Simultaneous Equations with Endogenous Hazards

Suppose that one samples unemployed job seekers and observes their current job search policy and their elapsed duration of search: this paper is a study of the econometric problems that arise in using such data to investigate inter-individual and inter-temporal differences in search policy and the effect of such policies on the probability of leaving unemployment. In particular we study the implications of the distinction between sampling the stock and sampling the flow of unemployed people and we propose a particular model. The theory is illustrated by some calculations using data on ‘reservation wages’ and elapsed durations reported by two samples of British unemployed people. These calculations are consistent with the job search model, but suggest that the effect of selectivity in prolonging unemployment for our samples was small.