Problem of public-private crowd-out for insurance surveys

Crowd-out refers to the degree which expansions of eligibility for public insurance lead individuals to substitute public coverage for private coverage. Many states are considering expanding their public insurance programs to deal with the increasing number of uninsured. As a result of this expansion, enrollment in public insurance programs may grow substantially, which has raised concerns among policy makers about the extent to which public insurance is “crowding out” private insurance.

This presentation will discuss how crowd-out is most commonly measured with survey data. Measurement problems will be covered including the treatment of individuals with multiple sources of coverage and the endogeneity of income based Medicaid eligibility models. The overview of estimation techniques will cover the crowd-out models most used in the policy literature, including two and three stage instrumental variables estimators. The instrumental variables discussion will cover the construction of the most commonly used instruments.