A study of U S Unemployment Insurance - Fraud and Errors

[[abstract]]本論文研究美國勞工保險2013-2014年各州之誤賠因素及失業率，藉由統計分析推估美國各州的詐保率，資料蒐集至美國勞工部門。 首先介紹詐保之概念並簡述BAM(Benefit Accuracy Measurement Program)制度，BAM制度是對於勞工保險理賠之賠償品質的監控制度，在美國此誤賠檢測制度是非常重要的。 敘述統計包含總和年度和各年度分別之數字摘要和圖形，再將美國本土分成四區進行相關分析，探討各誤差因素和失業率及詐保率之間的相關性，並以ANOVA分析美國本土之分區間各誤差因素和失業率及詐保率之間的差異性。 最後以詐保率為應變數，誤賠因素及失業率為自變數，採用逐步迴歸探討具顯著解釋力之變數，建立模型以進行對詐保率的監控及解釋。[[abstract]]This paper studies fraud and error factors of United States Labor Insurance and the unemployment rate of states which using the statistical analysis to estimate the United States in each insurance fraud rate of states by collects the United States labor depart-ment information that released in 2013-2014 and states data for research First it introduces some concepts types outcomes of Insurance Fraud and con-cerns about the Benefit Accuracy Measurement Program (BAM) The BAM is the qual-ity of pay by insurance benefit which is designed to determine the accuracy in the Un-employment Insurance programs In United States the BAM program is an important role for detecting the error The Descriptive Statistics and graphs in total and each year are shown Next we classify four regions in Unite State and analyzed the correlation between error causes Unemployment rate and Fraud rate Furthermore we test to see if there exist significant regional differences among the variables Secondly the major study on the research of United States unemployment insur-ance for the fraud rate and error causes Here uses some statistical model to fit analyzes and detects the circumstance of each cause At the end regression model are built to predict the insurance fraud rate by using stepwise method the result can be used to explore the variables for monitoring the rate of insurance frau