Three Essays On The Economics of Education

Abstract

This thesis examines heterogeneity in human capital outcomes across race and gender. Using administrative data from British Columbia, the first chapter investigates the income-achievement gap in provincial test scores among Grade 4 and 7 students of different racial backgrounds. The second chapter estimates the impact of switching post-secondary majors on labour market earnings for men and women. Finally, using university application data from Ontario, the third chapter investigates gender gaps in applications, offers and acceptances to engineering and computer science programs.

In Chapter 1, I show that there is considerable variation in test score gaps between children from families of high- and low-socioeconomic status (SES) across racial backgrounds. In particular, the gap in mean test scores between Indigenous children of high- and low SES is 0.7 standard deviations, while it is only 0.37 standard deviations for East Asian children. Further investigation into the gap among Indigenous students reveals a potential connection to broader socio-economic issues impacting Indigenous communities.

In Chapter 2, I study the impact of switching post-secondary majors on earnings. To address the endogeneity of switching, I employ a doubly-robust matching estimator to create a credible counterfactual group for switchers. Switching has a greater impact on the earnings of women, with women experiencing gains (losses) as large as $15,500 ($23,000) conditional on initial major. These results highlight the importance of major-choice as it relates to labour market earnings.

Finally, Chapter 3 investigates the gender gaps throughout the application process to undergraduate engineering and computer science programs. While we observe large gender gaps in applications to both programs, we also observe gender gaps in offers to engineering programs and acceptances to computer science programs. This suggests that both programs may face unique challenges in achieving gender parity in enrollment.