Determinants of flood adaptation: Parametric and semiparametric assessment

Abstract

Abstract Flood adaptation measures are indispensable to offset agricultural losses emanating from the increasing frequency and intensity of floods. Therefore, it is necessary to identify key precursors of adaptation for the success of flood‐related adaptation programs. Using cross‐sectional household survey data from rice farmers of eastern Nepal, this study explores existing adaptation practices along with the factors influencing farmers' adaptation strategies. Both parametric and semiparametric forms of the binary choice model are employed to assess key determinants of flood adaptation. The results reveal that existing adaptation practices are limited to conventional seed storage and few (bio)engineering techniques. Both household and farm characteristics, including flood impact indicators, significantly influence farmers' adaptation to flooding hazards. Thus, consideration of these multiple factors in adaptation‐related programs is anticipated to cope with future floods. Prioritising crop insurance over disaster relief programs and revitalization of existing adaptation measures are suggested. Furthermore, this study demonstrates that the semiparametric approach offers a better alternative by lifting the basic assumptions of parametric models. Findings from this article could aid in developing localised adaptation strategies that would ultimately contribute to the resiliency of farming households and the flood‐prone communities as a whole.

Keywords

• adaptation
• farmers
• flood
• semi‐nonparametric estimator
• semiparametric maximum likelihood