Matt Shum, Caltech
"A Closed-Form Estimator for Dynamic Discrete Choice Models: Assessing Taxicab Drivers' Dynamic Labor Supply"
Abstract: We propose a new closed-form estimator for dynamic discrete choice models in a semiparametric setting, in which the per-period utility functions are known up to a finite number of parameters, but the distribution of utility shocks is left unspecified. Compared to other existing estimators for these models, our estimator requires no iterative nonlinear optimization, rendering issues of starting values or convergence criteria irrelevant. Using our approach, we estimate an optimal stopping model for taxicab drivers’ labor supply decisions. Our results show that, once the inherent dynamic in taxicab drivers’ work decisions are accounted for, it is possible to obtain “nonstandard” (i.e., negative) wage elasticities from a model in which drivers’ utility functions do not have any explicitly “nonstandard” features, such as reference dependence or loss aversion.