Interdisciplinary Urban Sociology of Municipal Market Refuse Collection Levy Bunching and Trader Relocation to Unserviced Peripheral Sites in Accra: A Difference-in-Differences Analysis of Levy Thresholds

Municipal market authorities in Accra have experimented with refuse collection levies whose liability schedules create sharp thresholds, and traders near those thresholds are thought to relocate to unserviced peripheral sites. Evaluating such relocation with a difference-in-differences design is complicated by the fact that treatment assignment is a deterministic function of a running variable, so the parallel-trends assumption is untestable in the neighbourhood of the threshold and the estimator's operating characteristics are unknown. Across 2000 replications per cell and cluster counts of 8, 16, 24 and 32, the analysis characterises power, minimum detectable effect and interval coverage for effects of 3, 5 and 8 percentage points. The design attains 80 per cent power at 52 clusters for a 5 percentage point effect. The results indicate that the estimator is well calibrated at larger cluster counts but that the minimum detectable effect at 8 and 16 clusters exceeds the plausible range of levy-induced relocation, which means that null findings in small market samples cannot be read as evidence of no relocation. Bunching at refuse levy thresholds should be treated as a design problem rather than an estimation problem, and it specifies the cluster counts at which a credible evaluation becomes feasible.