Two years ago, Leila Clark ’18 proposed a referendum that would require the Undergraduate Student Government (USG) to create a committee with the Interclub Council (ICC) that would collect demographics of eating clubs’ memberships.
In the wake of the referendum’s approval, USG dragged its feet, and the ICC claimed this information would threaten individuals’ privacy if it were made public. Meanwhile, the administration — who had this data the entire time — kept quiet.
Over the course of six months, I conducted an empirical investigation into the demographic composition of Princeton’s eating clubs’ members. I utilized publically accessible information to estimate patterns of income, gender, major, varsity status, hometown and home state.
The findings show that people join clubs that have members similar to themselves.
Ivy is indeed the most international club. Colonial is mostly engineers and science majors. Cottage has the highest percent of athletes, with Cannon a close second.
Through Tigerbook, I found members’ hometowns and then pulled such locations’ median incomes from U.S. Census data.
Bicker club members who live in the United States are from towns with median household incomes that are $2,832 greater than those of sign-in club members.
Athletes tend to follow other athletes. Three clubs — Cottage, Cloister, and Cannon — have 75 percent of all varsity athletes on the Street.
Of the international students in eating clubs, 73 percent are in Bicker clubs.
Students from public schools are a minority in at least one eating club’s class. Fifty-eight percent of Tiger Inn’s Class of 2019 attended an independent day, independent boarding, or religiously affiliated school, according to a newsletter from this past spring. In contrast, 41 percent of Princeton’s Class of 2019 attended these types of schools. The median cost of TI members’ private secondary education is $36,685 per year.
Compared to all undergraduates, students concentrating in science, technology, engineering, and mathematics (STEM) and those in Bachelor of Science in Engineering (BSE) programs are overrepresented in sign-in clubs. Sixty-five percent of sign-in members study STEM and 35 percent are engineers.
School-wide, 50 percent of undergraduates study STEM, and 26 percent are engineers.
STEM concentrators are significantly underrepresented in the three oldest eating clubs. Only 24 percent of Ivy, 26 percent of Cottage, and 39 percent of TI study STEM subjects.
In 1967, The Harvard Crimson’s reporter at Princeton wrote, “there is a high proportion of math and science men in the bottom clubs” compared to the “top clubs.” At that time, he considered Charter, Cloister, Terrace, and Quadrangle to be the bottom, and Ivy, TI, and Cottage to be three of the clubs at the top.
Fifty-one years later, that observation still holds true despite the University’s admission of women and an increasing number of STEM students.
The controversy surrounding this issue doesn’t make sense. Before Stanford University’s eating clubs closed, they had been reporting their members’ demographics since the 1960s.
Princeton is behind the curve by a half century.
Everyone should be aware of eating clubs’ demographics because they impact students.
First, they can influence school policy. Currently, all juniors and seniors on financial aid receive a $2,000 boost each year to help pay for upperclass dining options.
Although this policy is helpful, many students still fall into an awkward financial doughnut hole: They come from families that don’t make enough money to easily stroke a $10,000 (or even $3,000) check to an eating club but also don’t demonstrate enough need to receive full financial aid, which would cover their bills.
Concrete data that’s published each year would show the effectiveness of administrators’ initiatives and tell officers if they should cut costs to make their clubs more accessible.
Second, the Street’s divisions could greatly impact students’ futures. Earlier this year, Kevin Carey wrote in The New York Times, saying “eating clubs are where many upper-income marriages begin.” His column demonstrated that Ivy League students — including Princetonians – of lower socioeconomic status are less likely to get married than their affluent peers.
But beyond marriage, membership in an eating club may provide different prospects in members’ careers, networks, earnings, charity, levels of tolerance, and more. Students could have very different futures depending upon the demographics of the eating club that they join.
Ultimately, we — Princeton students — are responsible for the Street’s stratification. We are the ones who judge others based on superficial traits, segregate ourselves by socioeconomic status, and create hierarchies of social prestige.
The student body needs to look in the mirror and understand how it segregates itself.
Interclub Council and University Responses
My investigation also uncovered administrators’ previously unknown practice of sharing demographic data with eating club presidents.
In a statement and interview, ICC chair Hannah Paynter ’19 said the University provides eating club graduate boards and undergraduate presidents with “confidential information” on club demographics.
Paynter said the ICC requested the data after meeting with Princeton’s Executive Director for Planning and Administration Christopher Burkmar ’00.
Burkmar deferred comment to the Office of Communications.
University spokesperson Ben Chang confirmed in an email that administrators share aggregate demographic data with the eating club presidents and graduate boards.
“This practice, which began last year, helps the clubs better understand the demographics of their members and supports recommendations of the 2017-18 Task Force on the Relationship between the University and the Eating Clubs,” he wrote.
The University regularly reviews demographic data, “to develop initiatives, enhance programs, and review student engagement.”
He said data is classified as “Confidential” under the University Information Security Policy. This level permits the sharing of confidential Information when, “necessary to meet the University’s legitimate business needs.”
Chang affirmed that the University did not violate any federal laws or disclose data on individual students.
“We thought that we were trailblazing by trying to get that kind of data because it’s private,” Paynter said. Although she acknowledged that the presidents would eventually inform undergraduates of this development, she remains opposed to making the data available to them.
