting behaviors of individuals on expert advisory committees at the U.S. Food and Drug Administration (FDA). Individuals on these committees sometimes have financial conflicts of interest, which may result in a principal-agent dilemma. Committee members also have institutional affiliations, a history of co-authoring relationships, and different areas of expertise, which may influence voting behavior. Using data on over 1500 uniquely identified individuals at 110 new product meetings over a seven-year period, I find that in a simple analysis, financially conflicted individuals are 18 percent more likely to vote favorably for new medical devices, but no more likely to vote favorably for new drugs. This pattern is driven by individuals voting favorably for competitors’ products and is consistent with a regulatory setting in which conflicted individuals help “pave the way” for subsequent entrants to move swiftly through the regulatory approval process. I then describe a preliminary model of individuals’ voting behaviors which incorporates both direct conflicts of interest and peer effects. Using this framework, I find reduced form evidence that the composition of an advisory committee adds additional predictive power to a model of how individuals vote. Peer effects models suggest that at high (low) levels of in favor voting within a meeting, the simple analysis is likely to to understate (overstate) bias related to conflict of interest.
The final chapter considers drivers of regional variations in healthcare spending in the United States and is based on joint work with David Cutler, Jonathan Skinner, and David Wennberg. There is considerable controversy about the causes of regional variations in healthcare expenditures. We use a set of detailed vignettes from patient and physician surveys linked to Medicare expenditures at the level of the Hospital Referral Region to test whether patient demand-side factors or physician supply-side factors better explain regional variations in Medicare spending. We find patient demand is relatively unimportant in explaining variations. Physician organizational factors (such as peer effects) matter, but the single most important factor is physician beliefs about treatment: 36 percent of end-of-life spending, and 17 percent of U.S. health care spending, are associated with physician beliefs unsupported by clinical evidence.
