中国健康保险欺诈的理论分析与实证研究

发布时间：2018-07-28 20:42

【摘要】：从世界范围来看,随着健康保险业的不断发展,保险欺诈已经成为健康保险业顺利发展的主要障碍。在中国,健康保险欺诈每年都造成了一定程度的经济损失,对医保基金安全构成了巨大威胁。而且,保险欺诈日趋蔓延的最直接影响是保险定价机制发生扭曲,最终损害诚实投保者的利益;就深层次的影响来说,保险欺诈将严重损害保险经营的最大诚信原则,侵蚀保险经营的契约基础。然而,与日趋严重的保险欺诈相比,大多数保险机构还处在市场份额的争夺战中,欺诈识别的高成本和收益的外部性使反欺诈还未成为其中心工作。而且,相比国际学术界反欺诈研究的热络,我国对健康保险欺诈的研究还主要集中在定性分析,很少进行实证研究。基于此,本文拟对中国的健康保险欺诈问题进行系统研究,并尝试和开启健康保险欺诈识别的实证研究工作,进而提出有针对性的反欺诈措施。通过本研究,不仅可以在一定程度上揭示我国健康保险欺诈的内在特征与规律,提高保险机构识别欺诈风险和反欺诈的能力与水平,有利于健康保险市场的稳定和医疗保险政策的实施,而且还对我国的保险诚信建设研究与反欺诈研究具有重要的参考价值。基于对文献资料的统计分析和实地调查研究,从健康保险欺诈概念的界定出发,依据欺诈发生的时间对健康保险欺诈的表现形式进行归纳分析,即购买保单时的恶意隐情投保和重复保险行为,出险报案时的伪造损失、夸大、扩大损失和故意制造保险事故行为,索赔申请时的伪造出险时间行为。另外,从不同保险合同参与主体的角度出发,对健康保险欺诈所造成的危害及产生的原因进行深入剖析。借鉴欧美国家健康保险欺诈与反欺诈的成功经验,对我国健康保险欺诈与反欺诈问题进行深入分析,梳理总结我国在开展健康保险反欺诈工作中所存在的问题,同时分析研究国外健康保险反欺诈的成功做法及其给我们的启示。在理论研究层面,利用完全信息静态博弈与不完全信息静态博弈的有关理论方法,以医患合谋这一典型的健康保险欺诈问题为例,从博弈分析的角度剖析健康保险欺诈产生的深层次原因以及引发欺诈的关键要素,并分别从最优保险契约设计的角度和保险公司经营管理的角度研究给出健康保险反欺诈的基本思路,即在保险契约设计过程中要依据投保人有无欺诈记录设置给付差额条款;在经营管理过程中保险经营机构要达到最优化状态使其对索赔申请案件的审核成本等于保险欺诈造成的损失。在实证分析层面,从保险公司经营视角出发,对多家保险公司健康保险理赔疑似欺诈案件特征进行梳理,并以XX保险公司2010~2014年五年间的综合健康险为例,利用经验分析法总结出27个欺诈识别指标,然后运用LOGIT回归分析法将欺诈识别指标精炼到9个,即投保人年龄是否在41~50岁之间、职业是否属于中等风险职业、居住地是否为县区、投保金额大小、索赔金额大小、发生保险事故时是否履行及时通知义务、投保日期与出险日期时间差距长短、索赔时能否提供完整资料、出险事故类型是否为意外事故。在此基础上,分别验证了BP神经网络模型与LOGIT-BP神经网络模型对健康保险欺诈识别的有效性。实证分析结果表明,这两种方法均为健康保险欺诈识别的有效方法,采用BP神经网络模型和LOGIT-BP神经网络模型识别健康保险欺诈行为的正确率分别为70%和80%,另外,LOGIT-BP神经网络模型能够获得比BP神经网络模型更好的识别效果。最后,分别从技术和法律等层面上研究给出健康保险反欺诈的对策建议。即借鉴和综合国内外目前在健康保险领域实施的反欺诈措施,探讨建立覆盖投保人、保险经营机构、政府管理机构三方的利益平衡机制;探讨建立信息共享、行业协调、参保公众的反欺诈参与机制和反欺诈培训体系的政策建议。
[Abstract]:In the world, with the continuous development of health insurance industry, insurance fraud has become the main obstacle to the smooth development of health insurance industry. In China, health insurance fraud has caused a certain degree of economic loss every year, which poses a great threat to the safety of medical insurance fund. And the most direct effect of insurance fraud is the most direct effect of insurance fraud. The risk pricing mechanism is distorted and ultimately damages the interests of the honest insured. In the deep influence, insurance fraud will seriously damage the maximum honesty principle of insurance operation and erode the contract basis of insurance operation. However, most insurance institutions are still in the battle of market share, and the fraud knowledge is still in the competition of market share compared with the increasingly serious insurance fraud. The externality of high cost and income makes anti fraud not a central task. Moreover, compared with the international academic circle of anti fraud research, the study of health insurance fraud in China is mainly focused on qualitative analysis and few empirical studies. Based on this, this paper intends to systematically study the health insurance fraud in China, and Try and open the empirical research work of health insurance fraud identification, and then put forward targeted anti fraud measures. Through this study, we can not only reveal the inherent characteristics and laws of health insurance fraud in our country to a certain extent, improve the ability and level of the insurance institutions to identify fraud risks and anti fraud, and be beneficial to the health insurance market. The stability and the implementation of the medical insurance policy also have an important reference value to the research on the construction of insurance integrity and the research on anti fraud in China. Based on the statistical analysis and field investigation of the documents, the concept of health insurance fraud is defined and the forms of health insurance fraud are entered according to the time of fraud. On the basis of the inductive analysis, the malicious insure