This paper evaluates discrete choice models as tools for analyzing the effects of tax and transfer policies on labor supply. An advantage of discrete choice models is that they distinguish changes in labor force participation from changes in the hours of work. Such models can also capture the heterogeneity in labor supply response among and within demographic subpopulations.
In this paper, two types of discrete choice models are estimated using cross-sectional data from the Current Population Survey: quadratic models and quasi-linear models. The models are then used to simulate the labor supply responses to hypothetical policy changes. The resulting labor supply responses are then compared with findings in the empirical literature. The paper particularly focuses on a hypothetical policy in which the earned income tax credit is contracted by about the same magnitude as it was expanded in the 1990s. We find a broad consistency between the labor supply responses in the simulations and in the empirical literature, which suggests that the omitted variable bias in estimating labor supply with cross-sectional data may not be as large as is sometimes suspected.
