Three Essays on Applied Microeconometrics: Education Policy, Intergenerational Schooling Effect and Value of a Statistical Life

• Main Authors: Wehn-Jyuan Tsai, 蔡彣涓
• Other Authors: 劉錦添
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/13920125922630636837

博士 === 國立臺灣大學 === 經濟學研究所 === 95 === My research involves applying advanced econometric techniques to further our understanding of some key issues in applied microeconomics. This dissertation is composed of three self-contained empirical essays. In chapter 1, we analyze the 1968 educational reform in Taiwan and its impact on gender disparities in schooling, as well as workforce outcomes (including sector/class of employment). The 1968 educational reform extended the island’s tuition-free compulsory education from six to nine years and has long been recognized as a critical role in the rise in educational attainment, resulting in considerable changes in the structure of the labor market. Our regression accounting for schooling (based upon the 1980 and 1990 censuses) suggests that the reform did not lead to convergence of the gap in the employment rate based upon gender, but instead, allocated those females educated to junior-high school level into the modern sectors, by reducing the male-female differentials in the contributions of schooling among this group. Using a structural-form estimation of the effects of the 1968 reform, this equates to an increase of 14 per cent, for each additional year of schooling, in the probability of females being employed in manufacturing, alongside a corresponding shift of females out of agriculture. In chapter 2, we exploit the natural experiment brought about by the Taiwanese educational reform in 1968 to evaluate the effects of parental schooling on educational achievements by the subsequent generation. Between 1968 and 1973, the Taiwanese government was involved in the greatest expansion of the public junior high school system on record. This policy introduced an exogenously regional variation in schooling between pre-reform and post-reform cohorts. We take advantage of the region-specific and cohort-based intensity of the program, as measured by the stock of public junior high schools per 1,000 of the population aged between 12 and 14 years in a given year, to construct instrumental variable (IV) estimates of the inter-generational schooling effect. Using the Joint College Entrance Examination (JCEE) files from 2000 to 2003, we examine the influence of compulsory education undertaken by parents on their children’s JCEE scores. Our estimates suggest that the schooling of the mother and the father have diverse effects on a child’s test performance scores. While the father’s schooling does have a determining role in the access of his child to post-secondary education, higher educational achievements by the mother are found to significantly improve the quality of the college or university which the child may ultimately attend. This measurement is undertaken by two different outcomes, whether children attend a public university, or whether they attend one of the top six public universities in Taiwan. In chapter 3, we use a unique longitudinally-linked employer-employee dataset, covering the period 1998-2000, to estimate the value of a statistical life for both male and female workers in Taiwan. Using the linked employer-employee data, we construct on-the-job injury and fatality risk measures that effectively capture the variation in job risks for the firm in which the worker was employed. Our panel of workers and firms also permits us to control for unobserved characteristics specific to both workers and firms. Based upon the estimation of the individual- and firm-level fixed effects, our results indicate that the value of life for male workers is US$0.03 million, while the value of life for female workers is US$0.14 million (both in 1998 dollars). Unlike the conventional industry or occupational risk measures which were invariably used in the prior studies, these value-of-life estimates recognize the variation in job risks between firms (with adjustment for the differences in job risks for each gender) as well as the unobservable heterogeneity of both workers and firms.