Old boys' clubs and upward mobility among the educational elite

What is this research about and why did you do it?

We ask how membership in exclusive social groups affects access to top positions in the economy and society and, if so, who can join. 'Old boys’ clubs' are central to qualitative accounts of elite closure — the idea that social groups restrict access to opportunity on the basis of shared traits and experiences. But there is not much quantitative evidence on who joins exclusive groups and how much it matters in the long run. This is largely because data on actual and potential members of exclusive groups are hard to get.

‍

How did you answer this question?

We gather archival data on undergraduate students at Harvard University in the 1920s and 1930s. These data track students’ academic and social lives in college and their professional lives as adults. We use the data to describe who joins exclusive campus groups known as “final clubs” and to compare long-run outcomes for members and non-members. We then study a policy that randomly assigned first-year students to different campus dorm rooms. We ask whether students assigned to live in close proximity to high-status peers are more likely to join exclusive social groups and pursue high-status careers.

‍

What did you find?

We have three main findings. First, we show extensive sorting into final clubs on the basis of pre-college social status. Students from exclusive private schools (US equivalents of, say, Eton) are overrepresented in final clubs, while religious minorities and academic top performers are almost completely absent. Second, we show that the final club earnings premium is large. Club members earn 32% more than non-members and are more likely to work in lucrative finance jobs and join country clubs (see figure). Third, we show that random assignment to high-status neighbours causes high-status students, but not low-status students, to behave in high-status ways—joining final clubs, joining country clubs, and working in finance. It therefore makes gaps by baseline social status larger.    

Finally, we zoom out to consider the full 20th century. Patterns from pre-WWII cohorts persist. High-status, low-achieving Harvard students remain more likely to join final clubs and work in finance, and elite university students from the highest income families (the top 0.1%) earn 76% more as adults than students from families with incomes that are high but outside the very top (the top 5-10%).

‍

This graph displays mean wage earnings for Harvard students starting college in the 1920s and 1930s, splitting by membership in campus final clubs and first-year academic performance. Academic performance improves as one moves from the left to the right. The top academic category is omitted from the final club member series because too few final club members achieve top grades.

‍

What implications does this have for the research on wealth concentration and economic inequality?

Integrating exclusive social groups is crucial for integrating top positions in the economy, but it is hard to do. Even prolonged residential exposure to high-status peers is not sufficient to help lower-status students join old boys’ clubs and attain the career outcomes with which membership is correlated. Elite universities have an important role to play in identifying talented students from underrepresented groups and helping them advance to leadership roles, but doing so effectively requires attention to social integration in addition to academic performance.

‍

What are the next steps in your agenda?

We would like to learn whether access to elite institutions at younger ages yields larger long-run gains for lower-status children, to understand how social ties help produce other outcomes, like innovation and governance, and to ask what policies might succeed in socially integrating elite institutions.

‍

Citation

This paper can be cited as follows: Michelman, V., Price, J., and Zimmerman S. D. (2022) 'Old Boy's Club and Upward Mobility Among the Educational Elite.' Quarterly Journal of Economics, 137(2), pp. 845-909.