Essays on Labor Economics

Xuan Jiang, Purdue University


This dissertation is composed of three independent chapters in the field of labor economics, focusing on educational decisions, gender differences and gender differences in educational decisions. The first chapter investigates gender differences in college major choice and job choice. Women are underrepresented in both STEM college degrees and STEM jobs. Even with a STEM college degree, women are significantly less likely to work in a STEM occupation than their male counterparts. This paper investigates whether men and women possess different ability distributions and examines how much the gender gap in major choice and job choice can be explained by gender differences in sorting on abilities. I use Purdue University's administrative data that contains every Purdue student’s academic records linked to their first job information. I apply an extended Roy model of unobserved heterogeneity allowing for endogenous choice with two sequential optimizing decisions: the choice between a STEM and non-STEM major and the choice between a STEM and non-STEM job. I find that both abilities are significantly weaker determinants of major choice for women than for men. High-ability women give up $13,000–$20,000 in annual salary by choosing non-STEM majors. Those non-STEM high-ability women only make up 5.6% of the female sample, but their total gains—had they made the same decision as men—explain about 9.4% of the gender wage gap. Furthermore, the fact that female STEM graduates are less likely to stay in STEM is unrelated to the differences in ability sorting. Instead, home region may be important in women’s job decisions; female STEM graduates who return to their home state are more likely to opt out of STEM. The second chapter exploits China's One-Child Policy to study the relationship between fertility expectation and educational attainment of the mothers of the “sibling-less generation”. One-Child Policy was China's most intensive family planning policy which implemented by the end of 1979 and only restricted Han families to have one child. I use two difference-in-differences approaches—one compares gender difference among Han, the other compares ethnicity differences between Han women and non-Han women—to estimate how Han women changed their educational choice in response to the policy. The OCP explains 53.6% of the 2.38 year average increase in education for women born between 1960–1980. Potential mechanisms include delaying entry to the first marriage, motherhood and increasing labor force participation. This study highlights the policy’s positive externality on women’s education. The third chapter studies how China's Open Door Policy's implementation at the end of 1978 affected the skill composition for workers born 1960-1970. Using measures of local labor markets' export exposure, we find that export growth increased high school completion rates but had no effect on middle school completion rates. For every $1000 increase in exports per worker, high school completion rates decreased by 4.76 p.p. for workers born in 1970 compared to those born in 1960, explaining about 10.4% of the national decline in high school completion for 1960s birth cohorts. This suggests a tradeoff between education and labor market opportunities in China. China's growth was likely dampened during the early industrialization of the 1980s and 1990s, as the Open Door Policy simultaneously reduced the availability of skilled labor.




Sarzosa, Purdue University.

Subject Area


Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server