Gender, Race, and Economic Status along the Computing Education Pipeline: Examining Disparities in Course Enrollment and Wage Earnings

Background and Context: Inequities in computing education have been identified based on gender, race, ethnicity, and economic status. However, extant quantitative research tends to treat demographics as siloed categories instead of accounting for the fact that many students are members of multiple minoritized groups. There is also a conspicuous lack of research that examines equity issues at multiple stages of the education pipeline. Objectives: Our research questions asked: 1) How are gender, race/ethnicity, and economic status related to the likelihood that students will enroll in computing courses in high school? 2) How are these factors related to the likelihood of enrollment in college computing courses? 3) How do these factors relate to wage earnings for students who majored in computer science in college? Method: This study analyzed education and workforce data in the United States using multilevel logistic and linear regression analyses to identify disparities among students from multiple minoritized groups at three stages of life: high school (N=135,961), college (N=199,230), and career (N=1,251). Findings: Compounding course enrollment disparities were present at the high school level for students who are Black or Hispanic/Latino/a and female, economically disadvantaged, or both. Similar results were observed at the college level. Only gender was a statistically significant predictor of wage earnings five years after college graduation, but results of this analysis may be attenuated by the fact that relatively few students who were members of multiple minoritized groups graduated with a computing degree. Implications: These findings demonstrate the importance of considering students’ intersecting identities when assessing equity in computing education. They also lay groundwork for understanding how early inequities persist across the education pipeline and how understanding disparities at one stage can inform interpretation of disparities at subsequent stages.