Hurricanes represent one of the costliest natural catastrophes in the United States. Over time, although hurricane fatalities have become less of a concern, partially attributed to improved warning and weather forecasting systems in coastal counties, declining trend in loss of human life has not been accompanied by a decrease in property damage. Much property loss, has been inflicted because of increased population, rising standards of living and the consequent accumulation of wealth in these coastal areas. Given recent socio-economic developments coupled with geophysical trends of hurricane intensities, damage figures will likely grow astronomically unless actions are taken to curb development and significant improvements are made in construction standards.

In this paper we examine the effectiveness of a series of non-structural (hazard identification, public awareness, warning systems, etc.) and structural projects (retrofitting, elevation, relocation, acquisition, dams, levees, infrastructure rehabilitation, etc.) funded through the FEMA Public Assistance and Hazard Mitigation Grants Programs in terms of mitigating hurricane induced property losses. We define three types of adaptation policies (Type I, Type II and Type III) adopting and modifying types of adaptation defined by Fisher-Vanden et al. (2011). Type I adaptation includes investment in mitigation planning, warning and awareness studies. Type II adaptation covers projects designated for rehabilitation of major structures (dams, levees, shorelines stabilization) and infrastructural projects (roads, utilities and bridges). Type III adaptation refers to adaptive/responsive measures and on adaptive side covers activities such as restricting development, relocation, zoning, and retrofitting based of HMGP, while on the responsive side includes all projects implemented under Public Assistant Programs (immediate relief and clean-up).

Our results reveal that ignoring adaptation one would potentially overstate the impacts of hurricanes on property losses. We find a 1% increase in FEMA total spending on adaptation programs to imply 0.16% percent reduction in property losses, conditional on observing positive losses in affected counties. The conditional elasticity of damage associated with the projects related to disaster warning and awareness studies were estimated at around -0.13%. Similar elasticities were estimated for the type II and Type III adaptation programs. The results essentially indicate that the FEMA should be indifferent in allocating resources across competing programs, because returns per dollar spending are similar across activities. In practice, we see a large share of the FEMA money spent on ex-post assistance programs to help communities to quickly clean-up and restore critical infrastructure, while spending on ex-ante mitigation programs is capped at 15% of the total FEMA spending. Shifting resources to ex-ante measures has potential to reduce social cost of federal assistance programs ex-post.