Improving efficiency evaluation in the presence of ratio data: inverse non-radial Enhanced Russell Models.

This article focuses on developing novel inverse Data Envelopment Analysis (DEA) models for efficiency evaluation of decision-making units (DMUs) when data are available in ratio form. It introduces non-radial Enhanced Russell Models (ERM) tailored for ratio data to address limitations of traditional radial DEA models, particularly the neglect of slack variables that can lead to inaccurate efficiency assessments. The paper formulates new inverse non-linear ERM models to estimate optimal input and output levels that maintain preset ratio-efficiency scores, supported by theoretical proofs, illustrative examples, and a real-world case study involving medical centers in Taiwan. The proposed models enhance precision in efficiency measurement and resource allocation decisions across diverse organizational contexts, offering managers flexible and computationally tractable tools for handling ratio data in performance evaluation.