HPM Seminar Series

Estimating Bayesian Demand Models in Portfolio Market Shares—An Application in Health Insurance Marketplaces

Tsan-Yao Huang
HSRP&A PhD Student
Division of Health Policy & Management

Abstract: The number of Americans covered by health insurance has reached a historic high since the implementation of the Affordable Care Act (ACA) due to creation of Health Insurance Marketplaces. Many studies on the economic behaviors of enrollees have been using extensive micro-level data from state-based Marketplaces. Yet, relatively little is known about individuals’ decisions to federally-facilitated Marketplaces due to either inaccessible enrollment data at individual level or market shares are aggregated as a portfolio by each firm — a measurement of market share that is aggregated over multiple products offered by the same company. Motivated by the empirical difficulties, I develop a Bayesian demand model that will overcome the data limitations and has the potential to infer consumer behaviors in federally-facilitated Marketplaces. The Bayesian approach allows me to simulate a group of consumers from aggregated market shares at firm level as if I could directly observe their choice behaviors at individual level. By using the hypothetical individuals I can infer the economic behaviors of agents in those markets when data are limited to portfolio market shares. My Bayesian model will help evaluate policies to reduce the number of the uninsured, and my approach will be applicable to a wide array of economic analyses in general.

Everyone is welcome to attend this seminar. Light refreshments served.
Unfortunately, live streaming is not available.