Research

Working papers

• Retention by Design: Long-term Contracts in Health Insurance (Job Market Paper).
  Abstract

  This paper studies how regulation designed to protect consumers from rising premiums as their health worsens shapes competition and welfare when insurers have market power. In Chile, insurers commit to long-term premium schedules, while consumers remain free to switch plans. Using administrative data on enrollment, insurance plans, and medical claims, we document that healthier individuals are more likely to switch plans. Leveraging demographic differences across groups of similar plans, we find that newly introduced plans help insurers retain switchers. To quantify how consumer switching and insurers' retention strategies jointly affect prices and welfare, we develop and estimate a supply and demand model in which insurers set prices once, anticipating future enrollment, costs, and plan introductions. In counterfactuals, we find that banning new plan introductions reduces insurers' costs but raises premiums by 13%, causing more enrollees to move from private to public insurance. Banning consumer switching also reduces costs but lowers consumer welfare, as individuals lose the flexibility to adjust coverage when their health needs change. Our findings show that policies can reduce insurers' costs without making consumers better off.
• Regulated Prices, Global Capacity, and Drug Shortages – with Pierre Dubois and Gosia Majewska, 2026 (Submitted)
  Previously circulated as Drug Shortages: Evidence from France.
  Abstract

  Drug shortages are widespread. We model capacity choice by a firm supplying multiple countries under regulated prices: supply shocks leave highest-price countries fully served, middle-price countries rationed, and lowest-price countries excluded. We develop a method to measure shortages using only national sales data. Applied to monthly French pharmacy sales (2019–2024), we find a 26% average shortage probability. Higher French prices reduce both shortage probability and magnitude; higher UK prices reduce probability but raise magnitude, and higher Polish prices reduce both, consistent with the model's capacity and allocation channels. Antipsychotic counterfactuals show targeted price increases substantially reduce shortages at limited cost.
• Gender-based Pricing Ban in Health Insurance: Evidence from Chile.
  Abstract

  Regulation in health insurance aims to reduce inefficiencies while addressing redistribution across demographic groups. We study a regulation in Chile's private health insurance market that bans pricing plans by gender. While this ban may benefit some demographic groups, it can also increase adverse selection and raise prices. Using administrative data from Chile, we estimate a discrete choice model of demand for health insurance plans that includes the distribution of expected individual medical costs. We combine it with a pricing model that accounts for insurer competition. Using this model, we evaluate how insurers adjust their pricing and plan offerings in response to the regulation, and how they would adjust their prices under alternative regulatory policies. Our results show that after the policy, consumer surplus increased for women and decreased for men. Focusing only on the price effect, without accounting for the change in plan offerings, would underestimate the losses of men. These findings highlight the importance of considering both price effects and insurer responses when designing health insurance regulations, providing insights for policymakers seeking to balance efficiency and redistribution.