Association for Public Policy Analysis & Management

We use data from Missouri to examine the role of high school math and science courses on graduation with a science, technology, engineering, or mathematics (STEM) degree. A robust literature establishes that increased high school and math coursetaking corresponds to college success and a technical field of study. Inference from these relationships are subject to bias, however, since students’ own course choices are endogenous to major choice in college. We provide a cleaner, causal estimate of the effect of access to STEM courses in high school on major choice. We find that while endogenous course-taking at the individual level positively predicts STEM outcomes in college, access to more STEM courses in high school has no observable overall effect on whether students pursue or complete a 4-year STEM degree. We further explore differences in this relationship by race, ethnicity, and gender.

The data used in this study is administrative microdata on 14 cohorts of students who graduated from a public high school and matriculated into a public Missouri college from 1996 to 2009 (N ≈ 155,000). Using panel data from state high schools, we construct a measure of the number of courses available in each high school in each year in addition to the actual courses taken by individual students. We track each student in the Missouri system and determine whether she graduated within six years after entry, and if so, her degree. In addition to regression adjusted estimates that control for a robust set of characteristics available in the data, we employ two panel data methods to estimate causal effects. First, using high school fixed effects, we leverage our long data panel to isolate within-high-school variance in STEM course availability to estimate its effect on STEM outcomes in college. Next, we instrument for students’ own course taking with within-high school variance in course availability.

These findings are important because one of the strategies to meet the growing U.S. demand for a larger technological and scientific labor force is to increase the participation of women and underrepresented minorities in science and engineering fields. While research has demonstrated that high school preparation is a critical determinant in student success, many of the connections between high school level characteristics and college going rates have not been thoroughly studied. This study helps contribute to this important knowledge gap by connecting course access to STEM degree attainment using new and unique data that overcomes potential threats to internal validity inherent in past research.