Evaluating the efficiency of Taiwan's compulsory education system: A dynamic network DEA approach incorporating dropout impacts

Abstract

This study aims to dynamically assess the production efficiency of Taiwan's compulsory education system from 2012 to 2022 by extending the network data envelopment analysis (NDEA) model into a dynamic framework. The proposed model integrates intertemporal carry-over variables and bad outputs (such as student dropouts) to address inefficiencies beyond the static input-output framework. Using the directional distance function (DDF) approach, we distinguish between retrieved and non-retrieved dropouts, emphasizing the long-term impact of non-retrieved dropouts as a carry-over variable. The empirical analysis evaluates performance disparities among various local governments and examines the influence of different local governments on overall and stagespecific educational performance. Our findings provide insights into the efficiency changes in Taiwan's compulsory education system, offering valuable guidance for policymakers to improve system performance. Additionally, this study contributes to the literature by introducing a refined dynamic NDEA framework that better captures the complexities of educational inefficiencies.