Cities are divided into local governments responsible for local commuting infrastructure that is used by both their residents and outsiders. In this paper, I study how metropolitan fragmentation affects the provision of commuting infrastructure and the distribution of economic activity. I develop a quantitative spatial model in which municipalities compete for residents and workers by investing in commuting infrastructure to maximize net land value in their jurisdictions. In equilibrium, relative to a central metropolitan planner, municipalities underinvest in areas near their boundaries and overinvest in core areas away from the boundary. Infrastructure investment in fragmented cities results in higher cross-jurisdiction commuting costs, more dispersed employment, and more polycentric patterns of economic activity. Estimating the model using data from Santiago, Chile, I find substantial gains from centralizing investment decisions. Centralization increases aggregate infrastructure investment and population. More importantly, for a given amount of investment, centralization yields large welfare gains due solely to more efficient infrastructure allocation.
Presented at: CEP-Warwick Junior trade workshop
Leveraging a rich dataset that covers the universe of loans in Chile, we document that variation in the cost of capital across cities is substantial and larger than differences in wages between cities. We then establish that less competition between banks is associated with higher interest rates at the city level, while a larger pool of firms in the city is associated with lower interest rates. We are building a quantitative model that can account for these facts. We emphasize the role of competition between banks at the city level and how the geographic presence of a bank in cities with high deposits determines the marginal cost of raising deposits at the bank level and, indirectly, on the interest rate they charge.
Presented at (by coauthor or myself): 12th European Meeting of the Urban Economics Association
This paper explores the role of local social capital and within-household production on geographic mobility. We start by showing that lower-income households are less geographically mobile than their higher-income counterparts. Zooming in on childcare needs, we document that lower-income families rely more on local friends and relatives for their childcare needs rather than using market providers. Further, households living close to their local social networks are less likely to move, with lower income households showing the most negative effects. We propose a dynamic model of households' joint location and childcare production and estimate it by matching key moments in the data. Then, we quantify how this local social capital mechanism accounts for the patterns present in the data. Finally, we study the recently proposed American Families Plan, which heavily subsidizes market-based child care for lower-income households.
Presented at (by coauthor or myself): 12th European Meeting of the Urban Economics Association