首先使用最小平方法進行迴歸分析，但因為本研究所採用的資料為兼具時間序列資料及橫斷面資料，若採用最小平方法會有偏誤的情形，故較適合追蹤資料迴歸分析，而想瞭解三種細懸浮微粒的被解釋變數與解釋變數之間的關係沒有很大的差異，而實證結果得知與過去文獻不同，本研究並不符合EKC假設，故所得與細懸浮微粒之間呈現U形關係；解釋變數人口年齡結構的分組，六歲以下的幼年人口並沒有統計顯著的意義，十五歲以下的青幼年人口及六十五歲以上的老年人口對細懸浮微粒濃度都有顯著的影響效果；本研究的解釋變數工廠數量並沒有統計意義，可能是每年的工廠數量變化較不大所造成的結果；本研究衡量交通變數的加油站數量，其對細懸浮微粒濃度也有顯著的影響。 ;This study mainly discusses about the factors how effect on air pollution, like age structure、income、traffic and so on. The relationship between economic growth and environmental quality has been drawn considerable attention for the last three decades, so this study also validates the EKC hypothesis. My contributions is to further disaggregate PM2.5 into three types and further disaggregate population into three particularly key age groups: 0-6, 0-15, and over 65, and by doing so demonstrate that population’s environmental impact differs considerably across age groups.
First, the Ordinary least squares method is used, but the data of study has both time series data and cross-sectional data, the estimate results of OLS is biased. This study model using a seventy – seven stations panel data set in Taiwan over the period 2012–2017. Findings indicate that income U-shaped relationship with PM2.5, so this study does not support the environmental Kuznets curve hypothesis. The variables of 0-15 and the number of gas stations has a positive relationship with the PM2.5, the variables of population over 65 years old has a negative relationship with the PM2.5.
