本文主要目的是採用共積、因果關系、及誤差修正VAR模型等方法,探討我國獨樹一幟的土地增稅,與住宅價格間的關系。除檢驗政府藉由提高公告現值,以打擊房地產投機行為,穩定住宅價格的政策是否可行,整合長期以來各文獻的爭論外,公告現值調整與住宅價格變動間的動態影響型態及方式為何?公告現值調整機制是否會加深房地產衰退的趨勢等,亦是本文的研究重點。 根據共積檢定發現,住宅價格波動與公告現值調整,具有長期穩定共同成長的關係。因果關係(Granger-causality)檢定則發現,兩者僅存在單向因果關係,公告現值的調整的確會受住宅價格變動的影響,但公告現值調整卻不會影響住宅價格波動,因此政府欲利用調高公告現值增加土地增值稅賦,作為抑制房地產投機行為,穩定住宅價格的政策,其實是不可行的。再者,根據誤差修正VAR模型估計可知,公告現值的調整僅受前期住宅價格變動的影響,並不受前期公告現值調整及其他各期住宅價格波動與公告現值調整的影響,此一結果說明了我國土地公告現值的評定,可能僅以趕上土地市價變動為目標,而且根據估計係數皆小於一可知,此一目標亦無法達成。因此,公告現值僵固的調整機制,應該不會加重房產景氣衰退趨勢。
So far, lots of papers have studied the relationship between land value increment taxation (LVIT) and housing price, however, there are no agreed conclusions. By employing the co-integration test, the Granger-causality test, and the error correction VAR model, we empirically review this issue by using the annual data of three kinds housing prices and the declared land values of Taipei city from 1968 to 1994. According to the results of all tests, we find an agreed conclusion in the relationship between any kind housing price and LVIT. By Granger-causality test, this paper reveals that the causality leans one way- the change of housing price will influence the one of the declared land value. By using error correction VAR model, further empirical test has revealed that only last change of housing price will influence the declared land value.