有鑑於不法營業人冒退營業稅情況甚為嚴重,各國稅局莫不積極防杜及加強查緝;惟限於查核人力配置不足,遂針對每期(月)營業稅退稅案件審查作業實施分級審查制度,希將有限資源運用於異常案件之審核,以有效提升稽徵效能。而異常案件之判斷,須有賴統計資訊與指標作為參考的依據,本研究遂針對營業稅退稅異常案件之特性,以稅籍異常、虛報進項、虛報零稅率銷售額作為主要研究構面,並採用模糊AHP法以專家問卷方式確立各項指標權重及優先排序。 最後,本研究建立包含3個構面、12項評估準則及35項評估指標之「營業稅退稅選案查核模式風險指標體系」,期能提供稽徵機關實際運用於營業稅退稅案件選案查核之參考,以有效改善選案品質提升查核績效,達到打擊不法維護租稅公平的目標。
Due to the huge number of illegal business tax refund cases has become a serious problem, the National Tax Administrations of all 5 regions in Taiwan make every effort to improve this condition. However, because all National Tax Administrations are short-staffed on tax auditing, they are forced to classify the business tax refund cases each month into several classes and only focus on the most abnormal cases. Whereas, to identify abnormal cases, reliable statistical indicators are needed. Therefore, this study consider the characteristic of abnormal business tax refund cases, and based on this observation we choose taxpayer files abnormal, Input tax falsely reported、zero-tax-rate sales amount falsely reported as major indicators of abnormal cases. We apply Fuzzy Analytic Hierarchy Process to evaluate the weight and priority of each indicator. Finally, we construct a business tax refund cases selection and auditing risk indicator system composed of 3 dimensions and 12 criteria and 35 indexes. We hope that the system could help National Tax Administrations improve the performance in the practice of business tax refund case auditing, and the goal of maintaining a fair tax is more achievable.