Peer-evaluation in centrally managed systems

A centrally managed system (CMS) is typically a system operating under a central management who controls the activity of all decision making units (DMUs) to enhance the overall performance of the system. An important distinguishing feature of CMS, pertinent to the efficiency evaluation, is that only one strong supporting hyperplane of the production possibility set (PPS), the one that the aggregated unit is projected on, matters. As a result the ordinary definition of cross-efficiency which is based on projection on different supporting hyperplane of PPS, renders itself inapplicable. We introduce the concept of peer-evaluation in CMS and define the individual cross-efficiency score of the DMU under evaluation by aggregating the individual self-evaluated efficiency and the peer-evaluated individual efficiencies obtained using the most favorable CRA plans of other DMUs. These cross-efficiency scores may not be unique due to the presence of several most favorable CRA plans for the DMU under evaluation. To address this issue, we propose several secondary goal models in both aggressive and benevolent perspectives. The effectiveness of the proposed models is examined using a real dataset.