The Short- and Long-Run Effects of Transportation Infrastructure on International Migration

This paper leverages the rise and fall of passenger trains in Mexico to (1) estimate the impact of railroad access on initial migration patterns to the United States and (2) evaluate the persistence of these patterns long after coaches and airplanes took over passenger transportation. Using an IV strategy based on least cost paths, historical border crossing records, and present-day migration surveys, I find that, in the early twentieth century, districts with railroad access had twice the migration rates of those without access. This difference is largely explained by increased population mobility toward the north, reflected in higher stage migration rates and higher shares of stage migrants. Today, municipalities with historical access to railroads continue to have about 1.8 more stage migrants per thousand inhabitants and 10 percentage points higher shares of stage migrants than those without access. I provide evidence showing that locations with historical railroad access developed migration economies over time, reflected in shelters, food courts, and health centers for migrants, which influenced the formation of migration corridors that have persisted for more than 100 years despite technological change in passenger transportation.