Railroads and Growth in Prussia

© Wikimedia Commons


Erik Hornung

Cities that gained railroad access in the early phase of railroad-network expansion in Prussia subsequently grew an additional 1.7 percentage points faster per year than their unconnected neighbours. This is the result of an empirical study by Erik Hornung.

In the absence of more direct measures of economic development, studies of pre-modern growth usually resort to using population growth rates as for most of history these were closely associated with per capita income growth. Hornung’s study sheds new light on the role of railroads in the industrial growth process for the historical German state of Prussia using a remarkably large dataset of roughly 1,000 cities.

The results suggest that population growth due to railroad access was induced by migration, whereas railroad access had no measurable effect on birth rates. At the county level, each additional railroad station resulted in an increase in urbanisation by 1.2 percentage points over a period of 15 years. Furthermore, the author finds that railroad access had a strong positive effect on industrial firms in terms of the number of employees but not on the number of firms in a city. This suggests that firms in railroad cities might have faced increasing competition; forcing them to increase productivity by increasing the establishment size.

One of the main challenges in the assessment of benefits from investing in transport infrastructure lies in separating the part of growth that is induced by the infrastructure itself from growth that would have occurred anyway due to the fundamental characteristics of cities. Fortunately, highly detailed census data allow a comparison among those cities that were similar in size and economic development prior to the coming of the railroad and which differed only regarding their access to the railroad. Subsequently, these cities grew 1.7 percentage points faster annually than their economic twins.

Published:   Journal of the European Economic Association, 2015, 13(4), pp. 699–736.