Rural women's cooperative membership thresholds and household dietary diversity in Uganda: a regression-discontinuity analysis of village savings-group eligibility rules

Village savings and loan associations in Uganda commonly impose minimum membership thresholds, yet whether crossing such a threshold improves household dietary diversity remains untested. Because no primary survey data were collected for this analysis, the paper specifies the estimand, the running variable, the bandwidth selection procedure, the forcing-variable manipulation tests and the local polynomial estimator that would be applied to a household survey with a sharp or fuzzy threshold. The design is calibrated against the DHS Program indicator 'Breastfed children 6-23 months fed 5+ food groups' (CN_IYCF_C_4FB), which was observed for 69 country-surveys across 31 African countries between 2010 and 2025, with a mean of 21.451 and a standard deviation of 8.448. The credibility of any threshold-based claim about cooperative membership and dietary diversity depends less on the regression-discontinuity estimator itself than on the institutional transparency of the eligibility rule, and it derives three testable propositions about threshold design, member selection and dietary outcomes. The contribution is a transferable identification strategy for cooperative research in data-scarce rural settings, not an estimated effect.