Essays in the Economics of Place Based Policies

Abstract

This dissertation contains three essays related to the economics of place-based policies. I focus on two place-based policies: economic development and natural disaster mitigation. Additionally, I propose a general method to characterize a geographic market structure that may be relevant in other empirical contexts.The first essay uses a characterization of the unobserved housing market to estimate the spillover effects of an economic development policy on residential properties within the same housing market. Using a stacked difference-in-differences approach, I find a positive impact on residential property values within the targeted areas and in spatially close areas. However, economic development activities may also affect residential properties within the same housing market. If these housing markets are not contiguous, spatial spillover effects will exclude them. I estimate a negative market spillover effect. This indicates that while the policy is effective close to the targeted area, it may draw investment that otherwise would have occurred within the same housing market. I show results on business license data to support this mechanism. Although the weighted sum of the direct and spillover effects is close to zero, the policy targets relatively disadvantaged areas, suggesting that it is redistributive.The second essay presents the methodology to estimate housing markets I use in the first essay. Over time, households or workers frequently move between homes or jobs in different geographic areas. These moves have information and can reveal the preferences of the economic agents. I use data on these moves to create a network of connected geographic areas in which two areas are connected if agents move between them. I apply a method from network theory to estimate the network's underlying group structure. With certain assumptions on the underlying model, these groups can be interpreted as housing markets (for household movement). For the empirical illustration, I show estimated housing markets in Chicago.The third essay studies the place-based policy of natural disaster mitigation, specifically studying how levees affect households' decision to purchase flood insurance. After levee construction, the risk of flood damage decreases, which should decrease flood insurance take-up. However, certain levees undergo accreditation, after which flood insurance prices decrease. This price change may induce households to increase their take-up of flood insurance. Mythili Vinnakota and I empirically study these two mechanisms, which have an ambiguous theoretical effect. Using original data collection on accreditation dates, we apply a difference-in-differences framework to study the effect of construction, accreditation, and both construction and accreditation on flood insurance take-up. Although construction decreases flood insurance take-up, accreditation increases take-up, showing that households respond to the decrease in flood insurance price. A back-of-the-envelope cost-benefit calculation indicates that levees pay for themselves after 13 years, even after accounting for the overall crowd-out of insurance.