Working Paper: NBER ID: w12381
Authors: Michael R. Haines; Robert A. Margo
Abstract: We use county and individual-level data from 1850 and 1860 to examine the economic impact of gaining access to a railroad. Previous studies have found that rail access was positively correlated with the value of agricultural land at a point in time, and have interpreted this correlation as evidence that rail access chiefly benefitted agricultural land owners in the manner predicted by the Hekscher-Ohlin or Von Theunen models. We use a difference-in-difference strategy, comparing changes in outcomes in counties that gained rail access in the 1850s to those that either gained access earlier or did not have access before the Civil War. \n\tMost of the estimated effects are small and the signs are not wholly consistent with either model, under the null hypothesis that agriculture was the chief beneficiary of rail access. For example, we find that rail access appears to have increased urbanization, raised the likelihood of participation in the service sector, decreased agricultural yields, and reduced the share of improved acreage in total land area, opposite to the patterns predicted by either the Heckscher-Ohlin or Von Theunen models.
Keywords: Railroads; Economic Development; Urbanization; Agricultural Yields
JEL Codes: N51; N71
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Rail access in the 1850s (L92) | Increased urbanization (R11) |
| Rail access in the 1850s (L92) | Decrease in agricultural yields (Q11) |
| Rail access in the 1850s (L92) | Reduced share of improved acreage in total land area (Q15) |
| Rail access in the 1850s (L92) | Higher land prices (R31) |