Non-radial Measures of Efficiency and their Applications in Data Envelopment Analysis

在这个学期的项目中，我完成了五篇论文。在第一篇论文中，我证明了BCC模型中按比例回报的状态受到添加权重限制的影响，并开发了一种用于识别RTS状态的新算法。在第二篇论文中，我讨论了DEA的前沿投影问题。当涉及到具有几个不同特征的组织时，这项任务需要注意。本文对这种情况进行了探讨，并提出了一种简单实用的方法来应对这种混合商业环境下的目标设定。提出了一种综合全局与局部条件的“调整投影”方法。第三部分论述了DEA中的“超效率”问题。我通过使用我在之前的论文中提出的基于松弛的效率度量（SBM）来解决这个问题。该方法与传统的基于径向效率度量的方法不同，前者直接处理投入/产出中的松弛，而后者不考虑松弛的存在。第四篇论文是日本的一个国家项目，讨论了几个政府机构迁出东京的计划，重点讨论了选址问题。为此，我提出了一种基于AHP和DEA保证域模型相结合的达成群体决策的共识方法。最后一篇论文涉及NTT的价格上限设置。公用事业的价格设定是至关重要的社会问题。本文借鉴DEA的成本效率模型，阐述了限额是如何确定的。

I have completed five papers in this term of project. In the first paper, I demonstrated that the status of returns to scale in the BCC model is affected by adding weight restrictions and developed a new algorithm for identifying the RTS status. In the second paper, I discussed the issues of projection onto the frontiers in DEA.This task needs care when organizations with several different characteristics are involved. This paper discusses such situations and proposes a simple and practical method for coping with target setting in such mixed business environments. A method of "adjusted projection" is proposed which synthesizes a global view with local conditions. The third paper deals with the "super-efficiency" issue in DEA.I approached to this problem by using the slacks-based measure of efficiency (SBM) which I have proposed in my previous paper. This method differs from the traditional one based on the radial measure of efficiency, in that the former deals directly with slacks in inputs/outputs, while the latter does not take account of the existence of slacks. In the fourth paper, a national project of Japan, plans for the relocation of several government agencies out of Tokyo was discussed concentrating on the site selection problem. For this purpose, I proposed a consensus-making method for reaching a group decision, based on a combination of the AHP and the assurance region model of DEA.The last paper deals with the price-cap setting of NTT.Price-settings of public utilities are crucial societal matters. This paper describes how the cap was determined by referring to the cost efficiency model of DEA.