Expertise, Regulatory Incentives, and Oversight in Health Care
Paper Session
Sunday, Jan. 4, 2026 12:30 PM - 2:15 PM (EST)
- Chair: Adam Sacarny, Columbia University
Managing Conflicts of Interest and Industry Influence in Government Advisory Committees: The Case of the FDA
Abstract
The FDA frequently convenes advisory committees consisting of outside experts to offer guidance in the review of prescription drugs. There have been long-standing concerns that some of these experts have financial ties to pharmaceutical companies, leading to biased recommendations. Using 21 years of data on FDA drug advisory committee meetings, I examine whether the industry financial ties of committee members are associated with biased voting. Using exogenous within-person variation in the scheduling of drug reviews relative to members’ industry relationships, I find that members who have ties to the drug sponsor are 15 percentage points more likely to vote in favor of the sponsor compared to members without any financial ties. In contrast, members with financial ties to competitor firms do not vote differently from those with no financial ties. This pro-sponsor bias is concentrated in votes on non-approval questions (e.g., post-approval safety deliberations), rather than on drug approval questions.I also examine the effect of changes in meeting-level conflicts of interest – specifically, the share of members with industry financial ties – on the probability that a majority (>50%) of members vote favorably for the sponsor. Using two novel instrumental variables, I find that increases in the share of members with ties to the sponsor lead to increases in the likelihood of favorable voting. By contrast, the share of members with ties to competitor firms has no effect – positive or negative -- on voting in favor of the sponsor. Taken together, these findings suggest that management of conflicts of interest on advisory committees need not require banning all industry ties. Bias from conflicts of interest can be mitigated by reducing the participation of members with ties to the sponsor, while still allowing participation of members with ties to competitors, who can offer similar relevant expertise.
First Do No Harm: The Labor Market Consequences of a Physician Disciplinary Record Authors: Amanda Y. Agan, Colleen Carey, and Gabrielle Sorresso
Abstract
State medical boards enforce care quality standards by disciplining physicians via publicly-disclosed administrative actions. We use a novel dataset of 70,000 state medical board actions in more than 40 states over the years 2008-2023 to report the first estimates of the effect of these disciplinary actions on physicians' careers. We find that 2.2% of physicians ever receive a disciplinary action over this time period. After a disciplinary action, a physician's billings to Medicare fall substantially, even for those physicians who are not subject to Medicare exclusion. We find that the declines are smaller for mild disciplinary actions (e.g., fines, continuing medical education) and larger for more severe discipline (e.g., revocations). We also observe that physicians lose hospital affiliations and begin practicing in smaller organizations. Overall, we determine that most disciplined physicians remain in the profession, although generally in lower-productivity settings with less oversight.Immigration Enforcement and Healthcare Provision: Evidence from the Home Care Industry
Abstract
In the U.S. home care industry, immigrants make up 32% of the workforce. Does tightening immigration policies distort their service quality? To answer this question, I assess the causal effects of Secure Communities, a county-based immigration enforcement program, on the quality of home care. Using establishment-level data and exploiting the staggered phase-in of Secure Communities, I found some evidence suggesting that the health outcomes of patients declined following the enforcement: Both the percentage of patients who need urgent unplanned medical care and the percentage of patients who had to be admitted to the hospital substantially increased, whereas the share of patients who get better at bathing and taking drugs significantly declined. One potential channel is that the quality of home health aide services offered by immigrant workers decreased: The probability of being employed in the home care industry declines among female undocumented immigrants. I also show such deteriorated health outcomes are not driven by a decline in the likelihood of providing such services at the facility level. Going forward, I will combine both administrative and survey data to check other relevant margins: the compositions of the workers and more measures of service quality among home care agencies.Discussant(s)
Diane Alexander
,
University of Pennsylvania
Abby Alpert
,
University of Pennsylvania
Hanming Fang
,
University of Pennsylvania
Marcella Alsan
,
Harvard University
JEL Classifications
- I1 - Health