Gender Peer Effects: Evidence from a Quasi-Random Classroom Assignment Policy∗

The literature on gender peer effects documents that exposure to opposite gender peers can lead to higher academic performances for boys but lower outcomes, especially in math, for girls. This result can be explained by two distinct exposure effects. One is inside-class time exposure (e.g. boy peers are disruptive and lower the quality of student-teacher interactions) and the other is outside-class time exposure effects that arise through social interactions with peers outside class periods (e.g. boy peers negatively influence other peers’ decisions on study efforts). I use student-level data from a Korean high school that, at the beginning of each academic year, quasi-randomly assigned students to either a single-sex or a coeducational homeroom. Students assigned to the same homeroom have the same class schedule throughout the year and take classes together in their homeroom — except for tracking subjects, math and English, where students are assigned to separate classrooms based on gender and academic skill in each subject. I find significant increases in girls’ native language test scores when the class is taught in a coeducational environment and a girl’s prior performance in native language was above her cohort’s average performance. However, I find no significant effect of assignment to a coeducational classroom on boys’ and girls’ academic achievements in tracking subjects. These findings suggest that gender peer effects are mostly due to exposure to opposite gender peers during class time. JEL Codes: I21, J16. ∗I am extremely grateful to Seema Jayachandran, Lori Beaman, and Jonathan Guryan for providing helpful comments and suggestions. I am thankful to Byungsuk Lee and Yoon Kyoung Lee for providing support with data collection. All errors are my own. †Email: [email protected] Web: sites.northwestern.edu/spa136