Hiring Practices, Duration Dependence, and Long-Term Unemployment

We develop an equilibrium search model of the labor market in which firms cannot observe worker quality prior to interviewing. Upon meeting an applicant, firms can interview workers at a cost to learn their type. Less productive workers are more likely to be turned away after an interview so that, on average, they are also unemployed for longer. In equilibrium, firms use the information contained in unemployment duration to screen workers before deciding whether or not to interview them. Thus, our model rationalizes new evidence on negative duration dependence in callback rates from Kroft et al. (2013a). We argue that the estimated duration dependence in callback rates helps to identify the degree of unobserved heterogeneity employers face in hiring process in our model. Since we can condition on unobservables within the model, we can decompose duration dependence in the job finding rate into its underlying sources: composition bias due to unobserved heterogeneity, and true duration dependence arising from employer screening.