夯实基本养老保险缴费基数对于促进基本养老保险缴费应收尽收、保障参保人员的合法权益具有十分重要的作用。但是目前基于计算机技术的缴费基数的核定方法仅仅局限在使用"保底封顶"原则,不能识别较为狡猾的欺诈行为;而人工的核定手段在面对众多参保人员时往往采用抽样核查方法,很不全面。为此,本文基于参保人员"保底封顶"换算后的缴费基数,采用核密度估计的方法拟合申报的工资分布,进而推算出全体参保人员的平均工资,最终通过与当地职工平均工资的比较从宏观上衡量参保人员缴费基数申报的真实程度,为详细核查确定重点。本文使用三个典型地区的176874个基本养老保险参保人员的缴费数据开展实验,结果表明:一些地区参保人员普遍存在缴费基数不实的情况。
Verifying the validity of the payment base of the basic endowment insurance is critical in ensuring the income of the basic endowment insurance and protecting the attendee’s legal rights and interests.However,the existing computer-aided method for verifying the attendee’s payment base is only based on the "upside on cap,downside at the end"principle.This kind of method fails to detect many slickness fraud actions.Meanwhile,the manually method can only verify a small sample of large amount of attendees.Therefore,based on the conversed attendees’payment base following the"upside on cap,downside at the end"principle,the paper simulates the declared wage distribution of attendees using the kernel density estimation method.Then the average wage of the attendees can be estimated.Finally,by comparing the estimated wage with the average wage in society,the declared wage can be macroscopic verified.Thus,some highly suspected groups can be detected.The experiment adopts the real data set of 176,874 attendees belonging to three different areas of the basic endowment insurance.The experimental results show that there is an obvious phenomenon of under-reporting of payment base in some areas.