The 2010 Deepwater Horizon (DWH) oil spill was the largest in U.S. history, releasing an estimated 4.9 million barrels of oil into the Gulf of Mexico. The scale of the disaster motivated diverse stakeholders to examine the human dimensions of the spill and how communities' resilience to similar threats could be improved. This examination is needed because, as long as humans depend on extracting oil and gas for energy, coastal regions are at risk for spills. In this report, the authors explore how communities, government officials, nongovernmental organizations, businesses, and scientists can build community resilience to large oil spills. Researchers found mixed evidence of distress associated with the DWH disaster and a variety of factors that affected the nature and severity of people's experiences.

Over the past three decades there has been a rapid increase in the number of disasters occurring worldwide that affect communities, households and individuals. The increase in disasters and the associated impacts are evident in our society. The impact of disasters can have more chronic impacts generating social and economic hardship, loss of employment, dissolution of personal relationships, and the long-term decline of physical and mental health. A study was undertaken to develop an understanding of the predictors of individual social vulnerability on individuals nested within communities. The Behavioural Risk Factor Surveillance System and 14 other community level data sources were used. The model investigated the influence of parish disaster history, operational resilience and socio-economic resilience on individual social vulnerability. Methods: The research design for the study was a multilevel repeated cross-sectional design with a three level nested structure. The software package MLwiN was used to conduct the multilevel analysis using empirical Bayes Markov chain Monte Carlo (MCMC) estimation. Using a representative sample of 34,685 individuals from 2004 to 2010, nested in 56 Louisiana parishes, the trend study allowed for an understanding of the subjective and objective factors that predict individual social vulnerability. Results: In each step, the model fit improved using the DIC statistic. Overall the results indicated that there were differences between parishes and their levels of individual social vulnerability; individual social vulnerability decreased from 2004 to 2010 and several statistically significant predictors of social vulnerability were identified. Statistically significant community level predictors of individual social vulnerability were lack of educational attainment, communities with less access to a household phone, community poverty and community unemployment. A trend was detected for age. Statistically significant two-way interactions were number of disasters and total population per square mile, and number of disasters and number of physicians per 100,000 population. A moderate trend was observed for the interaction effect of age and access to a household phone. Conclusions: With the significant increase of disasters worldwide it is imperative that factors causing social vulnerability are addressed. Results indicated that communities with lower levels of social vulnerability had higher levels of education, access to communication, and lower poverty and unemployment rates. Recommendations for future research are made, with policy and practice implications discussed.