Bourbon Royal Roads and Local Development in Southern Italy

This thesis explores the broader urbanization effects of state-led infrastructural development. Specifically, it examines whether carriage roads built or substantially improved under Bourbon rule between 1732 and 1797 were followed by greater local demographic concentration in the continental Kingdom of Naples. It combines a twelve-year municipal demographic archive reconstructed on fixed 1991 boundaries with a historical GIS of the Calabria, Abruzzi, and Grain Road corridors, while keeping the chronology of documented road works separate from the geometry used to measure proximity. The main analysis uses a balanced six-year panel from 1648 to 1859 and compares 28 municipalities whose centroids lie within 3 kilometres of the selected network with 35 municipalities in the adjacent 3--6 kilometre band. In the fully adjusted local difference-in-differences specification, the pre-to-post increase in population density is 45.58 inhabitants per square kilometre larger in treated municipalities, with a 95\% confidence interval equal to [8.90, 82.27]. Relative to 1732, the two estimated pre-treatment coefficients are small compared with their uncertainty and do not jointly reject zero, while the coefficients for 1797, 1828, and 1859 are positive and become larger across the observed dates. The result remains positive across nearby bandwidths, when population levels are used as the outcome, and when the post-road change is allowed to vary with distance from Naples, although the latter adjustment attenuates the estimated magnitude. Reclassifying exposure with official ISTAT-2011 municipal-capital points leaves all 28 baseline treated municipalities within 3 kilometres and yields a positive estimate on the 51-municipality stable-classification sample. Shifted-road placebo estimates are small relative to their uncertainty, and one-municipality deletion estimates remain positive. A three-ring comparison places the largest relative density increase in the 0--3 kilometre band, while the 3--6 kilometre estimates are imprecise and do not show a persistently negative pattern. The spatial evidence therefore does not establish either one-for-one displacement from immediately adjacent municipalities or net population creation. Population density is interpreted as a measure of local settlement concentration rather than income, productivity, or welfare. Taken together, the historical reconstruction and the empirical evidence are consistent with a causal interpretation in which Bourbon road improvements contributed to greater demographic concentration near the selected corridors.