DEA Methods for Evaluating Non-Homogeneous DMUs Using Known Internal Structures

In most applications of DEA with internal structures in the literature, the models are designed to evaluate the performance of homogeneous Decision Making Units (DMUs). In many instances, however, DMUs operating in a similar technological environment may have different internal structures. In such situations, DMUs are non-homogeneous. In the presented paper, we propose a one-step DEA method for evaluating non-homogeneous DMUs based on their real internal structures and allow for the specification of shared and dedicated inputs and outputs. An illustrative application of the method to a sample of financial holding companies in Taiwan is given. We believe tthat our method leads to a more realistic evaluation and support managers to focus on the sources that contribute to the observed inefficiency.