The Impact of Roads on Agglomeration

Abstract

This dissertation contains three essays on the impact of roads on agglomeration. In the firstessay, I show the US Interstate Highway System had a significant impact on the agglomeration and dispersal of different industries due to differing sensitivities to increased economic centripetal and centrifugal forces generated by a reduction in transportation costs. This study suggests the impact depends on truck-transportation utilization and backward linkages. I construct travel time estimates by representing the US highway system as a network each year during its construction utilizing a dataset of completion dates for each segment. I combine this with county level earnings data by industry to construct a measure of spatial inequality as a proxy for agglomeration. I conduct a panel regression with multiple industries across time including interaction efects, individual and time effects, utilize regional variation in the timing of highway completion to support the finding. The second essay examines the set of literatures regarding roads and the economy. Following the timeline of the development of thought, I examine and discuss some of the key works from early location theory, central place theory, urban economics, cost-benefit analysis, the new economic geography, market access, and graph theory. The third essay is an exploratory spatial monopolistic competition model in the spirit of the new economic geography. In a monopolistic competition framework firms produce different types of substitutable goods using labor and other goods as intermediate inputs, competing in price and wage over a two-dimensional space where transport is costly. Households commute to work for firms and use their income to consume each type of good with a preference for variety within each type. Firms enter and exit based on profitability, establishing clustering patterns over time based on the fixed distribution of households and the relative positions of other firms. This paper fits into the literature on production network and spatial equilibrium in that it utilizes an exogenous input-output matrix with endogenous market power to explore the locations of activities in relation to each other and agglomeration, but the spatial forces stem only from trade and does not feature any ad hoc benefits of agglomeration or household migration. I find that by altering the space of transport costs roads influence the patterns of regional activity by facilitating competition, and that industries with no linkages exhibit more dispersion.