A New Method for Recognizing the Input Congestion in Data Envelopment Analysis

Abstract

One of the most attractive issues in Data Envelopment Analysis (DEA) is the evaluation of congestion of Decision Making Units (DMUs), which occurs if the decreases in certain inputs lead to increases in certain outputs without worsening any other input or output and equivalently, if the increases in certain inputs lead to decreases in certain outputs without improving any other input or output. Most of the existing approaches for detecting the congestion of DMUs in the literature on DEA employ the traditional definition of congestion and consider the situation in which the inputs and outputs can only change proportionally. This study proposes a method for recognizing the congestion of each input of a selected unit under evaluation for the case when the input and output of the respective unit can change disproportionally. The potential application of the proposed method is illustrated based on a dataset related to 16 institutes of basic research of the Chinese Academy of Sciences (CAS) for the year 2010, which was also reported in the literature.