- This event has passed.
Arnaud MAUREL (Duke University) – "Conditional Choice Probability Estimation of Continuous Time Job Search Models "
Time: 03:00 pm – 4:15 pm
Date: 25th of May 2021
by visio
Arnaud MAUREL (Duke University) – “Conditional Choice Probability Estimation of Continuous Time Job Search Models”
Abstract: We propose a new method to estimate continuous-time job search models. Our approach is based on an adaptation of the conditional choice probability estimation methods to a continuous-time job search environment. To do so, the proposed framework incorporates preference shocks into the search framework, resulting in a tight connection between value functions and conditional choice probabilities. Our method, relative to standard estimation methods for continuous-time job search models, yields considerable computational gains. In particular, this method makes it possible to estimate rich, possibly non-stationary, job search models without having to solve any differential equation, and in some cases even avoiding any optimization altogether. We apply our method to analyze the effect of unemployment benefit expiration on the duration of unemployment and wages using rich longitudinal data from Hungarian administrative records. (joint with P. Arcidiacono, A. Gyetvai and E. Jardim).
Organizers:
Benoît SCHMUTZ (Pôle d’économie du CREST)
Anthony STRITTMATTER (Pôle d’économie du CREST)
Sponsors:
CREST
Time: 03:00 pm – 4:15 pm
Date: 25th of May 2021
by visio
Arnaud MAUREL (Duke University) – “Conditional Choice Probability Estimation of Continuous Time Job Search Models”
Abstract: We propose a new method to estimate continuous-time job search models. Our approach is based on an adaptation of the conditional choice probability estimation methods to a continuous-time job search environment. To do so, the proposed framework incorporates preference shocks into the search framework, resulting in a tight connection between value functions and conditional choice probabilities. Our method, relative to standard estimation methods for continuous-time job search models, yields considerable computational gains. In particular, this method makes it possible to estimate rich, possibly non-stationary, job search models without having to solve any differential equation, and in some cases even avoiding any optimization altogether. We apply our method to analyze the effect of unemployment benefit expiration on the duration of unemployment and wages using rich longitudinal data from Hungarian administrative records. (joint with P. Arcidiacono, A. Gyetvai and E. Jardim).
Organizers:
Benoît SCHMUTZ (Pôle d’économie du CREST)
Anthony STRITTMATTER (Pôle d’économie du CREST)
Sponsors:
CREST