Power distribution networks in many developing nations are characterised by significant technical losses and unreliable supply. Systematic, quantitative evaluations of infrastructure interventions are scarce, limiting evidence-based investment and maintenance strategies. This study aims to develop and apply a robust quasi-experimental methodology to empirically measure the impact of upgraded distribution equipment on system yield within a national utility. A difference-in-differences (DiD) model was employed, analysing high-frequency operational data from treatment and control feeder groups before and after a large-scale equipment retrofit programme. The core model is Y₈ₓ = ₀ + ₁ Treatᵢ + ₂ Postₜ + (Treatᵢ Postₜ) + ₈ₓ, where captures the causal effect. Inference is based on cluster-robust standard errors. The intervention caused a statistically significant increase in average daily yield of 8. 7 percentage points (95% CI: 6. 2 to 11. 2). This improvement was primarily driven by a marked reduction in technical losses, with no evidence of heterogeneous effects across geographic regions. The applied DiD model provides a rigorous methodological framework for evaluating capital projects in power networks, confirming that targeted equipment upgrades can substantially enhance distribution efficiency. Utilities should adopt quasi-experimental evaluation designs for post-implementation project audits. Future investment planning should prioritise clusters of feeders with the highest pre-intervention loss profiles to maximise aggregate yield gains. power distribution, technical losses, causal inference, quasi-experimental design, infrastructure evaluation, developing economies This paper presents a novel application of the difference-in-differences model to isolate the causal impact of physical infrastructure upgrades on electrical distribution yield, providing a replicable analytical tool for engineers and planners.
Ssebulime et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: