Investigating the relationships between climate hazards and spatial accessibility to microfinance using geographically-weighted regression
2019
Abstract Microfinance institutions (MFIs) in Bangladesh provide a variety of financial services to poor households that can help them cope with natural disasters (e.g. floods) and adapt to environmental changes (e.g. increasing soil salinity). However, due to the limited geographic range in which MFI branches can provide their services, households located far from a branch typically do not have access to microfinance. In this study, we measured how spatial accessibility (SA) to microfinance varied across 18 sub-districts (upazilas) of southwest Bangladesh, a region heavily affected by climate-related hazards including flooding and high soil salinity. Our objective was to identify if accessibility to microfinance was negatively affected by climate hazards due to, e.g., higher lending risks in hazard-prone areas. For this, we incorporated geospatial data sets related to flood hazard, soil salinity, population density, and transportation infrastructure as explanatory variables for regression modeling of SA. We tested both ordinary least squares (OLS) regression and geographically-weighted regression (GWR) approaches, and found that GWR was better able to predict SA. The GWR model for the SA measure “distance to nearest branch” had the strongest relationship with the explanatory variables (adjusted R 2 = 0.717), and in this model (and four of five other models tested), high flood hazard and high soil salinity were negatively correlated with accessibility to microfinance. To increase microfinance accessibility in these climate hazard-prone areas, additional funding for MFI outreach activities (e.g. utilizing national/international climate change funds), reduction of transaction costs, and further experimentation with adapting/packaging MFI services, may be required.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
36
References
11
Citations
NaN
KQI