Methodological Evaluation and Reliability Assessment of Transport Maintenance Depot Systems in Ghana: A Difference-in-Differences Approach

"background": "Transport maintenance depots are critical infrastructure for road network reliability in developing nations, yet systematic evaluations of their operational systems are scarce. Existing assessments often lack rigorous causal inference, relying on cross-sectional data that cannot account for unobserved heterogeneity. ", "purpose and objectives": "This study aims to develop and apply a robust methodological framework for evaluating the causal impact of a systematic depot upgrade programme on system reliability within the national transport network. ", "methodology": "A quasi-experimental difference-in-differences (DiD) model is employed, using panel data from treatment and control depot groups. The core estimating equation is Y{it = \0 + \1 + \2 + \ (\) +, where Yit is the reliability index. Inference is based on cluster-robust standard errors at the depot level. ", "findings": "The intervention significantly improved the mean reliability index by 18. 7 percentage points (95% CI: 12. 4, 25. 0) relative to control depots. The parallel trends assumption was validated using pre-intervention data, and results were robust to several sensitivity checks. ", "conclusion": "The DiD approach provides a rigorous methodological framework for infrastructure system evaluation in contexts where randomised controlled trials are impractical. The significant positive effect demonstrates the efficacy of structured upgrade programmes. ", "recommendations": "Infrastructure agencies should adopt quasi-experimental evaluation designs for major capital projects. Future depot investments should prioritise the specific technical and managerial components identified as drivers of reliability gains in this study. ", "key words": "infrastructure reliability, maintenance depots, difference-in-differences, causal inference, transport engineering, Ghana", "contribution statement": "This paper provides the first application of a difference-in-differences model to isolate the causal effect of a large-scale