节点文献

政府绩效相关的定量评价方法及应用研究

Research and Application of Quantitative Methods Related to Government Performance

【作者】 黄勇辉

【导师】 朱金福;

【作者基本信息】 南京航空航天大学 , 管理科学与工程, 2010, 博士

【摘要】 为了应对经济全球化、市场化、信息化的世界发展趋势的挑战,各国政府大力推行绩效管理,以强化政府治理能力,提高国际竞争力和服务水平。在西方国家政府绩效评估理论与实践的影响下,自20世纪90年代中后期以来,我国一些地方政府和部门对绩效评估日益关注,并借鉴西方政府绩效评估的成功经验,引入现代绩效评估的理念、方法和技术,以各种方式来推行这一实践,取得了一定成效。但和西方发达国家相比,我国对政府绩效评估的研究和实践起步较晚,其理论与方法研究尚处于探索阶段,存在着许多有待完善的空间,“重定性,轻定量”的评价方法研究尤为突出,定性分析居多,客观的定量研究欠缺。本文以政府的绩效评估为核心,以评价方法的选取、改进和应用为主线,在推进相关理论创新的同时,注重评价方法的合理性和实用性研究,紧密联系实际以解决政府绩效评估中的实际问题。本文的主要工作可概括为以下几个方面:首先分析了政府绩效评估的研究意义,介绍了国内外研究的现状,提出了论文的研究思路和组织构架。在政府绩效评估相关的理论基础上借鉴新制度经济学的思想,从交易成本这样一个新的视角深入分析了政府绩效评估的成本构成,在制度设计和制度运行这两大方面探索有效控制评估成本的基本途径。针对地方政府绩效评价指标体系的构建与指标权重方法展开研究。首先,阐述了政府绩效指标选取的基本原则,然后列举了厦门市区级政府绩效考评体系并加以分析。借鉴平衡计分卡这一工具,把地方政府绩效的主要影响指标概括为地方建设、公民服务、内部管理、创新和学习这四个方面,构建起一个具有空间立体特征的评价模型,真正实现对地方政府绩效的全方位、多维度、深层次的立体评价。其次,把结构方程模型引入政府绩效评估领域处理评价指标的权重问题,避免了人为主观因素对指标权重确定的干扰。最后研究层次分析法在政府绩效评价中权重问题的应用,实现定量与定性分析的有效结合。针对政府绩效评价中的非确定性特性,运用不确定性系统方法展开研究。首先把模糊理论引入到政府管理领域,将政府决策中定性的模糊语义科学合理地转化为可量化的结果,通过增设加权向量,改进后的模型使得评价对象中各项影响因素的差异性显现分明,便于政府部门找出问题的根源,有针对性地提出改进措施,为政府科学决策提供依据。然后,根据赵克勤提出的集对分析理论,运用比例分析法针对确定不确定系统进行深入研究与探讨,得出了不确定性因素转化为确定性成分问题的最终结论形式。改进和创新系统综合评价方法,并应用于政府绩效评估领域中解决实际问题。在政府绩效评估领域中大量的评价都可归结为综合评价问题。首先针对政府绩效评价中多指标评价对象,利用可拓综合评价法构建关联函数,建立起多指标物元模型,并应用于城市道路交通管理可持续发展(URTASD)的综合评价。该方法结构合理,实际意义明确,评价结果能直观明确地反映出URTASD的水平。然后,从理性决策者的角度出发,分析与探索了兼顾未来发展态势的决策问题,提出了集“过去”、“现在”和“将来”三个状况于一体的动态综合决策方法。该方法综合集成了主客观信息,使评价信息更全面,评价结论更具有可信度和导向作用利用各种评价方法的自身优势,促成方法间的有机组合,以评价政府绩效评价领域中评价对象不同的特性和评价侧面。在评价方法集成方面,首先将AHP扩展到模糊环境中,通过一定的补充与完善,综合运用AHP法和三角模糊数的排序理论,提出了一个多人模糊层次分析法模型,并应用于城市综合竞争力的优选排序,评价结果表明该方法的合理性和有效性。然后针对政府危机应对能力综合评价这类高维指标体系决策问题,建立起投影寻踪分类模型,并用基于实码加速遗传算法优化其投影方向,将多维数据指标转换到低维子空间,根据投影函数值的大小评价出样本的优劣,最大限度地避免了传统评判中权重取值的人为干扰。总之,本文主要根据政府绩效评价的一般逻辑思路展开,即先从评价的指标体系建立入手,然后研究指标的权重,最后针对政府绩效的各种特性与具体的评价问题,分别采用不确定性系统方法、综合评价法以及评价方法的集成等展开研究。经过改进或创新之后的理论与方法通过与政府绩效评价实践有效结合得以合理检验。

