Retornos crescentes urbano-industriais e spillovers espaciais: evidências a partir da taxa salarial no estado de São Paulo

Abstract

Agglomeration of people and activities in space is a common fact of life in the modern world, despite intrinsic inefficiencies that an agglomeration process may generate. For firms, these spatial concentration phenomena are justified by positive externalities that enhance local productivity and compensate for inefficiencies. In the literature, such externalities are generated by both urbanization economies (JACOBS, 1969) and localization economies (MARSHALL, 1890). This study intends to find evidence for such externalities in the urban development of cities in São Paulo state, which is polarized by the largest economic agglomeration in Brazil, the Metropolitan Area of the City of São Paulo, using two techniques based on data of the 2000 Brazilian Census. The first one, Explanatory Spatial Data Analysis, is a non-parametric approach focusing on inquiry of spatial spillovers of urban attributes that in the context of agglomeration economies enhances the productivity of contiguous areas. The main result shows a great economic area around São Paulo city with high productive complementarities that reinforce the relevance of strategic localization and transport infrastructure for regional development. The second one, using data on extended urban areas, focuses on urbanization economies embedded in relationship of productivity levels and the density of manufacturing activities through a wage equation approach To measure these external economies, this study uses a model developed by FINGLETON (2003) that, under typical Urban Economics hypothesis and assuming that variations on wage rates between cities expresses variations on productivity, makes possible estimations of urban increasing returns and spatial spillovers of efficiency levels of production. Using the instrumental variables technique in 2SLS estimations, the main results are in accordance to Jacobs theory since they suggest a positive relationship between productivity and the density of industrial activities. Furthermore, it shows the magnitude of spillover effects in neighboring areas and its decay with distance.