- 學位論文
臺灣教育外部性之實證研究
The External Effects of Education: An Empirical Study of Taiwan
摘要
臺灣近年來高等教育的改革開放的確取得了重大的成就。但可能同時產生了一些問題，大學生數目過多導致資源配置不效率以及大學文憑所能傳遞的訊號降低，此時教育的社會報酬將可能降低。 我們以Mincer (1993)、Ciccone and Peri (2006)及Acemoglu and Angrist (2001)的模型為基礎，並以行政院主計處民國81年及民國96年二年的人力資源運用調查資料來進行實證分析。 根據我們的研究結果顯示，Mincer式的人力資本薪資函數迴歸，民國81年及民國96年時個人教育年數的係數估計值、個人工作經驗的估計值皆為正值，符合人力資本理論所云，接受較多教育及有較多工作經驗的個人平均來說會有比較高的薪資報酬，教育的個人報酬為正。 而使用Ciccone and Peri (2006)的模型分析的結果，在我們分析的這15年間，平均教育年數的增加對整體樣本有正的外部性。亦即，平均教育年數的增加對個人的薪資有正向的影響，人力資本有正的外部性。同時，女性的外部性高於男性。 至於我們使用Acemoglu and Angrist (2001)的模型來看平均教育年數增加與個人教育年數增加對個人薪資所得的影響，結果發現，平均教育年數的估計係數值和個人教育年數的估計係數值的差異很小。此外，將樣本區分為國小、國（初）中、高中職及專科、大學、研究所二組分別估計，發現第一組社會報酬高於個人報酬，而第二組個人報酬高於社會報酬。 雖然教育「金錢」上的外部性不大，但鑑於教育對個人或是社會仍有其他的好處。所以，我們認為，平均教育年數的增加其實對整個社會而言是有正的外部性的。
並列摘要
The revolution on higher education of Taiwan really made a huge success, but some negative effects have been created accordingly. Exceed of university graduates leads to an inefficient of resources allocation and reduces the signal of a bachelor degree. The social return of education will probably be decreased. We use Mincer (1993), Ciccone and Peri (2006), and Acemoglu and Angrist (2001) as the basic model for our empirical research, and use the 1992 and 2007 manpower utilization survey which was conducted by the MOI. Based on this research, the regression coefficients of years of education and working experiences of the Mincerian human capital function are all positive. That matches what human capital theory says. On average, people who get more education and more working experiences will have higher salary level than those who receive less education and working experiences. The private return of education is predictably positive. Using the model of Ciccone and Peri (2006), samples in this study represent that the increase in average years of schooling of the 15-year period has positive externalities, which means, the increase in average years of schooling has positive effect on individual's wage, and also has positive externalities on human capital. Meanwhile, the externalities on females are greater than the externalities on males. While using Acemoglu and Angrist (2001)'s model to examine the effects of the increase in average years of schooling and the increase in individual's years of schooling on individual's wage, we find that the regression coefficients of average years of schooling and individual's years of schooling show little difference. Besides, samples are divided into two groups for regression analysis, of which the first group contains individuals who have a highest degree of elementary school, junior high school, senior high school, and senior vocational school, and the second group contains rest of all. According to the regression results, the social return of the first group is higher than private return, while the second group has a higher private return than social return. Although the “pecuniary” externalities of education are small, since that education has other benefits to individual or to the society, after all, the increase in average years of schooling actually has positive externalities to the whole society.