Automation and Agglomeration

We develop an economic geography model that integrates the constructed capital framework by Baldwin (1999. “Agglomeration and Endogenous Capital.” European Economic Review 43 (2): 253–80) with a task-based approach to explore automation and spatial agglomeration. First, we show that automated firms have an incentive to agglomerate in locations with large market demand. Consequently, they tend to locate in the central region instead of the periphery when trade freeness is high. Second, as the number of automated tasks increases, the agglomeration strengthens due to capital accumulation and scale economies driven by an increased demand for capital. Finally, since our model features demand- and production-side circular causalities, we derive a critical substitution threshold to determine which forces dominate the other.