"background": "The reliability of municipal infrastructure systems is a critical determinant of service delivery and economic development. In many regions, systemic failures in water, sanitation, and road networks are prevalent, yet quantitative, system-level reliability assessments remain methodologically underdeveloped, often relying on isolated asset condition reports. ", "purpose and objectives": "This case study aims to develop and apply a novel multilevel regression framework to quantitatively evaluate the reliability of integrated municipal infrastructure systems. The objective is to identify and quantify the key technical, financial, and institutional factors influencing system-wide performance. ", "methodology": "A longitudinal case study methodology was employed, analysing panel data from municipal asset registers, financial statements, and service delivery reports. The core analytical tool was a three-level hierarchical linear model specified as Reliability{ijt = \0 + \1Xijt + uj + vt + \₈₉ₓ, where i, j, and t index assets, municipalities, and time, respectively. Robust standard errors were used for inference. ", "findings": "The analysis revealed that operational expenditure per capita had a significantly stronger positive effect on system reliability than capital expenditure alone (p < 0. 01). A key concrete result is that a 10% increase in preventative maintenance funding was associated with a 3. 2% improvement in composite reliability score (95% CI: 1. 8% to 4. 6%). Institutional governance metrics emerged as a dominant theme explaining municipality variance. ", "conclusion": "The multilevel model successfully captured the nested structure of infrastructure system data, demonstrating that reliability is driven by a complex interaction of asset-level, municipal, and temporal factors. Sole focus on physical asset condition is insufficient for predicting system performance. ", "recommendations": "Municipalities should rebalance budgets to prioritise sustained operational and maintenance funding. National policy should mandate integrated reliability reporting using multilevel modelling techniques to enable comparative benchmarking and targeted interventions
Botha et al. (Tue,) studied this question.