Abstract

The bivariate probit model is frequently used for estimating the effect of an endogenous binary regressor on a binary outcome variable. This paper discusses simple modifications that maintain the probit assumption for the marginal distributions while introducing non-normal dependence among the two variables using copulas. Simulation results and evidence from two applications, one on the effect of insurance status on ambulatory expenditure and one on the effect of completing high school on subsequent unemployment, show that these modified bivariate probit models work well in practice, and that they provide a viable and simple alternative to the standard bivariate probit approach.