Rural polygynous household composition and daughters' school-continuation decisions in Senegalese communities: a sex-ratio shock instrumental-variable analysis of matrimonial-market exposure

The central identification problem is that household structure is endogenous to unobserved preferences and to the anticipated returns to daughters' schooling; a naive comparison of polygynous and monogamous households therefore conflates composition with selection. The paper specifies an instrumental-variable design in which community-level sex-ratio imbalances, generated by sex-differentiated migration and mortality, shift the matrimonial-market exposure of daughters in polygynous households without directly determining their schooling. The framework is developed as a proposed design rather than an executed estimation: no household survey data are analysed here, and no empirical estimates are reported. To anchor the design in measurable quantities, the paper draws on the DHS Program indicator 'First-year contraceptive discontinuation rate due to method failure' (FP_DISR_W_PRG), covering 28 African countries and 74 country-survey observations from 1992 to 2025, with a mean of 2.585 and a standard deviation of 1.493. These values are used to characterise the demographic and reproductive-health context in which matrimonial-market exposure operates. The contribution is a specification of the exclusion restriction, the first-stage relationship and the plausible threats to identification, together with a set of derived propositions that future empirical work can test.