A Quasi-Experimental Design for Cost-Effectiveness Diagnostics of Water Treatment Systems in Rwanda (2000–2026)

{ "background": "Evaluating the long-term cost-effectiveness of water treatment infrastructure in low-resource settings remains a significant methodological challenge, particularly for isolating the causal impact of specific interventions from confounding factors.", "purpose and objectives": "This working paper presents a novel quasi-experimental design to diagnose the cost-effectiveness of water treatment systems. The primary objective is to provide a robust methodological framework for engineers and policymakers to assess the financial and operational efficiency of such infrastructure.", "methodology": "The proposed design employs a difference-in-differences (DiD) approach, leveraging phased implementation across districts. The core statistical model is Ydt = \β0 + \β1 dt + \β2 dt + \β3 ( \× )dt + \γ Xdt + \εdt, where Ydt is the cost-effectiveness ratio for district d at time t. Inference relies on cluster-robust standard errors at the district level.", "findings": "This is a methodological paper; no empirical results from the full study period are yet available. Preliminary analysis of a pilot subset indicates a strong positive direction for the key interaction coefficient \β3, suggesting the intervention improves cost-effectiveness. The associated 95% confidence interval, however, is wide, underscoring the need for the full longitudinal dataset.", "conclusion": "The developed quasi-experimental framework provides a viable and rigorous method for causal inference in infrastructure performance evaluation, moving beyond descriptive cost-benefit analysis.", "recommendations": "We recommend the adoption of this DiD design for future engineering evaluations of public health infrastructure. Policymakers should mandate the collection of granular, time-variant operational data to facilitate such analyses.", "key words": "quasi-experimental design, cost-effectiveness, water treatment, difference-in-differences, infrastructure evaluation", "contribution statement":