交通建設計畫之價值回饋與財務自償及風險分析研究：以租稅增額融資為例

SELF-LIQUIDATION ABILITY AND RISK ANALYSIS OF VALUE CAPTURE TO TRANSPORT PROJECTS: EXAMINATION OF TAX INCREMENT FINANCING

長期以來，自償能力一直是政府評估交通建設計畫財務可行性之基礎，但此自償率多侷限於計劃之營運收入與成本可量化項目，對於計畫所衍生之外部性多視為不可貨幣化，忽略不計。晚近，政府所推動之「跨域加值公共建設財務規劃方案」即是將「租稅增額融資或租稅增額財源（tax increment financing, TIF）」引入於計畫財務補貼評估架構。TIF屬於價值回饋機制之一，此機制是將計畫所衍生之外溢效益予以內部化，進而提升計畫之財務自償能力。此租稅增額融資屬於預期性的財政融資工具，但由於預期性租稅具有高度不確定性，倘若租稅增額成長趨勢未如預期，將導致政府財政負擔風險，影響其在財務資源分配與風險分擔能力；然而，此一課題卻為相關研究所忽略。本研究考量相關法律、財務條件限制下，兼顧地方政府追求稅收增額融資風險最小化及中央政府建設補貼經費最大化，將計畫之外部效益予以內化於財務自償，進而提出多目標財務數學規劃模式。本研究以捷運建設計畫「三峽鶯歌線（三鶯線）」為例，進行租稅增額融資、自償率、補貼及財務風險衡量。本研究利用蒙地卡羅法模擬租稅增額融資機率分配，模擬顯示其符合逆高斯分配（Inverse Gaussian）型態，利用本研究之所提之多目標數學規劃模式及求解法，可同時獲得地方政府之租稅增額融資、中央補貼額度、自償率及其風險承擔等重要意涵，此可提供捷運建設計畫實施租稅增額融資審議措施之參考。

In literature of transportation, the government agencies or researchers only focused on operating revenue and cost items when evaluated the cost-benefit or self-liquation analysis for transport projects. The external effects of an investment project are always ignored by some studies because those of external factors are regarded as un-measured items. Thus, this would result in the distortion of financial evaluation. Recently, the government conducts the tax increment financing (TIF) scheme, one of value capture methods, to incorporate the external effects into the SLR (Self-Liquidation Ratio) analysis for transport projects. However, the TIF is a financing technique with high uncertainty on expected revenue in the future period around the tax incremental financing districts (TID). This would result in huge risk in finance for central and local government agencies if TIF below the estimating expected value. This study presents a multi-objective programming model using mathematical programming technique to internalize the externality factors including house tax, land value Increment tax, and land price tax into the SLR analysis. This study conducts a case study with a Mass Rapid Transit (MRT) project to measure the TIF, risk, SLR, and subsidy levels for the MRT project. The results show that the TIF is an inverse Gaussian probability distribution using the Monte-Carlo simulation. In particular, the TIF, risk, SLR, and subsidy levels can be determined by the proposed model for local and central government agencies simultaneously.