周建濤 秦炳山等
摘要:保險理賠過程中時常發(fā)生投保方欺詐現(xiàn)象,通過對實證調(diào)研的2000—2011年度258個意外、健康保險理賠訴訟樣本進行描述性統(tǒng)計和Logistic回歸,發(fā)現(xiàn)年齡41~60歲欺詐概率甚于其他年齡段,年齡41~50歲欺詐頻次最高,年齡51~60歲欺詐概率最大,工人欺詐概率甚于其他職業(yè),北京欺詐概率甚于外地,北京郊區(qū)縣欺詐概率甚于北京市區(qū);保險金額與欺詐概率正相關(guān),疾病身故保險欺詐金額最大,保險期間與欺詐概率負相關(guān),意外傷害住院醫(yī)療保險欺詐頻次最高,健康險欺詐概率甚于意外險;摔扭傷欺詐概率甚于其他傷害,交通事故機動車司機欺詐概率甚于交通事故非機動車傷害,被保險人索賠欺詐概率甚于投保人索賠,非律師參與欺詐概率甚于律師參與。
關(guān)鍵詞:欺詐概率;索賠人特征;保單信息;案件情況;健康保險;統(tǒng)計分析;保險金額;律師
中圖分類號:D90 文獻標識碼:A 文章編號:1007-2101(2014)06-0138-05
一、引言
自保險業(yè)誕生之日起,保險欺詐(Insurance Fraud)隨之而來。美國雪城大學(xué)(Syracuse University)“政府檔案交流中心”(Transactional Records Access Clearinghouse,TRAC)統(tǒng)計,2011年1-8月美國發(fā)生903起健康險欺詐訴訟,相比2010年、2006年、1991年分別增加85%、157%和822%;美國保險研究委員會(Insurance Research Council,IRC)發(fā)現(xiàn),2007年美國汽車人身傷害保險欺詐造成保險公司多支出48億美元~68億美元;英國保險人協(xié)會(Association of British Insurers,ABI)估計,保險欺詐每年給保險業(yè)帶來20億英鎊損失,未能識別的欺詐金額每年高達19億英鎊,每個投保人每年額外支付約44英鎊。廣東省是我國經(jīng)濟、保險發(fā)達地區(qū),索賠欺詐相對嚴重:東莞,2010年財產(chǎn)險賠付支出21.79億元,其中約4億元屬于欺詐索賠;深圳,2008年各保險機構(gòu)車險賠案,疑似騙賠案件976起,涉案金額超過2 000萬元,某些險種欺詐導(dǎo)致的賠款支出最高達到保費收入的5倍。
保險欺詐是否屬實,法院判決具有客觀公信力。實證調(diào)研部分保險理賠訴訟案件,深入研究欺詐動因,提煉欺詐識別因子,對于有效遏制欺詐、促進保險業(yè)可持續(xù)發(fā)展,具有重要的理論意義和實踐價值。
二、研究假設(shè)、模型設(shè)計與樣本選擇
(一)研究假設(shè)
假設(shè)1:保險欺詐受索賠人特征(性別、年齡、職業(yè)、婚姻狀況、所在地區(qū)等)影響。
假設(shè)2:保單信息(保險金額、保險費、險別、保險期間等)對索賠欺詐也有重要影響。
假設(shè)3;索賠欺詐還受案件情況(保險事故類型、索賠人類型、是否律師參與、是否存在第三方賠付等)影響。
(二)模型設(shè)計
三、描述性統(tǒng)計分析
對258個案件數(shù)據(jù)進行描述性統(tǒng)計,結(jié)果見表2。
表2顯示,男性欺詐概率(0.659)甚于女性(0.341),年齡41~50歲欺詐概率(0.437)甚于其他年齡段,工人欺詐概率(0.333)甚于其他職業(yè),已婚欺詐概率(0.881)遠甚于其他婚姻狀況,北京欺詐概率(0.570)高于外地(0.430),北京郊區(qū)縣欺詐概率(0.281)高于北京市區(qū)(0.148),外地地市欺詐概率(0.333)高于外地區(qū)縣(0.237)。
高保險金額欺詐概率(0.532)甚于低保險金額(0.491),高保險費欺詐概率(0.548)甚于低保險費(0.458),短期保險欺詐概率(0.800)甚于長期保險(0.200),健康險欺詐概率(0.993)高于意外險(0.733)。
