Association for Public Policy Analysis & Management

Little research has been able estimate the effect of participation in career and technical education (CTE) on high school graduation and college going. Earlier work has explored economic returns to participating in CTE in high school (Neumark & Rothstein, 2006; Bishop & Mane, 2005), while more recent work has focused on how CTE participants in community college have fared in the workforce (Kurleander, Stevens, & Grosz, 2014; Trimble & Xu, 2014). Only a few studies can provide plausibly causal estimates of the effect of CTE on academic outcomes, but findings do not point to a consistent answer (Kemple, 2008; Dougherty, 2015), and none look at college going. In this paper we use rich administrative data from Arkansas and an instrumental variables approach to estimate the causal effect of high school CTE course taking on high school completion, initial enrollment in college, employment and wages immediately after high school. We also provide estimates for traditionally marginalized groups of students who are disproportionately represented in CTE; youth from lower-income families and students with disabilities.

We use rich administrative data from Arkansas that includes data from K-12, higher education, career and technical education, and the department of labor. These data include student demographics, test scores, and course enrollment. Using three graduation cohorts we follow over 100,000 unique students longitudinally through high school and into their first year after on-time graduation.

As of the class of 2014, students in Arkansas were required to complete the SmartCore which requires all students to take at least six career-oriented courses while in high school. While Advanced Placement and language courses can satisfy some of these requirements, the bulk of qualifying courses offered are classified as CTE classes. As schools sought to ensure an adequate supply of qualifying classes, labor-market conditions, and variation in the available supply of teachers generated arguably exogenous variation in the type and availability of CTE coursework in a given year and school. We use within-school and across year variation in the number of CTE courses offered as an instrument for CTE course taking, and then use the exogenous portion of the variation in course taking to estimate the effects of taking an additional course on student outcomes. In all analyses or first-stage F statistic on the instrument exceeds a value of 40.

Our findings show that being induced to take an additional CTE course in high school is associated with higher probabilities of on-time graduation (two percentage points), higher probability of enrolling in a two- or four-year year college (three percentage points), and higher initial wages (four percent).  Overall graduation effects are larger (and appear to drive overall results) for students with disabilities and students from lower-income families, while college enrollment gains appear focused among higher-income and students without disabilities. More CTE coursework induced higher wages in among students with disabilities compared to typically developing peers, whereas wage effects were not different by family income, with the former likely driven by higher probabilities of being employed immediately after high school.