Effects of Peer Groups on the Gender-Wage Gap and Life After the MBA: Evidence from the Random Assignment of MBA Peers

Publication Date


Grant Type

Early Career Research Award


The paper explores the causes of the gender-wage gap among MBA students upon graduation and the evolution of this gap over the course of their careers. By exploiting the historical random assignment of MBA students to peer groups at an elite business school in the United States, this paper explores the effect of the gender composition of a student's peers on the starting salary at graduation, the field of study in business school, and on long-term outcomes up to 20 years after graduation. Preliminary results show that women who are randomly assigned to a peer group with a larger share of men have significantly higher salaries at graduation, are more likely to enter male-dominated industries and job functions in their first job after graduation, such as investment banking, venture capital and investment management, and are less likely to accept job offers in relatively more female-dominated industries, such as marketing and product management. Though the wage differences are not large in the first year post-graduation, the effects of the gender composition of the peer group on wages of female MBAs accumulate over time. In particular, in the first year after graduation, an increase in the share of male peers (from 0 to 100 percent) leads female students to choose industries at graduation that have an hourly wage that are $12 an hour greater on average, but ten years after graduation, are nearly $180 an hour greater on average than the hourly wages in the industries chosen by women with a lower share of men in their peer group. In addition, women with more male peers are less likely to have a career interruption, have fewer career interruptions, and are less likely to have children. These results reveal some underlying mechanisms through which the gender-wage gap, which starts out small at graduation, can accumulate to the documented magnitudes over the course of the lifecycle.