Text

theses

This dissertation discusses a wide variety of topics on demographic change and education. In chapter 2, I examin the fertility effect of Hurricane Katrina. After Hurricane Katrina in 2005, official birth records showed a significant increase in the fertility rate in hurricane destroyed areas. After carefully examining data quality and possible problems with the difference-in-difference method, I conclude that Hurricane Katrina had no significant impact on fertility rate in the affected states. A county-level analysis also supports this conclusion. In Chapter 3, I study the determinants of the supply of secondary education in China. I consider the unique characteristics of the Chinese educational system: strict hierarchies and consequent competition among schools. Schools on higher level, i.e. high schools, and schools of higher quality, i.e. key-point schools, are at the top tier of the hierarchical system, able to enjoy better resources and serve more capable students. I suggest that when schools and education authorities make joint decisions on enrollment, they face a tradeoff between quantity and quality, and a school's position on the hierarchy determines its specific weighting of quantity versus quality. Using census and aggregate enrollment data, I find strong evidence that cohort-crowding effect for different schools is consistent with their optimal decisions on enrollment. In general a 10% increase in cohort size reduces middle school enrollment rate by less than 1% and high school enrollment rate by 3-4%. Meanwhile, a 10% increase in cohort size reduces teacher/student ratio in primary school, middle school and high school by 3.9%, 3.7% and 2.5% respectively. The variation in different schools' responses to cohort size suggests that schools on top tier of the hierarchy sets higher premium on quality than quantity. Further analysis shows that factors affecting the weighting of quantity at different schooling levels, including the introduction of Compulsory Education Law and the number of high schools per county, also have predicted impacts on enrollment. My findings suggest that the strict hierarchy and disparity in resource allocation are the major reasons of inelastic supply in education and I prove that such a system also leads to intense competition among students and loss in efficiency compared to an egalitarian system. Chapter 4 analyzes the whole college going process from application to admission to matriculation, taking use of the national representative dataset Education Longitudinal Study of 2002 (ELS:2002) which contains students' full records on transition from high school to college. This paper addresses two main problem in estimation: first, for college's admission decision, I use information revealed from application to proxy student's unobserved ability and use conditional logit model to eliminate college fixed effect, solving the problem of endogeneity. Second, for student's matriculation decision, as each student's complete choice set is accessible, I use conditional logit model and nested logit model to obtain unbiased estimation on factors affecting a student's matriculation decision. The results show that students' application to different types of college has different implications on student abilities; gender, race, SAT score and academic performance in high school affect students' probability to be admitted. Also, factors like tuition and distance have significant impact on students' matriculation decisions.

ETD_8317

Ph.D.

Includes bibliographical references

by Han Liu

doi:10.7282/T3R49TW9

ETD doctoral

The author owns the copyright to this work.