Hurricane disaster impacts, vulnerability and adaptation: Evidence from US coastal economy.

摘要

My dissertation research focuses on understanding the costs (direct and indirect) of North-Atlantic hurricanes on the US coastal economy and the role adaptation plays in fostering economic recovery and local resilience.;The first part of the dissertation investigates the effectiveness of various public adaptation measures provided under hazard mitigation and public assistance programs of the Federal Emergency Management Agency (FEMA) and local governments in relation to direct costs of disasters, given by property losses. Results indicate that wealth and economic exposure, together with infrastructure and socio-economic vulnerability, are the primary drivers for high costs from hurricanes in the United States. While major hurricanes can bring immense hazard to exposed counties, results show that counties with prior land-falling hurricanes are more resilient to disaster shocks. Clear evidence has been found about the importance of regulatory-based loss mitigation strategies as exhibited by improved building codes and effectiveness of enforcement. Results suggest that where to build (zoning, land-use planning, etc.) is a significant policy direction when complemented by how to build (building codes, retrofitting, etc.) policies. While non-structural and regulatory based projects are found to be extremely efficient in term of mitigating property losses, results also indicate that public protective measures may not always exhibit loss-reducing features and can, on the contrary, exacerbate costs. This is seen throughout by loss inducing features of major structural and infrastructural projects implemented by FEMA in hurricane impacted counties.;The second part of the dissertation focuses on understanding local labor market impacts of hurricane disasters and the role adaptation plays in mitigating these higher order impacts. Employing synthetic control methodology allows identifying not only the effects of the 2005 hurricanes on county quarterly employment and per worker earnings, but also investigating time to recovery to a hypothetical "no-hurricane" employment level, that was forgone because of devastations. Significant negative effects of hurricanes on employment are found in counties at the extreme end of damage distribution. Opportunity loss of employment due to the hurricanes is estimated at around 7.3% for these counties and shock is found to be very persistent, no complete recovery even 7 years after the 2005 hurricanes. Results show that the traded sector employment contracts by 6%, while the local services sector employment drops by 5-8%. Results indicate that the construction sector employment booms for only 5.5 years after the devastation, nonetheless this boom is not sufficient to spur employment in other local services sectors. We find large local multipliers of adverse shocks, one job lost in the traded sector implies around 7.13 jobs lost in the local services sector. Results indicate that where long-term mitigation programs are implemented by FEMA counties are more resilient to disaster shocks. Counties, where flood hazard and non-structural mitigation projects are initiated by local governments, exhibit similar responses. Nonetheless, significant evidence of two sources of moral hazard problems are found, one stems from the provision and generosity of disaster related public assistance spending, while another originates from unemployment insurance payments.;Interestingly, the results from part I of the dissertation and those from part II are contradictory in that they indicate different sources of moral hazard. Specifically, when disaster impacts were proxied by property losses, we found that the FEMA Public Assistance programs promoted smooth community recovery, while major structural and infrastructural projects financed under the FEMA Mitigation program exacerbated losses and thus created moral hazard issues. On the contrary, when the impact is investigated in relation to local labor markets, results indicate that mitigation programs are more effective and moral hazard stems from public assistance programs and generous unemployment insurance payments. The significance of these finding is that it emphasizes the relativity of effectiveness of adaptation programs. There is no unified solution to address and cope with consequences of disasters; a solution provided in a certain context may create inefficiencies in another context. Continuous rebuilding of major structures and infrastructures, in the context of direct impacts of disasters, seem to perversely impact and limit individuals' adaptive behavior to protect their private properties. While in the context of labor markets, an individual's decision involves labor supply and subsequently public assistance transfers aftermath of hurricanes and unemployment insurance payments are found to be behavior distorting.;The dissertation is organized as follows. In chapter 1 effects of adaptation measures are investigated in relation to property losses of hurricanes. Chapter 2 deals with local labor market effects of hurricanes and reviews existing literature, presents conceptual framework and estimation strategy. Chapter 3 presents results and discussions related to hurricane impacts on county quarterly employment and per worker earnings, adaptation effects of higher order impacts of hurricanes as well as spillover effects of disasters. (Abstract shortened by UMI.).