土地增值稅自民國九十一年二月一日起兩年內減徵百分之五十（民國九十三年一月又延長一年至九十四年一月三十一日止）。以往研究發現，當稅制改變時，股市會反應稅制變動所帶來的影響，但鮮少有研究探討稅制變動是否帶來公司實際交易改變之影響。本研究針對土地增值稅減半徵收對我國上市櫃公司土地交易的影響進行實證分析。研究結果發現（1）在減徵政策宣布之後，但實際生效之前，我國上市櫃公司傾向延後處分所持有的土地（2）整體統計資料顯示，土地增值稅減徵政策有助於刺激上市櫃公司進行土地交易以及處分具有較高未實現持有利得之土地之現象，但實證迴歸的結果並不支持。（3）本文實證迴歸結果顯示，營業虧損的公司以及擁有較多出售用土地的公司較會在土地增值稅減徵期間進行土地交易。 基於上述實證結果，本文推論，在減徵正式實施後所增加的土地交易，其背後的動機多為彌補本業虧損或藉由在減徵期間進行土地交易，墊高土地成本以降低未來租稅負擔。因此，由我國上市櫃公司土地交易情形觀之，此次土地增值稅減徵對於欲達到刺激經濟成長的目標，成效並不顯著。 / The Land Value Increment Tax (LVIT) has been halved for two years from Feb. 1, 2002. (The tax cut policy has been extended from two years previously to three years until Jan. 31, 2005). Prior studies found that the impacts of change in tax rate will be reflected on company stock prices, but fewer studies were aimed at the impacts of change in tax rate on real transactions. This study examines the impacts of LVIT cut on Taiwan’s listed and OTC companies land transactions. It is found that (1) According to the empirical results, listed and OTC companies deferred their land transactions after the announcement of the tax cut policy. (2) Based on the descriptive statistics results, it is found that the tax cut policy induces listed and OTC companies to sell land in the implementation period and to postpone selling land with higher holding gains during the announcement period to gather greater disposal gains. However, the regression results do not support the same conclusions. (3) The empirical results of this study provides evidence that higher Land-for-sale and Operating loss are important motives behind increased land transactions by listed and OTC companies in response to the tax cut policy. Based on the above-mentioned empirical results, this study infers that, as a result of the LVIT cut, the motives behind increased land transactions conducted by listed and OTC companies are for the purpose of reducing tax burden in the future or covering loss from operating activities. Consequently, this tax cut policy contributes little to efforts to stimulate economic upturn, since most of the listed and OTC companies only take advantage of the policy to gather tax benefit or to cover loss -- a development that is not beneficial for overall economic growth.

建設公司為典型的內需產業，隨著國內經濟的發展而變動。在民國80年後房地產市場因經濟與政治的動盪而一路下滑，陷入長期的衰退時期，因此政府採取許多利多措施以振作該產業的發展，而土地增值稅減半徵收政策就是其中一項振興措施。土地增值稅減半措施可以降低土地的交易成本，被房地產市場認為是利多消息，而國內研究多從投資人的觀點探討政策對股價的影響，未曾探討該政策對身為房地產市場供給者之建設公司經營效率的影響程度。 因此，本研究針對1998年至2003年的36家上市櫃建設公司，採用資料包絡分析法與大邊界法的概念，畫出單一包絡曲線以涵蓋整體樣本期間，求出所有樣本的效率值並分析其長期趨勢，以探討減半徵收政策對建設公司成本效率、配置效率與技術效率的影響。 實證結果顯示，政策實施後之純技術效率顯著大於政策實施後，而政策實施後之規模效率則顯著小於政策實施前；以Tobit迴歸分析顯示，政策實施對建設公司整體而言有提升配置效率與純技術效率的效果，但規模效率是下降的，表示建設公司尚未達到最適生產規模，仍有調整提振的空間。另外，就不同組織型態下的建設公司探討政策對其影響程度，若就上櫃建設公司與上市建設公司兩個子群體分別觀察，分析結果發現上櫃建設公司在政策實施前與實施後之成本效率、整體技術效率、純技術效率與規模效率皆顯著大於上市建設公司，且政策實施對上市建設公司之純技術效率有大幅度提升的效果，而Tobit迴歸分析亦顯示上櫃建設公司之經營效率較上市建設公司為佳。從有無集團背景著眼分析顯示非屬於集團企業之建設公司在政策實施前與實施後之成本效率、整體技術效率與純技術效率皆顯著大於屬於集團企業之建設公司，而Tobit迴歸分析亦顯示屬於集團企業之建設公司在整體技術效率方面不如非屬於集團企業之建設公司。 / Developers as the typical domestic demand industry change with development of the domestic economy. Because of turbulence of the economy and politics the real estate market fell into long-term decline after 1990. So the government took many favorite policies to improve the development of this industry, and the Land Value Increment Tax (LVIT) cut was one of those policies. The LVIT cut could reduce the transaction cost of the land, so it was thought a profitable policy to the real estate market. Most domestic researches are based on the investors’ view to test the impact of the policy on stock market, but not focus on the influence on the suppliers of real estate market, developers. In order to trace the trend of developers’ efficiency, this research employs data of 36 listed developers from 1998 to 2003 and applies Date Envelopment Analysis (DEA) to construct a grand frontier. Therefore, based on the trend of grand frontier this research could examine the impact of LIVT cut on developers` cost efficiency, allocation efficiency and technical efficiency. The major findings of this research as follows. The statistical results show that pure technical efficiency after the policy is statistically significant greater than that before the policy, and scale efficiency before the policy is statistically significant greater than that after the policy. The Tobit regression suggests that the implement of the LIVT cut could improve pure technical efficiency and allocation efficiency but drop scale efficiency. The finding suggests that developers have not yet reached the best scale and still have adjustment space. In addition, this research examines the impact of the policy on developers under different organization type. First, to divide developers into TSE and OTC. The analysis result shows that whether before the policy or after the policy cost efficiency, technical efficiency, pure technical efficiency and scale efficiency of OTC are statistically significant greater than that of TSE, and Tobit regression also shows that business efficiency of OTC is greater than that of TSE. Besides, from views of group background, the analysis result shows that whether before the policy or after the policy cost efficiency, technical efficiency, and pure technical efficiency of independent developers are statistically significant greater than that of group developers, and Tobit regression shows that technical efficiency of independent developers is statistically significant greater than that of group developers.