Rural bride-price inflation and delayed male household formation in Ugandan sub-counties: a cohort-exposure instrumental-variable analysis of cattle-price shocks

Bride-price inflation in rural Uganda has been widely linked to delayed marriage among men, but the causal direction is contested: cattle wealth both raises the affordability of bride-price and signals the value of a marriage, so cross-sectional associations between cattle prices and male household formation cannot be read as causal. The design treats cohorts of young men as differentially exposed to cattle-price shocks during the years in which they accumulate bride-wealth, and uses the DHS Program indicator 'Years of exposure to the risk of mortality for women' (id MM_AMRT_W_EXP) across 34 African countries and 112 country-survey observations from 1992 to 2025 as the empirical anchor for the exposure measure. The sample mean of the indicator is 139,374.598 (SD = 74,624.488), with country values ranging from 38,177.00 in Sao Tome and Principe to 433,649.00 in Nigeria. The paper specifies the instrument, states the exclusion and monotonicity conditions under which the cohort-exposure design identifies a local average treatment effect, and derives the conditions under which cattle-price shocks can be used to separate the affordability channel from the signalling channel of bride-price. The contribution is methodological and theoretical: it clarifies what a cohort-exposure IV can and cannot identify in this setting, and it specifies the data and first-stage evidence that would be required before the design could be executed. The paper concludes that the design is credible only under a narrow set of conditions that current DHS indicators do not satisfy, and it sets out the measurement agenda that would make the design feasible.