Job Search in Thick Markets: Evidence from Italy

We analyze empirically the effects of both urban and industrial agglomeration on men’s and women’s search behavior and on the efficiency of matching. The analysis is based on the Italian Labor Force Survey micro-data, which covers 520 randomly drawn Local Labor Market Areas (66 per cent of the total) over the four quarters of 2002. We compute transition probabilities from non-employment to employment by jointly estimating the probability of searching and the probability of finding a job conditional on having searched, and we test whether these are affected by population density, size and land area, industry localization, labor pooling and family network size. Preliminary results indicate that labor pooling and the size of family networks always increase hazard rates, for both men and women. The positive externalities generated by population size and density often appear only beyond a certain threshold (e.g. 80-90th percentile of the distribution), and those related to industry agglomeration only in the industrial districts with a higher share of small manufacturing firms