本文的研究目的是以台中市污染控制場址與整治場址為例，探索污染場址周 邊地價的變動在污染公告前後是否產生顯著的差異、差異變動的趨勢、以及污染 公告對於該差異的解釋力。本研究同時採用量化與質化研究法，分析結果除了能 與已開發國家先行研究結果進行對話之外，也期許能提供環保署目前致力於推動 污染區再利用計晝時的政策參考。本研究先以台中市各污染場址所屬地價區段的 公告土地現值為基礎，經換算成土地推估市價後，以其變動率進行統計分析，再 透過專家深度訪談，為統計分析的結果提出解釋。經重複測量變異數分析顯示， 污染場址周邊土地的推估市價變動率在污染公告前後有顯著差異，但與常理相異 的是，其變動趨勢是上升的，而且上升的幅度也是漸增的。而經逐步迴歸分析後 則發現，絕大部分場址的污染公告對於周邊土地推估市價的變動都無顯著的解釋 力，而即使在少數場址中具有解釋力，也頂多抑制地價成長率，而無降低地價的 效果。經由專家深度訪談，本研究歸納出五個面向來解釋此種與大部分西方研究 大相逕庭的現象，分別是污染資訊不夠公開、不動產買方的污染容受程度大、整 體區域發展沖淡污染的負面影響、單一或少數污染場址無法對於整個區段地價造 成太顯著的影響、以及公告土地現值的資料特質。

This study intended to explore whether there were significant differences in values of land surrounding contaminated sites in Taichung City before and after the sites were officially declared as pollution control sites. The changing patterns of the land values and the explanatory power of the declaration on the changing patterns were also examined. The analysis of the quantitative and qualitative data helped to evoke discourse with existing literature and to provide policy suggestions regarding brownfield redevelopment to the Environmental Protection Administration. In this study, the Current Land Values announced by the government each year were calculated to be the estimated market values for the purpose of statistical analysis. The in-depth interviews were then conducted to extract explanations for results drawn from statistical analysis. Through One-way Repeated Measures ANOVA, this study found that the change rate of values of land around pollution control sites varied significantly before and after pollution declaration. Contrary to common belief, the land values increased after the declaration and so does the growth rate. The stepwise regression results showed that, for most of the pollution sites, declaration had no significant explanatory power on the changing land values. Through in-depth interview with six professionals, this study summarized five explanations for the statistical analysis results. Firstly, the information regarding possible pollution was not transparent in land transactions. Secondly, the land buyers may tolerate pollution because of a lack of environmental awareness or their own cost-benefit calculations. Thirdly, the positive effect brought by district development diluted the negative effect from pollution. Fourthly, single or few pollution sites did not cause significant effect on land values of the entire land-price section. Fifthly, the characteristics of the Current Land Value data may affect the statistical analysis results.