Journal Design Engineering Masthead
African Structural Engineering | 20 March 2009

Assessing Risk Reduction in Kenyan Water Treatment Systems

A Difference-in-Differences Methodological Evaluation
W, a, n, j, i, k, u, M, w, a, n, g, i, ,, K, a, m, a, u, O, c, h, i, e, n, g
Causal InferenceInfrastructure ResiliencePublic Health EngineeringQuasi-Experimental Design
Difference-in-differences model isolates causal effect of engineering interventions.
Rehabilitation programme linked to ~16% reduction in waterborne disease incidence.
Quasi-experimental design provides robust framework for post-hoc project evaluation.
Longitudinal data from control facilities is critical for causal inference.

Abstract

{ "background": "Waterborne diseases remain a significant public health challenge in many regions, with infrastructure resilience being a critical engineering concern. Systematic evaluation of structural and operational interventions in water treatment systems is required to quantify their effectiveness in mitigating public health risks.", "purpose and objectives": "This case study aims to methodologically evaluate the application of a quasi-experimental difference-in-differences (DiD) model to assess the risk reduction achieved by a major rehabilitation programme applied to a subset of water treatment facilities. The objective is to demonstrate the model's utility for isolating the causal effect of engineering interventions from secular trends.", "methodology": "A longitudinal panel dataset of facility performance and regional health outcomes was constructed. The core DiD model is specified as $Y{it} = \\beta0 + \\beta1 \\text{Treat}i + \\beta2 \\text{Post}t + \\delta (\\text{Treat}i \\times \\text{Post}t) + \\epsilon{it}$, where $Y{it}$ is the log-transformed incidence of waterborne disease. Inference is based on cluster-robust standard errors to account for serial correlation.", "findings": "The DiD estimator $\\delta$ was -0.18, indicating a statistically significant reduction in log disease incidence attributable to the intervention. The implied average treatment effect corresponds to an approximate 16% reduction in reported cases in regions served by rehabilitated facilities relative to the control group, with a 95% confidence interval of [-0.31, -0.05].", "conclusion": "The difference-in-differences approach provides a robust methodological framework for causal inference in post-hoc evaluation of engineering infrastructure projects, effectively controlling for common temporal shocks.", "recommendations": "Engineering project evaluations should incorporate quasi-experimental designs like DiD where randomised control trials are impractical. Practitioners should prioritise the collection of longitudinal data from both intervention and control facilities to enable such analyses.", "key words": "causal