Community policing councils and vigilante accountability in Kenya's informal settlements: a regression-discontinuity evaluation of civilian oversight boards and extrajudicial violence reporting

Civilian oversight boards attached to community policing councils in Kenya's informal settlements are intended to make vigilante violence visible to the state, yet the reporting channel they create is confounded with the very violence it is meant to record. the present analysis treats the problem as one of design rather than of data: it asks what a regression-discontinuity evaluation of civilian oversight boards could identify, and under what conditions, before any such evaluation is mounted. The analysis characterises the operating characteristics of a before-after-control-impact estimator under a declared variance structure, with a between-cluster standard deviation of 6.0, a residual standard deviation of 4.5 and an intraclass correlation of 0.64. Power was estimated across 2000 design cells spanning cluster counts of 8, 16, 24 and 32 and effects of 3, 5 and 8 percentage points. The design requires 52 clusters to detect a 5 percentage point effect at 80 per cent power, a requirement that exceeds every cluster count examined. The accountability literature has mis-specified the inferential problem by treating oversight boards as a treatment whose effect is estimable at the settlement level, when the intraclass correlation of 0.64 implies that settlement-level assignment cannot be rescued by adding respondents. The contribution is a bounded methodological result: a threshold rule for when civilian oversight board evaluations are worth mounting, and a specification of the reporting-observability mechanism that any such evaluation must model explicitly.