Rural intergenerational land-transfer timing and heirs' out-migration in Kenyan sub-counties: a regression-discontinuity analysis of succession-threshold ages

Because no observational data are available, the study characterises the operating characteristics of the RD design under a declared variance structure rather than estimating a substantive effect. The analysis specifies a sharp RD estimator around statutory or customary succession thresholds, derives the identifying assumptions, and evaluates power and minimum detectable effect sizes across a design grid of 8, 16, 24 and 32 clusters, effect sizes of 3, 5 and 8 percentage points, and 2,000 replications per cell. Under a between-cluster standard deviation of 6.0, a residual standard deviation of 4.5 and an intraclass correlation of 0.64, the design requires 54 clusters to detect a 5-percentage-point effect at 80% power. The RD estimand is local to heirs whose succession age falls near the threshold, that the intraclass correlation of 0.64 severely constrains cluster-level inference, and that the design is best understood as a template for future empirical work rather than a completed causal analysis. The contribution is a disciplined methodological specification that clarifies what such a study can and cannot identify before data collection begins.