健康與教育之長期效果: 台灣雙胞胎與手足樣本研究

Abstract

第一章:The Short-Run and Long-Run Effects of Birth Weight 出生體重的報酬一直是健康經濟領域的重要課題,並同時也是國家健康政策中不 可或缺的一環。 在這一章,根據出生於1979至1984年間的台灣雙胞胎與單胞胎手 足樣本,我試圖評估出生體重的短期至長期效果。 透過固定效果模型的運用,我 發現出生體重確實對小孩未來的健康狀態、教育表現以及薪資水準具有顯著正向 影響。 此外,我也進而檢視出生體重效果在不同類別的雙胞胎間是否有所差異, 以及觀察此效果是否呈現非線性。 第二章:An Examination of the Causal Relationship between Education and Health 高教育水準的人們是否擁有較好的健康? 許多研究指出教育程度愈高的人們確實 會在各種衡量健康的指標表現較佳,其中包括死亡、自我健康評估、健康相關行 為以及其他健康指標。 在這一章,我選擇平均一年的住院日數以評估一人健康狀 態,並同時透過工具變數法及雙胞胎 (及手足) 固定效果模型進行討論。 雖在工 具變數法的估計呈現不確定結果,但在利用雙胞胎及手足樣本的固定效果模型則 發現教育年數愈高的人們確實顯著地擁有較少的住院日數。 此外,教育對健康的 效果會在男性及高教育人們較大。 第三章:The Intergenerational Transmission of Education 過去已有許多研究透過不同的估計方法去檢驗教育是否會跨代傳遞,換言之,高 教育父母是否會導致其小孩同樣擁有較高的教育水準。 在這一章,我利用台灣的 雙胞胎與手足的兩代樣本來評估教育的跨代傳遞,發現父母若增加三年的教育年 數將僅使其小孩教育年數增加不超過半年。 此外,我自兩代手足樣本中進一步建 立一個三代樣本,試圖驗證Becker-Tomes模型的預測是否正確。 台灣三代樣本的 結果是過去相關文獻中首度支持該模型的預測,亦即,在其他條件不變下,祖父 母教育水準的提升將導致其孫子女教育年數的減少。

Chapter 1: The Short-Run and Long-Run Effects of Birth Weight Returns to birth weight have been a major topic in the field of health economics and become an important policy of national health during the past few years. In this chapter, I take advantage of Taiwanese twins and siblings over the period 1979–1984 to investigate the short-, medium-, and long-term effects of birth weight. My results show a positive causal relationship between an infant’s birth weight and health status, schooling, and labor market outcomes. I also check the robustness of my estimates by comparing the twin fixed-effects estimates across various sub-groups of twins and considering the possibility of nonlinear effects of birth weight. Chapter 2: An Examination of the Causal Relationship between Education and Health Do more educated people have better health status? Several studies show that well-educated people tend to have better health in terms of mortality, self-reported health conditions, health-related behaviors, and other measures for physical or mental health. In this chapter, I select the number of days hospitalized per year as the main indicator for health conditions and evaluate the education-health gradient via an instrumental variable method and a within-twin (within-family as well) estimator. The instrumental variable estimation shows mixed results, whereas the fixed-effects approach indicates that the effects of educational attainment are statistically significant on total days hospitalized per year. Moreover, the effects of schooling on health are bigger for both men and better-educated people. Chapter 3: The Intergenerational Transmission of Education Whether the intergenerational transmission of schooling exists has become a widely discussed topic via different empirical settings. In this paper, I use Taiwanese same-sex twins and siblings to examine parental schooling effects on their children’s schooling. The empirical findings based on 1951–1960 Taiwanese birth cohorts suggest that three additional years of parents’ education translates into an increase of children’s schooling by at most 0.5 year. Furthermore, I take advantage of a three-generation sample to test the Becker-Tomes model and find a strong evidence of supporting the important prediction of the Becker-Tomes model, that is, negative grandparental schooling effects.