Search and the City ∗

We show that search frictions can be the source of increasing returns to scale (IRS) that brings people together in big cities. We develop a model of an economy with several regions, which differ in scale. Within each region, workers have to search for a job-type that matches their skill. They face a trade-off between match quality and the cost of extended search. This trade-off differs between regions, because search is more efficient in larger regions. Then, interregional mobility and trade lead to a pattern of specialization where large regions undertake more search intensive activities. Empirical evidence for the United States corroborates the implications of the model. Search can explain about two thirds of the wage differentials between large metropoles and small cities. Princeton, the SED meeting in Stockholm, and the Theil memorial conference in Amsterdam for useful comments and suggestions on earlier versions of this paper.