Hazard-specific pathways shape the flood protection gap in the USA

Abstract Financial flood protection gaps are often reported as a single measure, masking important differences in how protection falls short. We combine building-level estimates of rainfall, riverine and coastal flood losses with property characteristics, FEMA flood zones and National Flood Insurance Program (NFIP) records to trace how expected losses are distributed across the United States. We distinguish expected losses outside reconstructed NFIP payout (“public shortfall”) from expected covered payout above the allocated building-coverage risk-premium base (“NFIP shortfall”). Public shortfall accounts for 81% of combined-hazard expected annual loss. Estimated NFIP participation is low for rainfall and riverine exposure (16 and 20%, respectively), compared with 59% for coastal exposure. Yet coastal NFIP shortfall reaches 32% of expected loss, versus about 11% inland. Across census tracts, similar financial gaps reflect different combinations of low participation, inadequate coverage and pressure within the insured pool. Effective policy must address these drivers differently across hazards and places.