A Three-Stage Data Envelopment Analysis Approach to Assessing the Performance of Educational Systems

Abstract

Assessing the efficiency of education systems is crucial for shaping evidence-based policies and ensuring sustainable human capital development. This study employs a three-stage framework to evaluate Iran's education system from 2005 to 2023, integrating both non-parametric and parametric approaches. In the first stage, Data Envelopment Analysis (DEA) is used to calculate partial efficiency scores for outputs such as primary completion, enrollment, and research productivity. In the second stage, regression analysis investigates the role of contextual variables-including unemployment, youth not in education, employment or training, adjusted savings, and digital access-in shaping efficiency outcomes. Finally, efficiency scores are adjusted and aggregated to produce a comprehensive ranking of performance trends. The findings highlight the varying influence of contextual factors across outputs and provide practical insights for policymakers to improve educational performance and resilience.