A Quasi-Experimental Design for Cost-Effectiveness Analysis of Power-Distribution Infrastructure in Kenya

"background": "Power-distribution infrastructure in sub-Saharan Africa faces persistent challenges in balancing capital investment with long-term reliability. Existing cost-effectiveness analyses often rely on retrospective data, lacking robust counterfactuals for new equipment deployments. ", "purpose and objectives": "This study developed and applied a novel quasi-experimental design to rigorously evaluate the cost-effectiveness of different medium-voltage transformer types deployed within a national utility's network. ", "methodology": "A difference-in-differences framework was employed, comparing operational and maintenance costs between intervention groups receiving modern amorphous-core transformers and control groups with standard silicon-steel units. The core model is C{it = \ + \1 (Treati \ Postt) + \ Xit +, where C is cost per kilometre of feeder. Robust standard errors were clustered at the substation level. ", "findings": "The intervention group demonstrated a statistically significant reduction in annualised maintenance costs of 18. 3% (95% CI: 12. 7% to 23. 9%) compared to the control, after controlling for feeder load and age. The higher capital cost of the modern transformers was offset within an estimated seven-year period. ", "conclusion": "The quasi-experimental design proved methodologically viable for infrastructure evaluation in a real-world utility context, providing stronger causal evidence than observational studies. The specific transformer technology analysed was cost-effective under prevailing conditions. ", "recommendations": "Utilities should adopt structured quasi-experimental pilots for major equipment procurements. Policymakers can mandate such evaluations in technical approval guidelines to improve investment efficacy. ", "key words": "cost-benefit analysis, difference-in-differences, electrical grid, infrastructure economics, sub-Saharan Africa", "contribution statement": "This paper provides the first application of a quasi-experimental, causal-inference framework for comparing power-distribution equipment costs in a sub-Saharan African context, yielding a replicable methodology and a definitive cost-offset period for a key