Our study details that – controlling for academic quality, high school and family background – these high-school students from Milan, Italy, were accidentally and randomly assigned to classes with a larger or smaller share of boys and girls. This accidental feature had a significant and long-lasting effect on the students: evidence suggests that a larger share of students of the same gender in the class increased the probability of choosing a high-earning college major.This finding is true for both men and women. In particular, as shown in below, women in high-school classes with more than 75% women had a 5-6% higher probability to choose high-earning majors than women in classes with a high percentage of men. Similarly, men in classes of more than 75% of women had a 6-7% higher probability of choosing high-earning majors.
Interestingly, these effects seem last longer for women. Those who attended classes with a large percentage of women and then choose high-earning majors also graduated from them and performed better than or as well as other women in those majors. Hence the long-run effect for women was a higher probability of graduating from high earning majors and therefore higher expected wage. On the other hand men who attended classes with high percentage of men had a higher probability of enrolling in high-earning majors but were also more likely to drop out. Thus, they did not have a higher probability of graduating from these potentially high-earning majors. Moreover women from classes with a large female percentage had higher wages on the labour market relative to other women above and beyond, that is, controlling for, the effect driven by their programme of study.
Overall, the expected wage premium for a woman who attended a high-school class with female share above 75% relative to an identical one who attended a class with a female share below 25% was €1,110 (about $1,470). This wage premium represents an expected 5.3% wage increase, as large as the return of one extra year of schooling in Italy!
Explaining the effect
The high school years (13 to 18 in Italy) are crucial for young women in shaping their preferences and self-confidence. In part, girls may shape those attributes by comparing themselves to male students. If there are fewer men for comparison, they may more readily consider themselves to be good at math, or interested in science, or fascinated by technology, attributes our society stereotypically assigns to men. Recent experimental research shows that gender-specific roles may be perpetuated at the high-school level (see Mobius et al. 2012 on how women and men adjust their beliefs about themselves). With fewer men around, some women are freer to choose to pursue what they love and what they are good at.
Policy implications: gender separation in high-school classes
Interestingly, the policy implications of these findings are straightforward. If an objective of schooling is to increase women’s career opportunities and thereby their salaries, our results would suggest that gender-separated classrooms would be an effective step in the right direction. Incidentally, men will also benefit, being encouraged to enroll in high-earning majors. Gender-separated classrooms would increase the probability of choosing high-earning majors for both women and men. At the very least, schools could offer students the possibility of choosing single-sex classes. Or they could offer students the possibility of entering a random lottery to be in single-sex classes and, through observing their subsequent college choice, we would be able to continue to test the effect of peer gender.
Anelli, Massimo and Peri, Giovanni (2013), “The Long-run Effects of High-School Class Gender Composition”, National Bureau of Economic Research, working paper 18744, January.
Black, Sandra E and Spitz-Oener, Alexandra (2007), “Explaining women’s success: technological change and the skill content of women’s work”, VoxEU.org, 1 September.
Mobius, Markus M, Niederle, Muriel, Niehaus, Paul and Rosenblat, Tanya (2012), “Managing Self-Confidence: Theory and Experimental Evidence”, September.