Rural development savings-group repayment enforcement and women's input-supply participation in Kenya: a difference-in-differences analysis of social-collateral sanctions

Savings groups in rural Kenya enforce repayment through social collateral: the withdrawal of trust, reputation and reciprocal access that follows default. The argument proceeds in three steps. First, a conceptual model specifies the mechanism linking sanction intensity to input-supply participation through three channels: liquidity insurance, reputational signalling and group-level coordination. Second, the paper synthesises sociological and development-economics scholarship on rotating savings and credit associations, social collateral and women's market participation to derive testable propositions. Third, it specifies a difference-in-differences estimator with group and household fixed effects, defines treatment as the introduction of a written sanction schedule, and identifies the parallel-trends assumption as the design's principal threat. It argues that the sanction–participation relationship is conditional on group size, sanction observability and women's outside options, and that these conditions should be measured directly rather than assumed. The design is intended for application to household panel data from Kenyan savings-group catchment areas.