Path Dependence in Economic Geography: Theory and Evidence from Post Roads

Talk by Sebastian Ottinger (CERGE-EI) as part of the Research Seminar Series of the IOS Economics Department.

Do history or geography determine the location and size of cities? We offer a novel perspective on this question to reconcile conflicting views in the theoretical and empirical literature. First, we develop a spatial model of urban systems and provide a novel characterization of its equilibrium in a general geographic setting. In the model, agents with identical preferences choose where to live and work, generating an endogenous commuting partition and allowing economic activity to concentrate in a subset of locations. We show that increasing returns can produce multiple equilibria in city locations, even when the city-size distribution remains unique. Then, we propose a new identification strategy to test for path dependence in urban location. Our empirical setting is the yams system of the former Russian Empire, where post stations were placed at regular horse-changing intervals across sparsely populated areas for logistical rather than economic reasons. Combining newly digitized data from the 1777 Roadbook with modern satellite night-light data, we find that locations within the historical stopping range are significantly brighter today than nearby locations along the same routes. This persistence of quasi-random city seeds into stable modern settlements provides direct empirical validation for the multiple spatial equilibria generated by our framework. In ongoing work, we use this setting to test the model’s secondary prediction: whether fixing city locations uniquely determines their relative sizes.