Urban-Biased Structural Change

Abstract
Over the past few decades high-income countries have experienced a structural shift of economic activity from manufacturing towards services. Using rich administrative micro data from France, we document three stylized facts about structural change: (i) It has been urban-biased -- areas with high population density have seen a faster shift into services than less densely populated locations; (ii) the urban bias is entirely driven by exporters; (iii) large manufacturing exporters locate in small cities, while large services exporters locate in large cities. Motivated by these findings, we build and estimate an open economy model of cities and exporters to quantify the role of agglomeration forces, falling international trade costs, and sectoral productivity growth in shaping urban-biased structural change. We find that falling international trade costs and falling agglomeration benefits have led manufacturing exporters to grow in small cities and services exporters to grow in large cities, generating urban-biased structural change. The differential location choices of manufacturing and services exporters play a key role in structural change – if they had the same location preferences, both aggregate and urban-biased structural change would be substantially muted.