Notes and Replication Exercises
I have collected some notes and replication exercises in a public Github repository
I have collected some notes and replication exercises in a public Github repository
This paper provides novel evidence that health status is an important driver of consumer inertia in health insurance markets. I show this by studying a large program in Colombia at a time when two large health plans experienced a persistent and severe decline in their ability to provide adequate access to medical care, eventually facing termination by the government. The program’s design – featuring a single standardized plan offered by competing health plans, mandatory enrollment, and virtually unrestricted monthly switching – provides an ideal setting to study consumer inertia. The study has three parts. First, I provide descriptive and causal evidence of the substantial quality differences between the terminated and the non-terminated plans. Using employer-driven steering into plans as a source of exogenous variation in plan choice, I show that switching out from one of the terminated plans can increase the probability of using care by up to 23 pp. Second, I provide descriptive and causal evidence that health risk increases choice persistence. The causal evidence focuses on how chronic illness affects the probability of switching plans. Using a flexible event-study approach, I find that cancer diagnoses decrease the probability of switching plans across all plans. Importantly, this “lock-in” effect increases choice persistence among those who could benefit most from enrolling with plans that provide better access to care. Third, I show that the correlation between illness and choice persistence contributed to the plans’ collapse by worsening the average risk of their enrollee pool.
Graduate Teaching Assistant, LSE, Department of Economics, 2019
Graduate Teaching Assistant, LSE, Department of Economics, 2020
Graduate Teaching Assistant, LSE, Department of Economics, 2020
Teaching Fellow, LSE, School of Public Policy, 2021