Disparate household recovery outcomes in rural communities: a longitudinal analysis of Hurricane Michael's impact on the Florida Panhandle

The unequal distribution and impact of disasters in the United States are well documented but there are gaps in scholarship on recovery outcomes in rural communities. We help fill this gap with a longitudinal analysis of self-reported household recovery outcomes in the Florida Panhandle following the destructive Category 5 Hurricane Michael, which made landfall in Mexico Beach, FL, on October 10, 2018. A total of 909 survey responses were collected approximately 5 months after the hurricane made landfall in the fifteen counties designated by FEMA as eligible for both individual and public assistance. The survey was conducted via Automated Telephone Interviewing (ATI) (landlines) (n = 451) and online (n = 458). The two samples were merged using propensity score matching. Five months after Hurricane Michael made landfall, more than 40% of the respondents said they fully recovered, 21.7% mostly recovered, 23.7% somewhat recovered, and 9.1% reported not recovering. The respondents from the ATI sample were recontacted with a follow-up survey 17 months after the hurricane's landfall. The panel data consisted of 156 respondents who were matched across the two phone surveys. Most participants (64%) in the matched sample reported they had recovered, while 34% did not. The results from a fitted random-intercept logistic regression model indicated that the severity of the damage, extended power outages, income loss, and difficulties getting access to food, fuel, and basic services had highly significant inverse associations with reported positive recovery outcomes. Homeowners' insurance had a highly significant positive association with recovery. Concerns over shortages of food supplies and fuel showed inverse associations with recovery. Compared to those who received disaster assistance, the odds of recovery for those whose applications were denied or still pending were significantly lower. The results help inform disaster management planning, policies, and practices to improve disaster recovery outcomes.