One airport, four efficiencies: Comparing efficiency dimensions in airport benchmarking

Evaluating airport efficiency through a single analytical dimension, whether operational throughput, airline response, service quality, or environmental sustainability, risks misrepresenting overall infrastructure performance. This study examines dimensional divergence by estimating four Data Envelopment Analysis (DEA) models for Brazil's twenty-nine largest airports from 2000 to 2024: an operational infrastructure-conversion model, a tactical airline-response model, a two-stage network model accounting for flight delays and cancellations, and a socio-environmental network model incorporating pollutant emissions for the 2018–2023 window. Applying within-year Spearman rank correlations, Kendall's coefficient of concordance, and a percentile-based dissonance index evaluated via a double bootstrap that re-estimates the DEA frontiers within each replication, with airport and year fixed effects, we demonstrate that single-dimension rankings conceal fundamental operational trade-offs. While operational throughput and tactical airline responses exhibit close alignment (mean Spearman ρ = 0.85), correlation with social effectiveness drops to 0.55, and environmental agreement falls to 0.40. Furthermore, system concordance is highly state-dependent, collapsing from about 0.86 in pre-crisis years to 0.52–0.63 during 2020–2021, when regulatory clarity is most critical. Crucially, dimensional dissonance is associated with physical scale rather than with ownership structure; governance regimes fail to predict dissonance even under within-airport fixed effects, whereas large capacity-constrained hubs are associated with systematic penalties when operational delays and cancellations are priced. Ultimately, unidimensional benchmarking introduces predictable structural bias into performance assessment, which indicates that regulatory incentives and corporate benchmarking need multi-model architectures to capture hub saturation and regional network realities accurately.