I estimate the economic impact of the construction of colonial India's railroad network from 1861-1930. Using newly collected district-level data on annual output, prices and internal trade flows I find that the railroad network had the following effects: (1) Railroads caused transport costs along optimal routes to fall by 73 percent for an average shipment. (2) The lower transport costs caused by railroads significantly increased India's interregional and international trade. (3) The responsiveness of a region's agricultural prices to its own rainfall shocks fell sharply after it was connected to the railroad network, but its responsiveness to shocks in other regions on the network rose. (4) Railroads raised real agricultural income by 18 percent in the districts where they were built. I find similar results using instrumental variables based on British military and famine-prevention motives for building railroad lines; and I find no effect, for any of the above four outcomes, in a `placebo' specification that uses railroad lines that were approved and surveyed, but were never actually constructed. Finally, I estimate the structural parameters of a Ricardian trade model using data on salt prices and trade flows only. The calibrated model explains 88 percent of the real income impact of railroads, suggesting that railroads raised welfare primarily because they enabled regions to exploit their comparative advantages by trading with one another.
Donaldson, D. Railroads and the Raj: The Economic Impact of Transportation Infrastructure. Presented at ‘Economic Change Around the Indian Ocean in the Very Long Run’ Conference, University of Warwick in Venice, 22-24 July 2008. (2008) 33 pp.