Efficiency Decomposition with Enhancing Russell Measure in Data Envelopment Analysis

Abstract

Problem statement: Most studies about the application of Data Envelopment Analysis (DEA) follow the concept of radial and non-radial efficiency measures. Because of some difficulties with the measure of weights, Russell measure utilizes the mean of all input/output efficiency measurements as the efficiency score of a specific business unit. However, this may render bias. Approach: This study provided a reasonable mechanism in the determination of user-specified weights for inputs to enhance the Russell measure which avoids the computational and interpretative difficulties. Results: Additionally, it can assist the decision maker in determining the relative importance of factors to improve the discrimination level of business unit efficiencies. Conclusion: The major advantages of the proposed technique are that it can not only provide a reasonable mechanism in the determination of user specified weights but also enhance the Russell measure. In addition, it can assist the decision maker in determining the relative importance of factors to improve the discrimination level of DMUs efficiencies and ensure the optimal scores of DMUs as the CCR efficiency.