疾病欺詐概率(0.321)高于傷害,摔扭傷欺詐概率(0.207)遠大于其它傷害,交通事故機動車司機傷害欺詐概率(0.163)略高于交通事故非機動車傷害(0.156),被保險人索賠欺詐概率(0.540)遠大于投保人索賠(0.193),非律師參與欺詐概率(0.319)大于律師參與(0.170),無第三方賠付欺詐概率(0.919)大于第三方賠付(0.081),無證人證言欺詐概率(0.852)遠甚于證人證言(0.148)。
對每個索賠案件欺詐總額(索賠總額-判決總額)及各子項欺詐金額(索賠金額-判決金額)進行描述性統(tǒng)計,結(jié)果如表3。
欺詐總額,從79.48元~399 138.7元不等,均值29 121.08元,不同案件欺詐金額差異很大。135例欺詐案件中,意外傷害住院醫(yī)療保險(63例)欺詐數(shù)量最多,其次為疾病住院醫(yī)療保險(35例)和傷殘保險(33例)。從子項欺詐金額看,疾病身故保險(1例)與意外傷害身故保險(25例)欺詐均值最高,分別為80 000元和68 267.09元;其次是傷殘保險(33例)和重大疾病保險(1例),分別為30 901.44元和10 000元;疾病醫(yī)療保險(1例)最低,均值499.5元。
四、Logistic回歸分析
為避免多重共線性,本文篩選出的各變量容忍度(Tolerance)均大于0.2,方差膨脹因子(Variance inflation factor,VIF)均小于5,Logistic回歸結(jié)果見表4。
表4顯示,極大似然估計值為195.624,Nagelkerke R2為0.615,說明模型擬合度較好;從P-value看,年齡41~50歲、年齡51~60歲、工人、北京市區(qū)、北京郊區(qū)縣、高保險金額、健康險、長期保險、摔扭傷、交通事故機動車司機傷害、被保險人索賠、非律師參與的回歸系數(shù)在5%置信度水平下顯著不為零,年齡41~50歲、年齡51~60歲、工人、北京市區(qū)、北京郊區(qū)縣、高保險金額、健康險、長期保險、摔扭傷、交通事故機動車司機傷害、被保險人索賠、非律師參與是欺詐概率識別的關(guān)鍵要素。
年齡51-60回歸系數(shù)為2.002,年齡51~60歲比其它年齡組欺詐對數(shù)發(fā)生比平均高2.002倍,可能是年齡51~60歲已進入老年,但尚未退休,對未來預(yù)期較悲觀,為應(yīng)對未來不確定性而欺詐。年齡41~50歲回歸系數(shù)為0.314,年齡41~50歲比其他年齡組欺詐對數(shù)發(fā)生比平均高0.314倍,可能是年齡41~50歲上要贍養(yǎng)父母,下要撫養(yǎng)子女,經(jīng)濟壓力較大,為緩解經(jīng)濟壓力而欺詐。從回歸系數(shù)看,年齡41~50歲欺詐明顯弱于年齡51~60歲,但結(jié)合表2描述性統(tǒng)計I,年齡41~50歲欺詐案件最多(59例),是年齡51~60歲(22例)的2.68倍,這兩個年齡段欺詐特點不同,應(yīng)采取不同的應(yīng)對措施。工人回歸系數(shù)為1.68,工人比其他職業(yè)欺詐對數(shù)發(fā)生比平均高1.68倍,可能是工人收入偏低,通過欺詐緩解自身經(jīng)濟壓力。北京郊區(qū)縣回歸系數(shù)為2.422、北京市區(qū)回歸系數(shù)為2.200,北京郊區(qū)縣比外地欺詐對數(shù)發(fā)生比平均高2.422倍、北京市區(qū)比外地欺詐對數(shù)發(fā)生比平均高2.200倍,北京欺詐甚于外地,可能源于北京高房價、高工作強度、高消費等壓力;北京郊區(qū)縣欺詐甚于北京市區(qū),可能是北京郊區(qū)縣聚集了相當數(shù)量外來務(wù)工人員,文化素質(zhì)較低,對保險條款理解膚淺,出險后不管事故嚴重程度如何,往往按保險金額上限索要。
