Abstract
{ "background": "Municipal infrastructure asset management systems (IAMs) in many African nations face chronic underfunding, yet rigorous evidence on the cost-effectiveness of different diagnostic and intervention approaches remains scarce. This creates significant challenges for engineering departments prioritising limited maintenance budgets.", "purpose and objectives": "This working paper presents a methodological evaluation of a randomised field trial designed to measure the cost-effectiveness of diagnostic protocols for municipal IAMs. The primary objective is to determine which diagnostic approach yields the most reliable asset condition data per unit of expenditure.", "methodology": "We conducted a clustered randomised field trial across multiple municipalities. Engineering teams were randomly assigned to implement one of three diagnostic protocols: a full physical survey, a risk-based sampling survey, or a community-participatory assessment. Cost data and diagnostic outputs were collected and analysed using a generalised linear model: $\\log(\\text{Cost}{ij}) = \\beta0 + \\beta1 \\text{Protocol}j + \\gamma X{ij} + \\epsilon{ij}$, where $i$ denotes assets, $j$ protocols, and $X$ a vector of controls. Inference is based on cluster-robust standard errors.", "findings": "The risk-based sampling protocol generated asset condition data with 95% concordance to full survey benchmarks, but at 62% of the cost. The community-participatory method, while lowest cost, showed significantly higher measurement error for technical structural parameters. The estimated cost differential between the risk-based and full survey protocols was significant at the 5% level.", "conclusion": "The trial demonstrates that methodological rigour in field diagnostics can be maintained while substantially reducing costs. Not all cost-saving diagnostic methods deliver data of sufficient quality for engineering decision-making.", "recommendations": "Municipal engineering departments should adopt structured, risk-based sampling diagnostic protocols for routine asset condition assessments. Further trials are recommended to validate these protocols for specific infrastructure types, such as bridges and drainage networks.", "key words