Estimating long-term K-12 student homelessness after a catastrophic flood disaster

Abstract

Despite efforts to end homelessness in the United States, student homelessness is gradually growing over the past decade. Homelessness creates physical and psychological disadvantages for students and often disrupts school access. Research suggests that students who experience prolonged dislocation and school disruption after a disaster are primarily from low-income households and under-resourced areas. This study develops a framework to predict post-disaster trajectories for kindergarten through high school (K-12) students faced with a major disaster; the framework includes an estimation on the households with children who recover and those who experience long-term homelessness. Using the National Center for Education Statistics school attendance boundaries, residential housing inventory, and U.S. Census data, the framework first identifies students within school boundaries and links schools to students to housing. The framework then estimates dislocation induced by the disaster scenario and tracks the stage of post-disaster housing for each dislocated student. The recovery of dislocated students is predicted using a multi-state Markov chain model, which captures the sequences that households transition through the four stages of post-disaster housing (i.e., emergency shelter, temporary shelter, temporary housing, and permanent housing) based on the social vulnerability of the household. Finally, the framework predicts the number of students experiencing long-term homelessness and maps the students back to their pre-disaster school. The proposed framework is exemplified for the case of Hurricane Matthew-induced flooding in Lumberton, North Carolina. Findings highlight the disparate outcomes households with children face after major disasters and can be used to aid decision-making to reduce future disaster impacts on students.