Risk, poverty or politics? The determinants of subnational public spending allocation for adaptive disaster risk reduction in Bangladesh

•Unique subnational analysis to understand rationale of adaptive disaster risk reduction spending allocation in Bangladesh.•Dataset for 483 sub-districts tracking various social protection programs’ for FY 2010–11 to 2013–14 had been utilized.•Flood hazard risk and socio-economic vulnerability are found positively correlated equally consistent robust determinants.•Political affiliation and urban proximity consistently failed to attract intra-regional ADRR funding allocation.•Public spending for ADRR can be useful intervention tool to other DRR programs e.g. insurance or broader social transfers. We examine the directly observable determinants of sub-national (central to local) public spending allocations for disaster risk reduction and climate adaptation in Bangladesh, a country with a very high exposure to weather risk. We use a comprehensive dataset for the 483 sub-districts (Upazilas) in Bangladesh, tracking disaster risk reduction and adaptation funding provided to each sub-district by the central government during fiscal years’ 2010–11 to 2013–14, disaggregated by the various types of social protection programs. We assess to what extent the primary determinants of such funding flows—such as current hazard risk, socio-economic vulnerability, and political economy considerations—contribute to these funding allocation decisions. We find that flood hazard risk and socio-economic vulnerability are both positively correlated with the sub-district fiscal allocations. We find that political factors do not seem to significantly correlate with these allocations and neither does proximity to the centres of Dhaka and Chittagong. Public spending for adaptive disaster risk reduction, as investigated here, can be a useful complementary intervention tool to other DRR programs, such as insurance or broader social transfers, provided that it is allocated rationally. Broadly, this appears to be the case in Bangladesh. We leave the measuring of the relative efficacy and efficiency of each financing tool for future work.