"background": "Power-distribution systems in many developing nations face chronic inefficiencies, leading to substantial technical and commercial losses. This case study examines a national programme aimed at modernising such infrastructure, focusing on the methodological challenge of robustly evaluating the impact of equipment upgrades on system yield. ", "purpose and objectives": "This work aims to develop and apply a quasi-experimental econometric model to isolate the causal effect of a large-scale equipment replacement initiative on the technical yield of a national power-distribution network. The objective is to provide a rigorous methodological framework for impact assessment in engineering infrastructure projects. ", "methodology": "A difference-in-differences (DiD) model is employed, leveraging phased implementation across regions to create treatment and control groups. The core specification is Y{it = \0 + \1 (\) + \ + \ +, where Yit is the technical yield. Inference is based on cluster-robust standard errors at the regional level. ", "findings": "The intervention produced a statistically significant positive effect on technical yield. The DiD estimator, \1, was 8. 7 percentage points (95% CI: 6. 2, 11. 2), indicating a substantial improvement attributable to the equipment upgrades. The parallel trends assumption was validated using pre-intervention data. ", "conclusion": "The applied DiD model successfully quantified the causal impact of the infrastructure modernisation, confirming its effectiveness. The methodology provides a robust alternative to before-after comparisons, which are vulnerable to confounding temporal trends. ", "recommendations": "Future engineering evaluations of large-scale infrastructure programmes should adopt quasi-experimental designs like DiD to strengthen causal claims. Utilities should consider phased roll-outs not only for logistical purposes but also to facilitate rigorous impact measurement. ", "key words": "difference-in-differences, power distribution, technical losses, impact evaluation, quasi
Ndiaye et al. (Tue,) studied this question.