Gender separation and academic achievement in higher education : evidence from a natural experiment in Iran

In 2011, a large university in Tehran launched a policy of gender separation at classroom level without publicly announcing it beforehand. The current paper utilizes this natural experiment to identify the causal impact of participation in single-sex versus mixed classrooms on students’ achievement. Despite the vast yet inconclusive literature on single-sex schooling, this paper addresses the dearth of the research in the context of higher education as well as the context of Muslim-majority countries where single-sex education is prevalent. Empirical findings show that when students’ characteristics and educational competencies are taken into account, attending a single-sex classroom improves both males’ and females’ average performances by around 0.36 standard deviation. While the academic benefit for females does not depend on their ability level, the effect is considerably heterogeneous among males with different initial ability. Nearly all positive effect for males is driven by upper-medium-ability male students performing significantly better in all-male classrooms.