Benchmarking Croatian airports: A data-driven approach through DEA window analysis

Abstract

This study benchmarks the efficiency of seven certified international airports in Croatia using Data Envelopment Analysis (DEA) with a window framework under variable returns to scale, explicitly addressing the challenge of a small number of decision-making units through dynamic window specification. Efficiency scores were calculated across overlapping three-year periods from 2019-2023, allowing performance dynamics to be captured during both crisis and recovery phases. The results reveal substantial heterogeneity across airports. Osijek Airport emerged as the most consistently efficient, outperforming larger hubs such as Zagreb and Split, while Zadar and Rijeka demonstrated notable efficiency improvements over time. In contrast, Dubrovnik Airport recorded persistently low efficiency despite traffic recovery, suggesting structural cost rigidity and seasonal vulnerability. These findings highlight the operational resilience of smaller regional airports and signal performance constraints among major hubs. By integrating traffic trends with efficiency outcomes, the study shows that rising passenger volumes alone do not ensure improved economic performance. The results provide a timely, evidence-based framework for airport managers and policymakers to enhance strategic planning, improve resource allocation, and strengthen the financial sustainability and competitiveness of Croatia’s air transport sector.