Disaggregated validation of disaster-resilience indicators using household survey data: A case study of Hong Kong

A disaster resilience index aggregates numerous observed individual indicators into a numeric value, for the purpose of gauging various communities’ disparate disaster resilience capacities as part of decision-making in resilience management. There have been abundant studies on the creation of such indices, but only a few have sought to empirically validate individual indicators’ practical efficacy in explaining disaster-related outcomes. Therefore, this study performs such disaggregated empirical validation of nine disaster-resilience indicators’ efficacy at explaining two outcome measures: the resistant capacity and recovery capacity of households in Hong Kong. It reveals that certain indicators including education, income, and place attachment can be empirically valid, but that their explanatory power varies substantially across the two outcome measures. For instance, place attachment has divergent relationships with households’ resistant and recovery capacities. The robustness of the indicators’ explanatory power is also unequal, due to the disparate effect sizes of the outcome measures and the indicators’ interdependence. Based on these findings, we provide recommendations on indicator selection and index creation that should be useful to those seeking to create parsimonious and robust sets of indicators that are explanatory of the actual resilience capacities of local communities.