Victims of Resilience: An Evaluation of Social Vulnerability's Applicability to Disaster Justice

John Roper^*David Casagrande^*

• Lehigh University, Bethlehem, United States

Background The Social Vulnerability Index (SVI) is one of the most widely used tools for determining how vulnerable populations are to disasters. We tested the ability of the SVI published by the Centers for Disease Control to predict population recovery within New Orleans census tracts after Hurricane Katrina in 2005. We contextualize our results within the distributive, recognition, and procedural disaster justice framework. Methods We hypothesized that populations in census tracts with more vulnerability (higher SVI scores) would be slower to return after the disaster (less resilient). Changes in population before and after the 2005 disaster were calculated using census data from 2000 and 2010. We ran linear multivariate regression models to test for relationships between SVI, flood damage, the change in population, and gentrification. Results SVI and flood damage successfully predicted whether population in census tracts recovered (ANOVA: F(2, 289) = 36.3, p < .001). The model was statistically significant, but it explained only 20.1% (R2 = 0.201) of the variation, indicating significant unexplained variance. Another regression model using SVI and flood damage successfully predicted whether census tracts would gentrify after the disaster (ANOVA: F(2, 284) = 15.69, p < .001), although variation around this linear relationship was also high (R2 = 0.10). A subset of census variables used in SVI and gentrification indices predicted population recovery better than the SVI or Gentrification indices alone (ANOVA: F(5, 292) = 257.5, p < 0.001; R2 = 0.82), with homeownership being the most important variable. Changes in SVI and gentrification between 2000 and 2010 were inversely correlated suggesting that vulnerability was replaced with gentrification after the disaster. Conclusion The SVI is useful for documenting distributive injustice when operationalized as reduced resilience. In the case of New Orleans after Hurricane Katrina, SVI did not account for historical processes like impacts of redlining on home ownership. Lower SVI values can be misleading if they result from gentrification and not improved resiliency of vulnerable populations. Correcting inequitable vulnerability requires procedural justice to overcome negative effects of historical processes like redlining or to avoid displacement of vulnerable populations by gentrification when attempting to promote resilience.