UC Santa Barbara

Every year, the federal government provides billions of dollars in disaster assistance to homeowners in the form of subsidized insurance, post-disaster relief, and mitigation grants. Such transfer programs can have negative resource effects if they encourage moral hazard, either for the recipients of the transfers or for the elected officials who work with bureaucracies to oversee their allocation. This dissertation explores how federal disaster programs influence the choices of landowners, as well as how the incentives of elected officials influence the allocation of disaster-related expenditures. In doing so I am interested how policy that intends to decrease the long-term local costs of disasters can hinder adaptation and increase the asset base at risk.

The first chapter explores how federal subsidized disaster insurance influences land-use decisions. I use satellite land cover data spanning 1973-2000 and historical flood maps to measure how rate subsidies offered by the National Flood Insurance Program in the 1970s induced land to be converted to developed use, increasing the exposure of capital and households to flood risk. Using ordinary least squares and instrumental variables strategies, I find that a year of subsidy availability in the 1970s had a positive and growing effect on the probability of development for inland floodplains over time, suggesting that induced stocks of housing acted as coordination devices for new development funds in later years. For coastal floodplains, I find that subsidy availability increased development probability in the short-term, but overall development growth rates in the region tempered the overall legacy of the subsidies. Calculations focusing on the Mississippi River Basin reveal an extra year of subsidy eligibility in the early stages of the NFIP increased expected flood costs by approximately $250 million dollars per year for the entire Basin, measured at the year 2000.

In the second chapter, joint with Sahaab Bader Sheikh, we test for the presence of tactical redistribution in the allocation of post-disaster relief through the Federal Emergency Management Agency, a process that the President has strong discretion over.

We find strong evidence that House electoral competition and Representative party alignment influence the amount of relief going to a region. We then calculate the resulting aggregate distortions from a baseline "politically-neutral" allocation and analyze which constituencies generally benefit from tactical redistribution. Relief packages increase by $450,000-$900,000 for zip codes in House districts with incumbent Representatives aligned with the President. Relief packages also increase for zip codes in House districts with incumbent Representatives unaligned with the President if the district is more competitive. We find that zip codes with more white, older homeowners tend to benefit from electoral influences, while more urban, nonwhite zip codes do not.

In the third chapter, joint with Sahaab Bader Sheikh, we investigate how elected federal officials at different levels influence the bureaucracies that allocate disaster-related expenditures. Using hazard mitigation grants from the Federal Emergency Management Agency from 1997-2020, we examine how the allocation of grants changes when the agency moves from being independent with direct Congressional oversight to being subsumed into the larger Department of Homeland Security in 2003. The restructuring represents an expansion of executive power over the operations of FEMA at the expense of Congressional influence. We that prior to the restructuring in 2003, Representatives successfully divert mitigation funds to their own constituencies to the order of 50%-150% of the median federal contribution per zip code. In addition, during this period they are also successful at using coalitions with Representatives within their state to secure hazard mitigation funds for other districts. Diverted funds through direct subcommittee membership and coalitions represent 7.2% of the total HMA budget. The 2003 restructuring of FEMA nullifies the benefits of both direct subcommittee membership

and coalitions, showing how the expansion of executive oversight results in the preferences of the President dictating the allocation of grants at the expense of Congressional preferences.