The Divided Ecowelfare State: A Network Analysis of Public-Private U.S. Disaster Relief

• In Event: Wildfire Risk, Housing, and Disaster Policy

Abstract

Introduction/Background
Climate change has intensified extreme weather, generating new social-ecological risks that demand novel ecosocial protection. In the United States, the ecosocial safety net—anchored by FEMA's Individual Assistance Program and the National Flood Insurance Program—faces mounting political pressure toward retrenchment. Building on Hacker's (2002) divided welfare state framework and critical state theory, we argue that the U.S. ecowelfare state is similarly bifurcated between public assistance and private corporate disaster philanthropy: a division through which state and capital manage competing imperatives of accumulation, legitimation, and ecoregulation.

Research Questions
We ask: (1) Is the private ecowelfare state concentrated among elite donors and recipients? (2) Does corporate philanthropy fill gaps in public assistance, or follow the same flows? (3) Are high-emitting firms more central to disaster philanthropy networks? (4) How does philanthropic distribution relate to county-level social vulnerability and damage?

Data/Methods
We compile a comparative, multi-disaster bipartite firm-to-NGO network spanning Hurricanes Dorian (2019), Ida (2021), Ian and Fiona (2022), the Hawaii Wildfires (2023), and Helene–Milton (2024). Firm-to-NGO transfers come from the U.S. Chamber of Commerce Foundation's Corporate Aid Tracker; NGO attributes from the Urban Institute 990 E-filer database; HQ-county greenhouse gas emissions from Vulcan Fossil Fuel CO₂ Emissions; public assistance outcomes (IHP, NFIP) from OpenFEMA; county property damage from the Federal Reserve Bank of New York; disaster risk from FEMA's National Risk Index; and three social vulnerability-style composites (economic, household, racial/ethnic) from ACS 5-year estimates. We estimate dyad-independent ERGMs, firm-level centrality regressions on log HQ-county CO₂ with sector and disaster-year fixed effects, and county-level two-part models with state and disaster-year fixed effects and state-clustered standard errors.

Results/Findings
Private U.S. ecowelfare is high-magnitude but extremely centralized. Recipient-side Gini coefficients exceed 0.79; the top ten NGOs capture over 93% of cash; a persistent core of 21 firms recurring in four or more events contributes 37% of all observed cash; and in five of six events, the median recipient county receives 100% of its corporate cash from a single firm. Publicly traded firms supply 78.8% of cash despite forming ties at rates indistinguishable from private firms—a two-stage elite system. ERGM and firm-panel models return a null on HQ-county carbon context: elite donor status does not reduce to carbon intensity. Two-part models show that philanthropy is spatially convergent with FEMA but functionally decoupled at the county scale, and counties with higher risk are less likely to receive any philanthropy (OR = 0.88, p < 0.001).

Conclusion/Implications
These findings operationalize the divided ecowelfare state as an observable relational infrastructure: corporate disaster philanthropy functions less as a universalistic safety net than as a bilaterally concentrated, hub-dependent system anchored in a small-donor core and a single pooling intermediary. Three implications complicate substitution narratives. Private capital has not flowed toward the protection gaps retrenchment would widen; it systematically underinvests in counties with high risk; and its structural fragility means it cannot bear the load of a retrenched federal program. As FEMA reform debates intensify, these dynamics are critical for advocating universalistic ecosocial protection.

Author

• Christopher Taylor Brown, University of California, Berkeley
  Presenting Author