Mapping community development aid: Spatial analysis in Myanmar.

• Community-centered development in Myanmar favors areas with less vulnerable populations and more luminosity. • Among villages with similar luminosity, aid goes to villages in greater need. • Aerial luminosity, consistently available and not constrained to a specific scale, is a robust proxy for regional needs at a granular level. • High-resolution spatial measures have stronger predictive power for aid location than surveyed wealth measures. • Donors adopting developmental state theories reach out to conflict areas while donors adopting neoliberal theories focus on poverty reduction. Development aid has greater potential to alleviate poverty when targeted towards areas of concentrated need. However, the body of research examining subnational aid targeting is often constrained to state- or district-level analysis due to the lack of aid or wealth data in small regions. This study fills this gap by examining the poverty orientation of aid at the granular spatial units where the pro-poor interventions are implemented. Specifically, it explores the extent to which community-centered development (CCD) in Myanmar takes place in poor villages. This paper estimates aid presence and density, given regional poverty, by linking detailed aid location with nightlight luminosity and wealth-related data. Overall, densely aided areas tend to shine brighter and have a lower share of vulnerable populations. However, for villages with similar luminosity, aid goes to areas with fewer resources. This study also finds heterogeneity in the targeting practice among donors who adopt different ideologies. The market-oriented World Bank project supports poorer villages whereas the Korean International Cooperation Agency's project, drawing from the developmental state, includes villages close to conflict zones. Some evidence of needs-based allocation in this study runs counter to previous findings that aid favors richer states and districts. The result of this study suggests that analyzing aid at a fine-grained level with high resolution spatial data can provide a more nuanced view of aid targeting. [ABSTRACT FROM AUTHOR]