Do spatial constraints affect the job search efficiency?

Theoretical and empirical works on job search often neglect the role of unemployed environment like spatial constraints meets while searching for a job. This paper proposes a job search model where both the spatial search area and the reservation wage are assumed to be endogenous. We exploit data from a French survey conducted by Research Direction of Employment Ministry (DARES) to estimate the structural parameters of the model. First we estimate the choice of the search area, i.e. the choice between passive (receiving job offers through the local public employment agencies) and active strategies (extending the job search area using others search channels). Using a bivariate probit model, we highlight that this choice depends significantly on individual attributes and spatial constraints and that it affects job quality. Besides, the independence of irrelevant alternatives is not rejected. Secondly, using Gamma duration model we treat both individual and unobservable heterogeneity and the multiple destinations after unemployment (long-term jobs, short-term jobs and subsidised jobs). Moreover, estimations are made on sub-samples in which individuals are homogenous according to their level of diploma and their situation towards the unemployment insurance. Controlling for unobservable heterogeneity and selection bias, the econometric results show that the passive strategy is more efficient for lowgraduated people accessing a subsidised job. However, the active strategy is more efficient for high-graduated individuals accessing a full-term or a short-term job.