Supplementary Materials for Three Essays on Career and Technical Education at Community Colleges

Abstract

The three essays comprising this dissertation focus on different aspects of the career and technical education (CTE) pipeline at American community colleges. Each essay involves an original empirical analysis using longitudinal administrative data from a state system to study the behaviors and outcomes of community college students. The first essay examines the extent to which first enrolling in noncredit education instead of credit-bearing coursework affects community college students’ academic progress, completion, and transfer outcomes over time. Using administrative data on nearly 2 million community college students from California, I employ selection-on-observables empirical methods to contrast the long-run academic outcomes of noncredit and credit students after adjusting for compositional differences between the two groups. Within 8 years of initial enrollment, noncredit students completed 12 fewer credits, had 7 percentage-point (pp.) lower rates of credential completion, and 7 pp. lower rates of transfer to 4-year colleges than they would have had they first entered credit education. The results are robust to sensitivity analyses involving selection-on-unobservables, alternative estimators, alternative measurement windows, and alternative subsamples, including students who participated in CTE coursework and students who explicitly reported aspirations for credit coursework and credentials. The second essay investigates how fluctuations in local unemployment rates and earnings levels of local workers influence the enrollment behavior of students in CTE educational pathways. The analysis interrogates the theoretical proposition that improving (worsening) labor-market conditions reduce (increase) community college enrollment by driving the opportunity cost of remaining enrolled upwards (downwards). Using administrative data on nearly 190,000 Colorado community college students and publicly accessible information on labor-market conditions, I implement event-history modelling to assess the relationship between local labor market conditions and community college students’ stop out, reenrollment, and enrollment intensity behaviors over time. I find that higher unemployment rates were weakly associated with male and female students being less likely to stop out, being more likely reenroll, and attempting more credits while enrolled. Results pertaining to unemployment rates comported with theory and were also robust to alternative model specifications. I also find that higher local earnings levels were associated with students being less likely to stop out and more likely to reenroll. These results are both directionally misaligned with theory and sensitive to model specification. The final essay, co-authored with Dr. Peter Riley Bahr, uses administrative data on nearly 130,000 noncredit students from 2-year colleges in Texas to examine variation in the economic returns to participating in noncredit occupational training on the basis of gender, field of study, type of training, and other factors. After implementing implement individual fixed effects regression models, we find that noncredit occupational training was associated with modest earnings gains for students, approaching $540 per quarter 5 years after enrollment. The magnitude of these earnings gains are noteworthy given students’ typically brief enrollment spells. We also find that earnings gains for noncredit students were strongest in fields such as transportation and engineering sciences, each averaging more than $1,500 per quarter after 5 years, while gains in other fields such as business and marketing and cosmetology, culinary, and administrative services were statistically indistinguishable from zero. Returns were additionally stronger for male students, for contract training participants, and for students who engaged in longer durations of training.