A Quasi-Experimental Policy Analysis of Risk Reduction in Tanzanian Transport Depot Maintenance Systems

{ "background": "Inadequate maintenance systems in transport depots present significant operational and safety risks across sub-Saharan Africa. The Tanzanian context lacks rigorous, evidence-based evaluations of policy interventions aimed at mitigating these systemic engineering risks.", "purpose and objectives": "This policy analysis aims to methodologically evaluate the efficacy of a revised depot maintenance protocol in reducing operational risk. The primary objective is to quantify the causal impact of the policy intervention using a quasi-experimental design.", "methodology": "A difference-in-differences (DiD) design was employed, comparing treatment and control depots before and after policy implementation. The core statistical model is Riskit = \β0 + \β1 (Treati \× Postt) + \γ Xit + \εit, where robust standard errors were clustered at the depot level to account for serial correlation.", "findings": "The intervention yielded a statistically significant reduction in aggregate risk scores. The DiD estimator \β1 was -0.18 (95% CI: -0.27 to -0.09), indicating an 18% reduction attributable to the policy. Thematic analysis of audit data revealed enhanced procedural compliance as a key mediating factor.", "conclusion": "The quasi-experimental approach robustly demonstrates that targeted maintenance policy reform can substantially mitigate engineering risks in depot systems. The findings provide a validated methodological framework for causal policy evaluation in infrastructure management.", "recommendations": "Policymakers should institutionalise quasi-experimental evaluations for major engineering policy rollouts. The specific maintenance protocol analysed should be adopted nationally, with continuous monitoring via the risk metric defined in this study.", "key words": "policy evaluation, difference-in-differences, infrastructure management, causal inference, maintenance engineering", "contribution statement": "This paper provides a novel application of a quasi-experimental design to isolate the causal effect of an engineering maintenance policy in a real-world African infrastructure context, generating a validated methodological template for similar