A decade of air-pollution sustainability performance across OECD countries: A weight-stability-benchmarked long-term MCDM evaluation

Air pollution sustainability is commonly assessed through separate energy, environmental, and health indicators, limiting the ability to evaluate multidimensional performance consistently over time. This study addresses this gap by developing a weight-stability-benchmarked, two-level hierarchical multi-criteria decision-making framework for evaluating air pollution sustainability across 38 OECD countries from 2015 to 2024. Nine sustainable development goal indicators were organized into three dimensions: energy transition (ET), environmental exposure (EE), and health system capacity and population vulnerability (HSCV). Five normalization schemes and three weighting families were benchmarked across 35 configurations. The maximum coefficient of variation (CV) across global indicator weights was used to assess admissibility, while the average CV was used to select among admissible configurations. Under the primary 10% maximum-CV threshold, C–N5, combining CRITIC weighting with composite N5 normalization, was selected and achieved the lowest average global CV (3.9%; maximum CV 8.5%). No configuration met the stricter 5% or 7.5% thresholds, while C–N5 remained the configuration with the lowest average global CV at feasible thresholds of 10% or greater. The selected weights were applied in a two-level TOPSIS model to derive annual and periodic sustainability indices. Iceland, Norway, Sweden, and Finland formed the leading group, whereas Ireland, Estonia, Türkiye, Hungary, and Czechia occupied the lowest performance tiers. No statistically significant pairwise monotonic associations were detected among ET, EE, and HSCV. The results provide a longitudinal diagnostic framework for comparing multidimensional sustainability performance, with conclusions conditional on the selected indicators, hierarchy, study period, and stability criterion.