Multi-dimensional Nondiscretionary Factors in Data Envelopment Analysis: A Slack-Based Measure

Abstract

Various extension using data envelopment analysis (DEA) are available in the literature for performance measurement in public sector. The existence of multiple non- discretionary factors in DEA based public sector models leads to overestimation of the efficiency. This paper proposes a two-step procedure to handle non-discretionary factors in DEA models. In the first step, the multiple non-discretionary factors are transformed into a single artificial factor and then, using a slack-based model the performance assessment is done. To highlight the comparison, an analysis using simulated data is performed. The results show that the proposed two-stage approach is relatively better than existing approaches when we apply the slack based measure of efficiency.