College major choice has become one of the major life decisions, as the earnings differentials across majors have been found exceeding college premium in many countries. This dissertation comprises two essays exploring the determinants of college major choice in the context of the centralized National College Entrance Exam (Gaokao) in China.

The first essay explores media influence on college major choice by examining how the quality of medical school students responds to newspaper coverages of hospital violence in China. Following the rationale that more newsworthy stories tend to crowd out less eye-catching ones, we use news coverage on other exogenous noteworthy issues as an instrumental variable to identify the pure effect of hospital violence news. We find that the mean col...[ Read more ]

College major choice has become one of the major life decisions, as the earnings differentials across majors have been found exceeding college premium in many countries. This dissertation comprises two essays exploring the determinants of college major choice in the context of the centralized National College Entrance Exam (Gaokao) in China.

The first essay explores media influence on college major choice by examining how the quality of medical school students responds to newspaper coverages of hospital violence in China. Following the rationale that more newsworthy stories tend to crowd out less eye-catching ones, we use news coverage on other exogenous noteworthy issues as an instrumental variable to identify the pure effect of hospital violence news. We find that the mean college entrance exams (CEE) score of clinical medicine degree programs decreases by 0.93 percentage points if incoming students are exposed to one percent more news articles on hospital violence within one year before CEEs, compared to control programs. The decrease is larger in universities located in cities where violence occurred. A closer look at news contents reveals that editorials tend to deter students from choosing medicine majors, while reports on the conviction of criminals moderate this deterrent effect.

The second essay evaluates a conditional grant program in China, which commits students to teaching in their home province. Using a triple difference method, we find that teaching majors obtain better students due to the conditional grants. The policy effects are larger not only for colleges located in cities with lower living cost, but also in provinces with bigger share of disadvantaged students—those who are rural, female, and have more siblings. These results suggest that the Chinese free teacher education program does successfully attract high-quality students into teaching force, and these high-quality teacher trainees are very likely to be credit constrained.