Rural community seed-bank governance rules and women's varietal choice in Zimbabwean wards: a regression-discontinuity analysis of seed-loan repayment thresholds

Community seed banks in Zimbabwe condition continued access to preferred varieties on loan repayment, and the point at which a household crosses from defaulter to eligible borrower is a governance rule rather than a natural fact. Because no primary data were collected for this run, the paper is a methodological and conceptual contribution: it specifies the running variable, the cutoff, the bandwidth logic, the continuity assumptions and the threats to identification that arise when seed-bank committees also allocate extension advice and social sanction. The design is illustrated against the structure of a real DHS indicator, 'Women who have a bank account' (CO_MOBB_W_BNK), drawn from 36 country-survey observations across 26 African countries between 2015 and 2025, with a mean of 15.914 and a standard deviation of 12.211. That indicator is used only to characterise the distribution of women's formal financial access in the settings where seed banks operate, not as an outcome of seed-bank rules. Repayment thresholds are a neglected institutional lever, that their effects on varietal choice are identifiable under stated conditions, and that the design reframes seed-bank governance as a question about rule design rather than about household preferences alone.