Diversity and Cooperation After Natural Disasters: Evidence from Harvey

• In Event: Intragroup Dynamics and Group Boundaries

Abstract

How do natural disasters affect cooperation in ethnically diverse settings? A well-established body of research documents the negative effects of diversity on cooperation and collective action (Easterly and Levine, 1997; Alesina et al., 1999; Alesina and La Ferrara, 2005; Putnam, 2007). A second expanding agenda shows that natural disasters can trigger changes in social norms, religiosity, time and risk preferences (Gelfand et al., 2017; Callen, 2015; Cassar et al., 2017; Bentzen, 2015; Ager and Ciccone, 2017). We contribute to these literatures by studying a specific setting --the aftermath of Hurricane Harvey, one of the most costly and deadly Hurricanes in US history (108 deaths, $125B in damages).
Using data from an original three-wave panel survey of residents in the Houston area in the aftermath of Hurricane Harvey, US Census data, and the Harris county voter file, we provide evidence that ethnic fragmentation hinders cooperation whereas segregation results in higher levels of spontaneous collective action. Further, using a conjoint experimental design, we test whether the positive effect of segregation is due to common participation in civic associations or diminished perceptions of racial threat.
Further, we go above and beyond previous studies by documenting that natural disasters have heterogeneous effects across different ethnic groups, as well as groups’ majority status and economic mobility. Our analysis reveals a novel pattern: The effect of segregation on collective action strategies is contingent on a group’s majority status --when in the majority, African Americans living in segregated places cooperate more with their neighbors, whereas they cooperate more with their extended family when in the minority. This substitution effect is absent for other groups. In turn, a group’s economic status has heterogeneous effects: Upwardly mobile African Americans living in segregated places cooperate less with their neighbors whereas upwardly mobile whites cooperate less with their families. This suggests that mobility can weaken different social ties for different groups.
Finally, we employ the Harris County voter file to study the effects of Harvey on voter turnout. Using a differences-in-differences strategy, we find that individuals living in affected areas were less prone to vote in the 2018 election, but that such effect is reversed at higher levels of ethnic fragmentation. A possible interpretation of this puzzling finding is that different forms of collective action are substitutes and, while fragmentation may hinder horizontal (peer-to-peer) collective action, it may boost vertical collective action.
Overall, our findings reveal novel patterns and constitute fertile ground for novel theorizing about the drivers of collective action in diverse settings.

Authors

• Pablo Balan, Tel Aviv University

• Pablo Martin Pinto, University Houston

• Agustin Vallejo, University of Houston