June 11, 2019, 15:30–17:00
Room MS 001
Econometrics and Empirical Economics Seminar
Abstract
We derive conditions for identification of sophisticated, quasi-hyperbolic time preferences in a finite horizon, dynamic discrete choice model under a set of economically motivated exclusion restrictions. Identification is reduced to characterizing of the zero set of two bivariate polynomial moment conditions. The number of discount function parameters in the identified set is bounded by known features of the data distribution. We show that though the discount function parameters are identified, it seems hard to precisely estimate these parameters compared to the geometric discount factor.