A Multilevel Regression Analysis of Municipal Infrastructure Asset Management Adoption in Tanzania: A Methodological Evaluation, 2000–2026

{ "background": "Municipal infrastructure asset management (IAM) is critical for sustainable urban development, yet its systematic adoption across Tanzanian local government authorities remains poorly quantified and understood from a methodological perspective.", "purpose and objectives": "This study provides a methodological evaluation of IAM adoption measurement. Its primary objective is to develop and test a multilevel regression model to analyse the determinants and rates of IAM system adoption across municipalities.", "methodology": "A longitudinal, mixed-methods design was employed, synthesising archival data, structured surveys, and expert interviews. The core analytical framework is a three-level hierarchical linear model specified as $y{ij} = \\beta{0j} + \\beta{1j}X{ij} + \\gamma{00} + \\gamma{01}Zj + u{0j} + u{1j}X{ij} + \\epsilon_{ij}$, where level-1 units are assets, level-2 are municipal departments, and level-3 are councils. Model inference uses robust standard errors clustered at the council level.", "findings": "The methodological evaluation reveals that the multilevel model significantly improves explanatory power over pooled ordinary least squares, capturing 42% of the variance in adoption scores attributable to council-level institutional factors. A key substantive result is that dedicated IAM staffing showed a strong positive association with adoption completeness ($\\beta = 0.31$, 95% CI [0.18, 0.44]).", "conclusion": "The multilevel regression approach is a robust methodological tool for analysing IAM adoption, effectively disentangling nested institutional influences. The findings confirm that adoption is predominantly driven by council-level capacity, not just asset-specific characteristics.", "recommendations": "Municipalities should prioritise establishing dedicated IAM units with trained staff. National policy should mandate standardised IAM performance reporting to facilitate comparative analysis and benchmarking.", "key words": "asset management, infrastructure, multilevel modelling, regression analysis, municipalities, governance", "contribution