*Names in bold indicate Presenter
My analyses comprise two distinct pieces. First, I take advantage of a uniquely rich dataset to characterize students in different fields of study. I use individual-level data from the University of North Carolina that is linked to K-12 data at the North Carolina Education Research Data Center. The descriptive analyses provide insight into the types of students that select particular majors. I describe how students invest in skills over time, and segment the analyses by categories such as socioeconomic status, sex and race. The dataset spans 2000-2008, so I can produce secular trends for students’ course-taking behavior and major choice during this period.
The second contribution relates students’ major choice to changes in local labor market conditions. The standard approach in this literature is to model the decision as a rational process that depends on students’ ability, tastes and expectations of future earnings. The models make a strong assumption that students form earnings expectations for their chosen major and all counterfactual majors. In practice, students are not well informed about future earnings and their subjective estimates are often quite misguided. I relax the informational assumptions and suppose that students adopt heuristics in making major choices. For instance, their choices may be swayed by local wage shocks in occupations related to their preferred majors. I use different university campuses and exploit variations in local wages to assess whether these conditions matter for major selection. Preliminary reduced-form analyses examine the effect of demand shocks on major choice during the Great Recession.