insurance and the repeated insurance behavior when buying the policy, the forgery loss, exaggeration, expansion of the loss and the intentional manufacture of insurance accident, the forgery time behavior of the claim, and the harm caused by the health insurance fraud from the angle of the different insurance contracts involved in the owner. With the successful experience of health insurance fraud and anti fraud in European and American countries, the problems of health insurance fraud and anti fraud in China are deeply analyzed, and the problems existing in China's health insurance anti fraud work are summarized and summarized, and the success of foreign health insurance anti fraud is analyzed and studied. At the theoretical level, the theoretical method of static game of complete information static game and incomplete information is used, and the typical health insurance fraud of doctor and patient conspiracy is taken as an example. From the point of view of game analysis, the deep reasons of health insurance fraud and the key factors that cause fraud are analyzed. From the angle of optimal insurance contract design and the management and management of insurance companies, the basic idea of health insurance anti fraud is given. In the course of the design of insurance contract, the difference clause should be set up according to the insured person without fraud. The audit cost of the claim case is equal to the loss caused by insurance fraud. On the empirical analysis level, from the perspective of the insurance company, the characteristics of the suspected fraud cases of several insurance companies' health insurance claims are combed, and the comprehensive health insurance of the XX insurance company in five years 2010~2014 is taken as an example, and the experience analysis method is used to sum up the 27. A fraud identification index, and then the use of LOGIT regression analysis to refine the identification of fraud indicators to 9, that is, whether the age of the insured is between the age of 41~50, whether the occupation is a medium risk occupation, whether the residence is a county, the size of the insured amount, the amount of the claim amount, and whether the timely notification obligation is fulfilled when the insurance accident is insured and the date of insurance and the date of the insurance. The difference between the date and time is long, and can the claim be complete and whether the type of accident is an accident. On this basis, the validity of the BP neural network model and the LOGIT-BP neural network model to the identification of health insurance fraud is verified. The results of the empirical analysis show that these two methods are all identified by health insurance fraud. Effective methods, the correct rates of identifying health insurance fraud by using BP neural network model and LOGIT-BP neural network model are 70% and 80% respectively. In addition, the LOGIT-BP neural network model can obtain better recognition results than BP neural network model. Finally, the health insurance anti fraud is given from the technical and legal aspects. The countermeasures are to draw on and integrate the anti fraud measures currently implemented in the field of health insurance at home and abroad, and to discuss the establishment of the balance mechanism for the three parties covering the insured, the insurance business and the government management institutions, and discussing the policy suggestions of establishing the information sharing, the industry coordination, the anti fraud participation mechanism of the public and the anti fraud training system.
【学位授予单位】：青岛大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：D924.35;F842.6

【相似文献】