Google Scholar: https://scholar.google.com/citations?hl=en&user=l6dHsy4AAAAJ
I study the long-run effects of major urban improvements on existing households. I exploit family location near the site of New York's Central Park (1858-76) prior to an east-west divergence in land values, neighbor wealth, and amenities. Between 1860-70, east side property holders gained 70% more wealth than their west-side counterparts, while non-holders remained similar. Regardless of property, son mobility by 1900 and grandson mobility by 1940 are unchanged. Gains for propertied households were partially offset by reduced labor force participation. Limited social integration with wealthy newcomers may explain why improved neighborhood conditions failed to increase mobility for existing households.
Available at SSRN: https://ssrn.com/abstract=5254607
I study economies of scale on U.S. railroads using a newly digitized panel of firm-level financial and operating data from the Interstate Commerce Commission, 1888–1940. I separately identify two types of scale: economies of density (higher output on a fixed network) and economies of size (joint expansion of output and network). I instrument for output using plausibly exogenous variation in U.S. cotton and wheat yields, and I identify the effect of size using the unanticipated 1901 oil discoveries in the American Southwest, which induced a regional rail construction boom. I find increasing returns to density and constant returns to size: a 1 percent increase in output raises total costs by 0.55 percent at fixed network size, while a proportional increase in output and network length raises costs by 0.98 percent. A separate Differences-in-Differences analysis of 152 consolidations finds no cost effect, even ten or more years after the event. My results imply that U.S. railroads reached their minimum efficient scale considerably earlier than previously thought, weakening the cost-efficiency case for large firms.
Forthcoming 2026