Spatial takeoff in the first industrial revolution

Using the framework of Desmet and Rossi-Hansberg (forthcoming), we present a model of spatial takeoff that is calibrated using spatially-disaggregated occupational data for England in c. 1710. The model predicts changes in the spatial distribution of agricultural and manufacturing employment which match data for c. 1817 and 1861. The model also matches a number of aggregate changes that characterise the first industrial revolution. Using counterfactual geographical distributions, we show that the initial concentration of productivity can matter for whether and when an industrial takeoff occurs. Subsidies to innovation in either sector can bring forward the date of takeoff while subsidies to the use of land by manufacturing firms can significantly delay a takeoff because it decreases spatial concentration of activity.