Essays on the Political Economy of Urbanization and Climate Change

The world has been urbanizing at an incredible pace during the last century. Meanwhile, the global rise in temperatures has led to the increased probabilities of gradual and sudden natural disasters, putting large shares of developed lands at risk. While the benefits from agglomeration economies are well documented, less is known on how local stakeholders make land-use decisions in the context of climate change. Understanding how economic agents in charge of land conversion cope with climate threats while trying to preserve urban opportunities is a paramount challenge for the next decades. This dissertation aims to shed some light on a few of the mechanisms at play, looking at spaces threatened by diverse environmental catastrophes. In this regard, the second chapter of this thesis, 'The Political Economy of Coastal Destruction,' studies the impact of political cooperation on coastal development choices, made in Spain between 1979 and 2015. We argue that political cooperation between municipal neighbors is fostered by local political alignment. We rely on a fuzzy regression discontinuity design in close elections to assess the impact of political homophily on coastal development. We show that coastal municipalities who decide on coastal development in isolation may overdevelop as they fail to internalize the positive amenity spillovers caused by land preservation. Within the first-kilometer fringe, local governments sharing their neighbors' ideology develop 63% less than otherwise similar but politically isolated governments. This effect vanishes as we consider farther distances from the coastline, suggesting that amenity spillovers are an essential driver of this result. While overdevelopment induces higher exposure to hazards when locating in disaster-prone areas, appropriate preparation can mitigate the chances of suffering from a natural catastrophe. However, mitigation measures do not only reduce but also signal the inherent risks of a location. I focus on the trade-off between risk reduction and risk disclosure in the third chapter of my thesis, 'Does media coverage affect government preparation for natural disasters?'. I demonstrate that in the absence of information circulating about local dangers, local governments, who seek to protect property values in their jurisdiction, have an incentive not to prepare to avoid signaling the latent risks to otherwise uninformed investors. To test this hypothesis, I construct an exogenous measure of newspaper coverage of storms, which is a good predictor of the number of newspaper articles published about these events. I show that conditional on being hit by a storm, a one-standard-deviation increase in my Coverage measure leads to a 54% increase in the number of mitigation projects implemented in a ZIP code. This result is primarily driven by neighborhoods with high pre-treatment levels of vacant houses, renters, and housing-units owned with a mortgage, suggesting that non-resident investors are the firsts to respond to the information shock. Considering that real estate interests could capture governments' preparation incentives, I questioned whether individuals learn from past disasters when making a development decision. In the last paper of this thesis, 'The Dynamics of Land Development around Flood Zones,' we study the land conversion response to an inundation. Exploiting a rich dataset on historical flood records in Spain, we show that new development drops at the municipal level by -14.64% in the year following an inundation, and peaks down at -26% in the sixth year. The decrease in land conversion is, on average, permanent. This outcome is primarily driven by municipalities with higher historical flood frequencies, and by floods occurring after the central government regulated constructions around flood zones, in 1986. New development neither occurs farther away from flood zones nor on the higher ground. These results could be consistent with several underlying mechanisms. In particular, if individuals do account for disaster history when making a development decision, it is puzzling to observe they prefer not to build rather than building away from the acknowledged source of dangers. We speculate that a misinterpretation of the risks caused by an availability bias, or an aversion to amenity losses, could explain this response