– Results reveal that foreign peers lower the probability natives graduate with a STEM degree, mostly inducing switches to Social Science.
– A 1 standard deviation increase in foreign peers reduces the likelihood of native-born students to graduate with STEM majors by 3 percentage points — equivalent to 3.7 native students displaced for 9 additional foreign students in an average course.
– STEM displacement is offset by an increased likelihood of choosing Social Science majors.
– The earnings prospects of displaced students are minimally affected, as they appear to be choosing Social Science majors with equally high earning power.
– Comparative advantage and linguistic dissonance may operate as underlying mechanisms.
– The probability of graduating with a STEM major is positively correlated with peer SAT math and negatively correlated with peer SAT verbal and High School GPA.
– Including foreign peer composition across all other classes did not change the results, indicating that the transmission of foreign peer impacts on STEM major choice occurs within introductory math classes as opposed to other courses.
2017 - Foreign Peer Effects and STEM Major Choice
This paper uses administrative student records from a public, land-grant university in California. The authors focus on introductory math courses, which are defined as the set of math courses that are listed as satisfying a university-wide quantitative course requirement and also satisfy mathematics prerequisites required for STEM majors.
The authors limit the analysis to new freshmen enrolled in an introductory math course in fall 2000 through fall 2006. The sample consists of 16,830 first-term native freshmen.
IV: Gender; Race/ethnicity; Nativity (country of origin); High school GPA; Course enrollment; Foreign peer exposure (measured within classes); SAT math, verbal, and combined scores
DV: The outcome variable is a dummy variable equal to 1 if the student graduated with a STEM major within 6 years from the first term of enrollment, and 0 otherwise.
Authors control for course-by-term indicators (σct) and course-by-professor fixed effects (σcp).