Key points are not available for this paper at this time.
Social vulnerability (SV) describes how socioeconomic characteristics influence individuals' or communities' ability to prepare for, withstand, and recover from hazardous events. Composite indices, which include multiple indicators to represent SV, are used across national, state, and local policy levels to guide decision makers in allocating limited resources. While several established SV indices (SVIs) exist for the United States, custom-made SVIs have been published across disciplines. Whether using a widely established SVI or custom-made, comprehensive documentation of the choices and assumptions made during the modeling process is often lacking. Often invisible to end-users these decisions have significant implications for SVI output, especially when applied to local site selection. Consequently, they influence resource allocation and our scientific understanding of the determinants, drivers, and consequences of disaster vulnerability. We addressed this issue by documenting and comparing the methodological choices, underlying assumptions, and their implications across multiple phases of SVI modeling for three existing indices. We provide guidance to help navigate the choices and assumptions through data selection and preparation (Phase 1), index construction (Phase 2), and validation and application (Phase 3). We argue that this process—a decision cascade in which methodological choices and assumptions influence subsequent phases of SVI modeling—explains differences in SVI scores. To demonstrate the implications based on our decision cascade framework, we applied our methodology to the selection of local study sites in Southeast Texas, where three SVIs diverged. Local knowledge was required to interpret divergent SVI results and inform site selection.
Rosenheim et al. (Mon,) studied this question.