Enterprise Performance Evaluation And Ranking Based On A Modified DEA Method

It is one of the core issues in economics and management research on enterprises performance, and also an issue which must be addressed to pursuit the goals of higher performance. This paper aims at methodologies research. By analysis and comparison of various performance evaluation methods and taking the data envelopment analysis(DEA) method as the theoretical tool, a modified evaluation method based on DEA is proposed. The modified method is raised based on the practical problems of enterprise performance evaluation. Most of the performance evaluation methods are for absolute results of enterprise performance. Few considered the relative performance evaluation based on inputs and outputs. In reality, most enterprises’ performance cannot be evaluated directly, and the production function cannot be defined. Therefore, enterprises are in need of an evaluation method which can solve the problem of ‘black box’. This method can help to evaluate the performance objectively and accurately,and help the managers get a clearer understanding of the enterprise running and provide a valid basis for decision making. Suggestions and comments for performance improving and resource allocation are also been given.Data Envelopment Analysis(DEA) is considered one of the most successful methods for relative efficiency evaluation by far. It can easily handle the problem of relative efficiency evaluation of enterprises with multiple inputs and multiple outputs and whose production functions can’t be defined. Based on previous research, and considering some of the practical problems of performance evaluation, a modified DEA-based method for performance evaluation and ranking is proposed. Firstly, the theories of performance evaluation methods and DEA evaluation and ranking methods are reviewed systematically, and the theories of enterprise performance evaluation are introduced, including the definition of the concept of enterprise performance evaluation, the performance evaluation factors, and the performance evaluation index and its construction requirements. Then the principles and applications of DEA method are described in detail. Based on the above theoretical knowledge, it can be concluded that DEA method can find a good solution to the ‘black box’ problem. And it is an absolute advantage in dealing with multi-input and multi-output performance evaluation. But the weights and efficiency gained through classical DEA models cannot be widely accepted because of the diversity of weights between different units. Especially in efficiency ranking, evaluation based on different weights is less accepted. Furthermore, decision-makers concern about the performance of a single decision-making unit as well as the overall consumption and output of the enterprise, namely its overall efficiency. So, when analyzing resource allocation, decision-making units can be united rather than mapped to the surface of the efficient frontier of DEA respectively. However, the need of largest overall efficiency has not yet been taken into account when scholars conducting the research of performance evaluating methods. Also, mutual evaluation between homogeneous units will affect the overall evaluation in an organization, and peer evaluation will be more fair and objective. Considering the expectations of largest overall efficiency and the acceptance of evaluating results by the majority of the decision-making units, I propose a more realistic performance evaluation method which can make the common weights accepted by the majority and meet the demand of maximum overall efficiency of decision-makers. The common weights got thought this method are unique and more stable. The balance between self-evaluation and group-evaluation is taken into account,so that the result of the evaluation is more realistic.In the modified method, the entire organization is regarded as a total virtual unit and added to the evaluating system to form a new evaluating system. Then I create and solve a new DEA model with an objective function to maximize the total efficiency. Since the weights are usually not unique when ranking through common weights, a second objective is put forward to minimize the sum of the deviation between the efficiency based on the common weights and the efficiency based on the classic DEA for all the DMUs, thus solving the problem of common weights diversity. Then the efficiency is calculated according to the unique common weights. To guarantee the balance of self-evaluation,overall evaluation and group evaluation,I introduce the cross-efficiency method and get an average cross efficiency of each unit. Finally, the efficiency based on common weights and the average cross-efficiency of each unit get their weights respectively according to the evaluation purposes,actual situation and the preference or experience of the organization’s decision-makers. The DMUs are totally ranked by the linear combination of the two sets of efficiency. After the proposing of the new approach, an example is given to prove the effectiveness of the modified method. The conclusion and future research trend and prospect are given in the end.