【Abstract】 In order to adapt to world development trend of economic globalization, market, information, governments promote the implementation of performance management, strengthen the governance capability, and enhance international competitiveness and service levels. Affected by the theory and practice of the Government Performance Evaluation (GPE) in Western countries, since the mid-to late 90s of the 20th century, some of our local governments and departments have paid increasing concerns to performance evaluation, and learned from the successful experience of Western governments performance evaluation, introduced modern performance evaluation concepts, methods and techniques to implement by a variety of ways in practice, have achieved some success. However, compared with the Western developed countries, China’s GPE research and practice started late, our theories and methods exists a lot of deficiencies too, in the evaluation methods, especially the use based largely on the qualitative assessment and the lack of quantitative analysis. In this dissertation, the government’s performance evaluation regarded as the core, selection, improvement and application of evaluation methods regarded as the main line, advancing the relevant theories in the same time, close attention have been paid to the study of rationality and practicality of the evaluation methods used. The main work of this dissertation can be summarized as the following aspects:Firstly, the paper analyzes the research significance of government performance evaluation, summarizes the status of research at home and abroad, presents research ideas and organizational structure. The second, the paper describes the theoretical basis related to GPE, uses the concept of new institutional economics for reference to analyze the cost composition of GPE by transaction costs, and explores the basic approach to control effectively cost of performance evaluation from a new perspective through the system design and operation.The doctoral dissertation focuses on construction index system of local GPE and method research of index weight. First of all, by using concept of the Balanced Scorecard (BSC) for reference, the main effect indices of Local Government Performance (LGP) indicators are summarized for these four aspects about local development, the civil service, internal management, innovation and learning, and an evaluation model with a tridimensional characteristic is built up with all-round, multi-dimensional, tridimensional evaluation of LGP. Secondly, the Structural Equation Modeling (SEM) is introduced into the government field to deal with index weights so as to avoid interference of artificial and subjective factors on the weights of indicators. Finally the Analytic Hierarchy Process (AHP) is researched too, and a quantitative and qualitative analysis is achieved effectively. According to the non-deterministic features of GPE, the uncertainty system methods are studied in the paper. Based on the theory of Set Pair Analysis (SPA), Proportional Analysis Method (PAM) is used to differentiate such uncertain composition. Thus the uncertainty system is translated into the certainty one in theory. Based of original model, a weighting vector is established in the model of fuzzy evaluation. The improved model makes the difference obviously between all influential factors of evaluation object. By the new model, the government can find effortlessly the cause of the question in evaluation object, and presents objective measures for improving the government performance.The dissertation studies the comprehensive evaluation methods in the field of government management. On the basis of considering the URTASD (Urban Traffic Administration Sustainable Development), the extended Multi-factorial Evaluation method is utilized to assess the level of URTASD. A matter-element model is build up by analyzing the elements that influence URTASD.Then the multi-ingredient assessment is transformed into the single-object decision, and the results can directly show the sustainable development level of the urban road traffic administration. Based on perspective of rational decision-maker, this dissertation analyze the decision problem that give attention to future trends, and proposal dynamic decision making approach that include past, present and future status. This approach integrates the objective and subjective information. This dissertation studies the combination evaluation methods in the field of government management. Analytic Hierarchy Process (AHP) extended to the fuzzy by supplementary and improvement, Fuzzy Analytic Hierarchy Process (FAHP) forms. Finally, sequencing result about the competitiveness of multi-cities by FAHP is a more reasonable conclusion. Through applying Projection Pursuit Cluster (PPC) model combined with real coding based Accelerating Genetic Algorithm (RAGA) in the comprehensive decision-making of city emergency response capabilities, the authors turn multi-dimension data into low dimension space. Thus, the value of projection function can evaluate each item good or not. The RAGA—PPC model can furthest avoid jamming of weight in the traditional method of multi-dimension data.In short, this article are based mainly on performance evaluation carried out the general logic thinking, that is, starting with the establishment of evaluation index system, and then study the weight of indicators, the final against the various features and specific problems of the government performance, methods were used such as non-deterministic systems approachs, comprehensive evaluation methods and evaluation method of integration. Improved or innovative theories and methods get reasonable tests through effective integration with evaluation practice.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络