人口老化對於所得分配之影響―以臺灣二十三個縣市為例

Abstract

本研究欲觀察人口老化與所得分配之關係，檢視人口老化對於所得分配之影響，而文中主要分別從「內政部統計年報」、「家庭收支調查表」、「人力統計資源年報」以及各縣市統計要覽獲得各種解釋變數，建立出台灣1998年至2005年總計8年台灣各縣市之資料，並且以各縣市之吉尼係數為被解釋變數，採用最小平方法之方式藉此觀察台灣各縣市所得分配與人口老化之關係。 \n 根據實證結果發現，人口老化與所得分配為顯著並且呈現正相關，亦即當人口老化更加嚴重時，將會導致較大的吉尼係數，造成所得不均之惡化。 其他解釋變數為顯著者分別有：平均每人可支配所得、平均每人可支配所得平方項、婦女勞動參與率以及各區域變數，且除了平均每人可支配所得平方項為負相關外，其餘皆與吉尼係數為正相關，其中所得方面之變數顯示出台灣中存在著Kuznets曲線；婦女勞動參與率的擴張則會導致所得分配更加惡化；中部、南部以及東部地區相對於北部區域來說其所得分配之狀況皆較不平均。而其他解釋變數如社會福利變數、教育變數以及工業變數所得之結果皆不顯著，顯示上述三個變數對台灣各縣市所得分配之影響在本研究中無法確定。

This research discusses the relationship between aging population and income distribution, and examines how the aging population affects income distribution. The independent variables were acquired from the Statistical Yearbook of the Interior, The Survey of Family Income and Expenditure, the Statistical Yearbook of Labor Force and the Statistical Yearbook from each twenty-three counties and cities in Taiwan to construct a data set from 1998 to 2005. Using the ordinary least squares method and choosing the gini coefficient as a dependent variable to observe the relationship between income distribution and aging population.\n According to the result of the empirical study, we discover that the effect of aging population is significant and has a positive relationship with income distribution. Income inequality is worse when the aging population increases. Other significant independent variables are disposable income per person, the square of disposable income per person, female labor participation rate and the regional dummy variables. Besides the square of disposable income per person, all variables have a positive relationship with the gini coefficient. In addition, the variable of income shows that Kuznets curve appears in Taiwan. The expansion of the female labor participation will cause the income distribution to change into inequality. Also the north region has the worst income distribution in Taiwan. The results of the social welfare variable, the education variable and the industry variable are insignificant so that in this article we can not observe how the three variables effect the income distribution in Taiwan.