高保險金額回歸系數(shù)為1.352,高保險金額比低保險金額欺詐對數(shù)發(fā)生比平均高1.352倍,欺詐和保險金額正相關(guān),高保險金額更會誘致欺詐,例如疾病身故保險欺詐金額80 000元,意外傷害身故保險欺詐金額68 267.09元。長期保險回歸系數(shù)為-1.198,長保險期間比短保險期間欺詐對數(shù)發(fā)生比平均低1.198倍,欺詐與保險期間負相關(guān),短期保險更易導(dǎo)致欺詐,例如意外傷害住院醫(yī)療保險欺詐63例,可能為彌補保費和免賠額而欺詐。健康險回歸系數(shù)為2.811,健康險比意外險欺詐對數(shù)發(fā)生比平均高2.811倍,可能是健康險承保常免于體檢,便于被保險人隱瞞所患疾病。
摔扭傷回歸系數(shù)為2.561,摔扭傷比其他傷害欺詐對數(shù)發(fā)生比平均高2.561倍,國際欺詐識別研究表明,摔扭傷系軟組織傷害,最難界定,更易欺詐。交通事故機動車司機傷害回歸系數(shù)為1.729,交通事故機動車司機傷害比交通事故非機動車傷害欺詐對數(shù)發(fā)生比平均高1.729倍,可能是交通事故機動車司機就是事故直接責任人,迫于息事寧人的賠償壓力而欺詐。被保險人索賠回歸系數(shù)為2.608,被保險人索賠比其他人索賠欺詐對數(shù)發(fā)生比平均高2.608倍,因為被保險人就是保險事故當事人,比投保人、受益人等更了解意外事故真實情況和自身疾病詳情,索賠時更能夠避重就輕,掩蓋事實真相或夸大受傷、疾病程度。非律師參與回歸系數(shù)為2.239,非律師參與比律師參與欺詐對數(shù)發(fā)生比平均高2.239倍,非律師參與人多是索賠方的親戚、朋友、同事等,對保險、法律也不甚了解,可能基于同情保險事故而包庇、縱容索賠人欺詐行為,親親相隱。
五、結(jié)論和建議
本文對實證調(diào)研的258個意外、健康保險理賠訴訟樣本進行描述性統(tǒng)計和Logistic回歸,發(fā)現(xiàn)索賠人年齡、職業(yè)、所在地區(qū)、保險金額、險別、保險期間、保險事故類型、索賠人類型、律師是否參與等是欺詐概率識別的關(guān)鍵因子。
1. 年齡41~50歲、年齡51~60歲欺詐概率甚于其他年齡組,其中年齡41~50歲欺詐頻次最多,年齡51~60歲欺詐概率最大。工人欺詐概率甚于其他職業(yè)。北京欺詐概率甚于外地,北京郊區(qū)縣欺詐概率甚于北京市區(qū)。建議保險公司理賠時,重點審查年齡41~50歲、年齡51~60歲、工人、北京、北京郊區(qū)縣等投保方特征。
2. 保險金額與欺詐概率正相關(guān),身故保險金額較高,欺詐概率遠甚于其他保險。保險期間與欺詐概率負相關(guān),意外傷害住院醫(yī)療保險期限較短,欺詐概率遠甚于其他保險。健康險欺詐概率甚于意外險。建議保險公司理賠時,高度關(guān)注高保險金額、短期保險、健康保險等保單信息。
3. 摔扭傷欺詐概率甚于其他傷害,交通事故機動車司機傷害欺詐概率甚于交通事故非機動車傷害。被保險人索賠欺詐概率甚于投保人索賠,非律師參與欺詐概率甚于律師參與。建議保險公司理賠時,重視保險事故類型(摔扭傷、交通事故機動車司機傷害)、索賠人類型(被保險人索賠)、律師是否參與(非律師參與)等案件情況。
參考文獻:
[1]M. Artís,M. Ayuso,M. Guillén. Detection of Automobile Insurance Fraud with Discrete Choice Models and Misclassified Claims. The Journal of Risk and Insurance,2002,(69):325-340.
[2]S.B. Caudill,M. Ayuso,M. Guillén. Fraud Detection Using a Multinomial Logit Model With Missing Information. The Journal of Risk and Insurance,2005,(72):539-550.
[3]J.Ai,P. Brockett,and L. Golden. Assessing Consumer Fraud Risk in Insurance Claims with Discrete and Continuous Data. North American Actuarial Journal,2009,(13):438-458.
