Out-of-Pocket Payment Timing and Antenatal Care Dropout in Tanzanian Dispensaries: A Regression-Discontinuity Estimation of Fee-Collection Schedules and Visit Continuity

Out-of-pocket payment timing at Tanzanian dispensaries has been proposed as a lever on antenatal care dropout, but the credibility of any evaluation of fee-collection schedules depends on the operating characteristics of the design used to estimate their effect. the present analysis characterises the before-after-control-impact (BACI) estimator for visit continuity under a declared variance structure, using a replication-based design with 2000 replications per cell across 8, 16, 24 and 32 clusters and effects of 3, 5 and 8 percentage points. The between-cluster standard deviation was fixed at 6.0, the residual standard deviation at 4.5, and the intraclass correlation at 0.64. Power curves, minimum detectable effect curves, coefficient plots, a scatter fit, a receiver operating characteristic and a residual diagnostic were generated from the design. The design achieved 80 per cent power at 5 percentage points with 54 clusters, and the intraclass correlation of 0.64 dominated the variance budget, so that cluster counts below 24 produced power below conventional thresholds for the smallest effect considered. The contribution is a design-based account of what BACI can and cannot identify when fee-collection timing is the exposure, and a specification of the cluster counts and effect sizes at which a dispensary-level evaluation of antenatal visit continuity would be informative.