Do all tracks lead to Montevideo? Expansion of Railways and their impact on regional productive development in Uruguay between 1870-1952

Abstract

This thesis investigates whether the expansion of Uruguay’s railway network caused sustained improvements in regional economic performance and reduced regional inequality during the pe riod of railway expansion, consolidation, and institutional transition (1870–1952). The period spans the expansion and consolidation of the agro-export model (c. 1870–1920s) and the early decades of import-substitution industrialisation (c. 1930–1952), alongside a major institutional shift in railway governance from private concessions to increasing state intervention and eventual nationalisation in 1948, with the subsequent creation of a State monopoly company in railways in 1952. I assemble a novel department-by-year dataset from railway company annual reports and state railway accounts, recovering investment flows and network characteristics. To measure multidi mensional railway endowment, I construct a composite Railway Infrastructure Index. Demand side connectivity is captured through a market-access measure based on minimum travel times across the evolving transport network. Because regional GDP per capita is observed only in benchmark years, I recover annual outcomes using modern spatiotemporal imputation to build a consistent panel for econometric analysis. The empirical strategy estimates the effect of railway exposure on (i) departmental GDP per capita levels and (ii) departmental GDP per capita relative to Montevideo (a convergence mea sure in a highly monocentric economy). Identification combines two-way fixed effects, a het erogeneous spatial Durbin framework to model spillovers, and an instrumental-variables strategy based on least-cost paths to address endogenous network placement. Results show that railway expansion raised average regional income in an economicallymean ingful way, with sustained effects over time and remaining robust to alternative definitions of market access, including monocentric access-to-Montevideo measures. By contrast, the evidence that railways systematically reduced regional inequality is weaker and more method-dependent: convergence effects, when present, are concentrated in a specific phase (1870–1887) rather than persistent across the full period. A central conclusion is heterogeneity, railways produced clear local winners and losers, suggesting that network design and initial conditions shaped the distri bution of gains even when average development effects were positive.