[4]J.P. Boucher,M. Denuit,M. Guillen. Number of Accidents or Number of Claims? An Approach with Zero-Inflated Poisson Models for Panel Data,The Journal of Risk and Insurance,2009,(76):821-846.
[5]H.J. Smoluk. Long-Term Disability Claims Rates and the Consumption-to-Wealth Ratio. The Journal of Risk and Insurance,2009,(76):109-131.
[6]周建濤,巨珣,董楠.保險公司應(yīng)對欺詐的訴訟研究[J].北京工商大學(xué)學(xué)報(社會科學(xué)版),2011,(4):117-121.
責任編輯、校對:張增強
Abstract: The insurance claim process is often accompanied by the claimant exaggerated claims. This article regresses the 258 accident and health insurance sample data of claiming litigations between 2000 and 2011 with descriptive statistics and Logistic model, finding that the 41-50 and 51-60 year-old is worse than the other year-old, and the 41-50 with the highest fraud frequency and the 51-60 with the highest fraud probability; the worker' fraud is worse than the others'; Beijing is worse than other areas, and Beijing suburb worse than Beijing city; fraud is significantly positive with insurance amount, the death insurance from illness with the largest fraud amounts, fraud is significantly negative with insurance term, casualty hospitalization insurance with the highest fraud frequency, health insurance worse than accident; strain/sprain fraud is worse than trauma, troublemaker in a traffic accident worse than the injured, the insured fraud worse than the insurant, fraud with non-lawyer worse than that with lawyer.
Key words: fraud probability, the insured characteristics, policy information, the case circumstances, health insurance, statistical analysis, insurance amount, lawyer
[5]H.J. Smoluk. Long-Term Disability Claims Rates and the Consumption-to-Wealth Ratio. The Journal of Risk and Insurance,2009,(76):109-131.
[6]周建濤,巨珣,董楠.保險公司應(yīng)對欺詐的訴訟研究[J].北京工商大學(xué)學(xué)報(社會科學(xué)版),2011,(4):117-121.
責任編輯、校對:張增強
Abstract: The insurance claim process is often accompanied by the claimant exaggerated claims. This article regresses the 258 accident and health insurance sample data of claiming litigations between 2000 and 2011 with descriptive statistics and Logistic model, finding that the 41-50 and 51-60 year-old is worse than the other year-old, and the 41-50 with the highest fraud frequency and the 51-60 with the highest fraud probability; the worker' fraud is worse than the others'; Beijing is worse than other areas, and Beijing suburb worse than Beijing city; fraud is significantly positive with insurance amount, the death insurance from illness with the largest fraud amounts, fraud is significantly negative with insurance term, casualty hospitalization insurance with the highest fraud frequency, health insurance worse than accident; strain/sprain fraud is worse than trauma, troublemaker in a traffic accident worse than the injured, the insured fraud worse than the insurant, fraud with non-lawyer worse than that with lawyer.
Key words: fraud probability, the insured characteristics, policy information, the case circumstances, health insurance, statistical analysis, insurance amount, lawyer
[5]H.J. Smoluk. Long-Term Disability Claims Rates and the Consumption-to-Wealth Ratio. The Journal of Risk and Insurance,2009,(76):109-131.
[6]周建濤,巨珣,董楠.保險公司應(yīng)對欺詐的訴訟研究[J].北京工商大學(xué)學(xué)報(社會科學(xué)版),2011,(4):117-121.
責任編輯、校對:張增強
Abstract: The insurance claim process is often accompanied by the claimant exaggerated claims. This article regresses the 258 accident and health insurance sample data of claiming litigations between 2000 and 2011 with descriptive statistics and Logistic model, finding that the 41-50 and 51-60 year-old is worse than the other year-old, and the 41-50 with the highest fraud frequency and the 51-60 with the highest fraud probability; the worker' fraud is worse than the others'; Beijing is worse than other areas, and Beijing suburb worse than Beijing city; fraud is significantly positive with insurance amount, the death insurance from illness with the largest fraud amounts, fraud is significantly negative with insurance term, casualty hospitalization insurance with the highest fraud frequency, health insurance worse than accident; strain/sprain fraud is worse than trauma, troublemaker in a traffic accident worse than the injured, the insured fraud worse than the insurant, fraud with non-lawyer worse than that with lawyer.
Key words: fraud probability, the insured characteristics, policy information, the case circumstances, health insurance, statistical analysis, insurance amount, lawyer