Abstract:
We consider a discrete time mixed proportional hazard (MPH) model of duration. We prove that the baseline hazard and the frailty distribution of unobserved heterogeneity are nonparametrically identified using multiple-spell data, and use this to develop a GMM estimator of the baseline hazard. Our approach imposes no restrictions on the shape of the baseline hazard or the unobserved frailty distribution, allows for competing risks and spell-specific observable characteristics, and applies to right-censored data. The GMM specification is linear in the baseline hazard, which makes estimation and inference straightforward. We also develop tests of whether the MPH model is the data generating process.

*To join the seminar, please contact at econ.events@umich.edu