DOI: 10.20965/jdr.2026.p0993 ISSN: 1883-8030

Exploring Crop and Flood Insurance as a Supporting Strategy Against Flood Disaster Risk: A Propensity Score Matching Analysis

Maria Angeles Catelo, Pierre Giuseppe Gilles, Tomohiro Tanaka, Miho Ohara

Smallholder farmers in flood-prone agricultural regions face severe flood risks that threaten household welfare and long-term economic stability. This study evaluates the causal impact of crop insurance on post-disaster coping mechanisms among smallholder farming households in Laguna, Philippines ( N =135). Methodologically, the study implements propensity score matching as its baseline framework, and extends the analysis using a multi-dimensional Mahalanobis distance nearest-neighbor matching algorithm ( nn =3) to address localized perfect prediction anomalies inherent in the baseline logistic covariates. Internal validity is rigorously assessed through a Rosenbaum bounds sensitivity analysis to evaluate the vulnerability of the matched pairs to hidden unobserved endogeneity. The empirical results reveal a structural hierarchy in post-disaster household coping mechanisms. Crop insurance exerts a negative and statistically significant effect on immediate food consumption shocks (ATT=-0.236, ρ=0.030), reducing the probability of severe household food security contractions by 23.61 percentage points. This welfare-stabilizing effect proves highly resilient against hidden selection bias, remaining statistically robust up to an extreme log-odds threshold (γ>2.0). Conversely, the treatment effect on reducing the probability of cuts in domestic non-food expenditures is marginally insignificant (ATT=-0.167, ρ=0.119), while the probability of post-flood debt accumulation remains statistically unaltered (ATT=0.031, p =0.753). These findings indicate that although current indemnity programs effectively serve as short-term consumption-smoothing shields that protect food security, they fail to alleviate systemic credit dependencies. Because conventional crop insurance schemes are calibrated to variable production inputs rather than comprehensive asset recovery, policy frameworks must transition toward parametric, asset-based coverage bundled with formal credit instruments to strengthen comprehensive post flood disaster coping mechanisms.