Disasters increase rates of population morbidity and mortality, especially among people living with disabilities. Engaging in pre-disaster preparedness behaviors, such as possessing an emergency supplies kit, having a disaster plan, and participating in community disaster planning, can help people protect against the negative health consequences of disasters. Despite their heightened vulnerability to the impact of disasters, studies suggest that people living with disabilities or other health limitations are less likely to participate in a range of disaster preparedness behaviors. Research examining the pathways that explain why people with disabilities are less prepared is lacking.
Based on a review of the literature, I hypothesize that people with disabilities have social psychological propensities, that I define as lower self-efficacy and response efficacy, both of which mediate the probability they will enact preparedness behaviors. I additionally hypothesize that community-level factors, such as possessing lower neighborhood social capital and living in a community more socially vulnerable to disasters, intensifies the negative relationship between disability status and disaster preparedness. Conversely, living in a more advantaged neighborhood and one where residents are more exposed to risks, such as wildfires, reduces the negative relationship between disability and disaster preparedness.
To test these hypotheses, I utilized individual-level data from the 2013 Public Health Response to Emergent Threats Survey, a household survey conducted in 16 different communities in Los Angeles County, linked with community-level data from the Social Vulnerability Index (SVI) and the Healthy Places Index (HPI). Both the SVI and HPI are multi-component indexes that rank census tracts using publically available measures of community vulnerability and advantage, such as neighborhood, economic, and housing conditions.
Using these merged data sets, I conducted hierarchical linear regression analyses of a 10-item disaster preparedness behavior index regressed on four separate measures of disability status: self-rated health, presence of activity limitations, presence of a health problem requiring the use of special medical equipment, and considering yourself to be a person with a disability. I utilized Baron and Kenny’s product method to conduct mediation analysis for both perceived self-efficacy and response efficacy. I additionally included measures of perceived neighborhood social capital, community social vulnerability, community advantage, and wildfire risk as interaction terms in multilevel regression models to test for moderating effects. Each of the models controlled for individual age, gender, and race/ethnicity, as well as accounted for the two-level structure of the data by including a random intercept for community.
The results from the multilevel models determined that in comparison to those with excellent self-rated health, those with good (B=-0.4284, p=0.002), fair (B=-0.6936 p<0.001), or poor (B=-0.7660, p<0.001) self-rated health participated in fewer disaster preparedness behaviors, with a decreasing trend for lower ratings of health. The presence of activity limitations was additionally negatively associated with engaging in disaster preparedness behaviors (B=-0.2294, p=0.035). The results were non-significant for the two models whose main independent variables were the presence of a health problem requiring the use of special medical equipment and considering yourself a person with a disability, suggesting that these may be weaker measures of disability status. Although the results from the mediation analyses were mostly non-significant, it was determined that self-efficacy partially mediated the relationship between self-rated health and disaster preparedness (mediation coefficient=-0.1451, p<0.001). Finally, the moderation analyses established than neither perceived neighborhood social capital nor community social vulnerability were significant moderators of the focal relationship. However, living in a community that possesses a greater percentile ranking for advantage (B=0.0155, p=0.025) and one that has a higher proportion of the people at high risk for wildfires (B=0.0237, p=0.014) attenuated the negative relationship between poor self-rated health and disaster preparedness.
This dissertation addresses current gaps in the literature by elucidating the pathways that describe why people with disabilities are less prepared for disasters. Based on these findings, I recommend that preparedness programs and policy be twofold, both targeting self-efficacy as a more proximal precursor to engaging in preparedness behaviors as well as distal factors related to community advantage. Investing in programs and policies that invest in community housing and civic engagement can support preparedness behaviors, while bolstering overall health status linked to social and environmental determinants of health. Upstream place-based approaches can decrease disaster health disparities and ultimately enhance resilience to disasters among people with disabilities.
