區域經濟成長與環境品質：臺灣降尺度環境顧志耐曲線假說之實證分析

環境顧志耐曲線（environmental Kuznets curve，以下簡稱EKC）假說為探討環境污染是否會隨著經濟成長而呈現先遞增爾後遞減之趨勢，惟現有研究多著重於國家尺度研究，對於一國境內各區域之實證結果進行差異比較分析者為數不多。有鑑於地區性空氣污染物受區域產業發展及氣候與地理條件等因素影響甚鉅，而國家尺度僅能呈現全國平均結果，無法揭露各區域之極端差異，為填補此一文獻落差，本文旨在探究臺灣各區域EKC假說的實證結果差異。我們依據「都市及區域發展統計彙編」，將臺灣劃分為北、中、南、東四大區域，進行各區域PM10濃度及所得兩者關係之EKC假說驗證，並納入小客車數、第二級產業勞動與用地面積、降雨天數作為其他解釋變數。實證結果顯示，臺灣整體及四大區域之EKC假說實證結果有相當大的差異，且影響各區域之PM10濃度的因素不甚相同，職是之故，有效的區域空氣污染防制措施需根據區域產業經濟發展狀況以及氣候與地理條件因地制宜，以期達成區域經濟發展與污染脫鉤之目標。 The hypothesis of Environmental Kuznets Curve ( EKC ) aims at exploring whether pollution would first increase with economic growth and then decrease after some turning point(s). A survey on the current literature shows that most of the EKC studies focus on the empirical tests at national level. Few pay special attention to the differences in empirical results across regions within a country. Given the fact that regional air pollutions(e.g., PM10 ) are significantly influenced by regional industrial development, weather, and other geographical factors, the studies at national levels may only show the average results for the whole country but belie extreme regional differences. To fill the gap in the literature, this study aims at exploring whether or not there are differences in the empirical results of EKC hypothesis across regions in Taiwan. Following regional classifications in “Urban and Regional Development Statistics”, we divide Taiwan into four regions, consisting of northern, central, southern, and eastern regions, and examine the validity of EKC hypothesis for each of them. Central to the analysis is to examine the inverted U-shaped relationship between regional PM10 concentration and income per capita using panel data regression with other explanatory variables, consisting of quantity of cars, labor and land used in the manufacturing sectors, and the days with rainfall. Our empirical evidence shows that the test results of EKC hypothesis are quite different across regions. In addition, the factors affecting regional PM10 concentration are dissimilar. Accordingly, to decouple economic growth and air pollutions, the policies for PM10 abatement should be designed based on regional attributes such as regional economic and industrial development, as well as climate and geographical factors.