Modifying Social Vulnerability Indices to Complement Physical Exposure Risk Analyses

Abstract

Multivariate social vulnerability indices, used to compare communities’ susceptibility to adverse disaster outcomes, cannot be combined with physical exposure risk estimates without double-counting some effects and potentially obscuring risk sources. Further, the techniques used to create the indices appear to have methodological concerns, including diluting the signal of poverty in defining overburdened communities, creating conflict with some federal and state policy definitions. This research examined ways to modify social vulnerability indices to isolate the residual social risk that could serve as a multiplier on flood or other quantitative risk assessments, and thereby create an equity-weighted risk metric. We advanced this concept by reviewing the social vulnerability studies that tested the explanatory power of sociodemographic indicators on disaster outcomes, when controlling for physical exposure risk. We then suggested a new type of vulnerability index, or Equity (E) Index, that isolated the indicators with the strongest evidence that they represented residual social risk and examined their potential magnitude as multipliers on physical exposure risk. We also used multiple decision science techniques to minimize some undesirable index construction issues. Whereas the literature provided mixed evidence for all indicators, poverty, disability, and old age had the highest confidence that they amplified physical risk. Other factors such as race, ethnicity, acculturation, mobility limitations, children under five, and female single parents had an intermediate level of support. Indicators of median home value and low education were less frequently tested, and gender and percentage of renters had ambiguous signs. An initial exploration of empirical models suggested that, on average, the E index approximately doubled the effect of physical harm when assessing outcomes for the most vulnerable groups, but further research and validation are required. The intention of this research is to provide an evidence-based approach for targeting flood risk mitigation investments to the places where they will reduce the most harm. We use literature evidence to propose an updated measure of social vulnerability and create an equity weighting factor that can be multiplied by a measure of physical risk, such as population exposed to a given size flood, to estimate project benefits. The equity weighting factor increases the predicted benefits for socially vulnerable populations to reflect how the outcomes of flood prevention vary by sociodemographic factors. The approach differs from earlier methods of assessing social vulnerability because it isolates a small set of factors that have been shown to exacerbate flood risk or slow recovery, after controlling for degree of physical risk (e.g., storm size, flood depth). Our initial results suggest that social vulnerability factors can roughly double the degree of harm from floods for the most vulnerable groups, but additional evidence is needed to test this result. We also address potential biases in index construction methods to ensure equity weights align with policy goals. Ultimately, incorporating social vulnerability into flood risk management choices requires thoughtful measurement for effective and equitable outcomes.