Crowd-Out and Emergency Department Utilization

Abstract

A policy proposal to address the over-use of emergency departments by the uninsured is expanding public insurance. However, the uninsured are typically not the only recipients of expansions -- crowd-out occurs, and the two groups face radically different price changes. Using the Low Income Health Program in California, we examine how the prior insurance status of the newly eligible Medicaid population impacts emergency department usage. We use machine learning to estimate crowd-out and find utilization increased, but this increase was driven entirely by those crowded-out from private insurance. This has increased healthcare expenditures in California by $329.6 million per year.

Keywords: Crowd Out, Emergency Departments, Medicaid, Public Insurance, Machine Learning

JEL Classification: I13, I18, I38, G22