In Chapter 1, I investigate whether physical capital anchors the spatial distribution of economic activity and how capital destruction affects local economic activity in the short and the long term. I investigate these questions by examining the 1975 frost that damaged coffee trees in the Brazilian state of Paraná. I find that capital destruction can have persistent effects on the spatial distribution of labor. I identify the effects of capital destruction by comparing the evolution of local economies that had different coffee tree densities before the frost. The frost resulted in a large and persistent displacement of agricultural workers. There was a permanent decline of the local coffee industry after the one-time shock, which reduced labor demand as coffee requires more labor than alternative land uses. The results are consistent with a model in which agglomeration economies within the coffee sector lead to multiple spatial equilibria and in which persistent capital investments keep the spatial distribution of the coffee industry from changing. In Chapter 2, I examine to what extent the geographic distribution of pre-Columbian societies determined the location of New World cities. I provide evidence that some modern cities in southern Brazil concentrate around a pre-colonial trail. Historical accounts suggest that this indigenous path was an important factor in explaining the location of the first European settlements, so the concentration is suggestive of persistence through path dependence. To separate the causal effects of the path from the effects of any geographic fundamentals that could correlate with it, I construct a counterfactual by using a region where European (Spanish) settlements were abandoned after a 17-century slave raid. I show that proximity to the indigenous path is associated with higher population density and urbanization in a region that was never abandoned, but no such association exists in the area the Spanish abandoned. In Chapter 3, I study how the gradual construction of transportation infrastructure affect the distribution of economic activity across the many sites it serves. I present a model of urban emergence and growth along an expanding railroad. The emergence of new towns depends on competition with other towns along the railroad, which implies that towns that are railroad endpoints for longer become persistently larger. I explore over 100 years of railroad expansion in Brazil to document this key prediction of the model. Each additional year that a municipality was a railroad endpoint is associated with a municipality urban population that is 0.083 log points larger in 2010. To support the conclusion that the association reflects the causal relation the model predicts, I confirm other testable predictions of the model, examine a historical case study, and use an instrumental variable approach based on the expansion rate of Brazilian railroads.

Creator

• Barsanetti, Bruno

DOI

• https://doi.org/10.21985/n2-18qs-b849

Subject

• Geography
• Economics

Language

• en

Alternate Identifier

• http://dissertations.umi.com/northwestern:15052
• etdadmin_upload_734725

Keyword

• Persistence
• Agglomeration economies
• Economic geography
• Economic development
• Brazil

Date created

• 2020-01-01

Resource type

• Dissertation

Rights statement

• In Copyright