A DEA Approach for Eco-Efficiency Analysis with Categorical Variable

The objective of this paper is to address a typical eco-efficiency analysis problem, in which undesirable outputs and categorical variables should be considered simultaneously. By analyzing the impacts of categorical variables on decision making units´ (DMU) desirable and undesirable outputs, a modified DEA (Data Envelopment Analysis) model is proposed. In the proposed model, a new approach of defining a reference set is developed. Based on the proposed model, the efficient input/output targets for DMU managers to improve DMUs´ efficiencies are also provided. A Chinese city case is employed to illustrate the proposed model.