8283 Bunche Hall
Los Angeles, CA 90095
Keywords: Latin America, Comparative Latin America
My research has been aimed at creating a tight connection between econometrics and economic theory, relaxing, at the same time, the parametric restrictions that one typically imposes when estimating a model or using the model for prediction.
The methods use structural economic models to estimate and test properties of unobserved nonparametric functions and distributions. The estimators and tests exploit observations on individuals' choices, prices, and equilibrium outcomes generated by the unobserved functions. My most recent work has developed methods to estimate the values and nonlinear effects of unobserved variables such as tastes, abilities, and productivity.
The estimates obtained from these methods can be used to predict choices and equilibrium outcomes when the structures, or some of the functions and distributions, change. They can also be used to evaluate the costs and benefits of structural changes for particular individuals or firms, and to determine what conclusions from parametric methods do not depend on their parametric assumptions.