Paynter noted that the ICC has already formed partnerships with the Princeton Latinos y Amigos, Freshman Scholars Institute, and Scholars Institute Fellows Program.
In reference to stereotypes, Paynter said, “I think the most frustrating thing is that the clubs are already putting forth these concerted efforts to change their make-ups and ensure no one feels unwelcome”.
* * *
Overall, eating club members hail from towns that collectively have a median household income of $76,543 and an average of $86,103.
For reference, the median household income in the United States is $61,372, according to the U.S. Census.
An estimated 54 percent of all eating club members are male and 46 percent are female. Slightly less than 30 percent are varsity athletes. Twelve percent are from abroad. Nearly half concentrate in STEM subjects. About a quarter are pursuing a BSE degree.
For all undergraduates, the University reports that 51 percent are male and 49 percent are female. Thirteen percent are international students, 18 percent are varsity athletes. There are no major differences in the gender ratios between selective and non-selective clubs.
Ivy Club (est. 1879)
Members come from towns with a median household income of $76,082 and an average of $85,188. Twenty-nine percent have hometown household incomes over $100,000. Thirty-six percent of hometown household incomes are below the U.S. median.
It is critical to keep in mind what I will call the “city effect”; large cities have a wide range of incomes, meaning that the incomes of families in the upper income range will be imprecisely represented by the median for the city overall. Given the reasonable assumption that Princeton students are generally pulled from the wealthier echelons of their hometowns, this analysis likely underestimates the family incomes of students from major cities.
Students from Chicago, Los Angeles, New York City, and Washington, D.C. comprise 22 percent of Ivy’s American membership.
Males and females are split 50-50. International students are 28 percent of the club. Of Ivy’s international students, 45 percent are from the United Kingdom, and 30 percent are from London alone.
Thirty-three percent of Ivy members are varsity athletes.
One quarter concentrate in STEM subjects, and 13 percent are engineers. Slightly more than 40 percent are Wilson School, history, economics, and politics (WHEP) concentrators.
Ivy president Mimi Asom ’19 did not respond to a request for comment.
Cottage Club (est. 1886)
Like Ivy, Cottage’s income estimate is also probably an underestimate. Twenty-three percent of American members of Cottage reside in cities with a population over 300,000.
Cottage members come from towns that have a median household income $75,260 and an average of $87,494. Thirty-four percent have hometown household incomes above $100,000. Forty-one percent have hometown household incomes below the U.S. median.
Males are 58 percent of the club, and 43 percent are female. International students compose 14 percent of members, of whom 54 percent are Canadian. Seventy-eight percent are varsity athletes.
Twenty-six percent study STEM subjects, but only 12 percent are engineers. Fifty-five percent are WHEP concentrators.
Cottage president Casey Swezey ’19 did not respond to a request for comment.
Tiger Inn (est. 1890)
TI members come from towns that have a median household income of $77,680 and an average of $88,533. More than 33 percent had hometown household incomes above $100,000. Thirty-eight percent had hometown household incomes below the U.S. median.
Half of members are male, and half are female. Almost 7 percent are international students. Twenty-nine percent are varsity athletes.
About 40 percent study STEM subjects, and 23 percent are engineers.
TI president Maggie McCallister ’19 did not respond to a request for comment.
Cap & Gown Club (est. 1891)
Cap and Gown members come from towns that have a median household income of $75,313 and a mean of $82,943. Thirty-one percent have hometown household incomes above $100,000. Thirty-seven percent have hometown household incomes below the U.S. median.
Fifty-nine percent are male, and 41 percent are female. Around 12 percent are international students. Eight percent are varsity athletes.
Fifty-three percent of members study STEM subjects, and 31 percent are engineers.
Cap president RJ Hernandez ’19 did not follow up on a request for comment.
Colonial Club (est. 1891)
Colonial members come from towns that have a median household income of $81,458 and an average of $87,809. Thirty percent have hometown household incomes over $100,000. Thirty-two percent have hometown household incomes below the U.S. median.
Fifty-nine percent are male, and 41 percent are female. One-tenth are international students, of whom 33 percent are Canadian. Less than 1 percent are varsity athletes.
Three-quarters of members study STEM subjects. About half are engineers.
Outgoing Colonial president Kimberly Peterson ’19 did not follow up on a request for comment.
Cannon Dial Elm Club (est. 1896; reopened 2011)
Cannon Dial Elm members come from towns that have a median household income of $77,680 and an average of $88,948. Thirty-seven percent have hometown household incomes above $100,000. Thirty-six percent have hometown household incomes below the U.S. median.
The split between males and females is 47 percent and 53 percent, respectively. Eight percent are international students, of whom 63 percent are Canadian. More than three-quarters are varsity athletes.
More than one-third study STEM subjects, and 16 percent are engineers. Forty percent are WHEP concentrators.
Cannon president Julia Haney ’19 did not follow up on a request for comment.
Charter Club (est. 1901)
Charter members come from towns that have a median household income of $69,626 and an average of $82,426. A quarter have hometown household incomes over $100,000. Thirty-five percent have hometown household incomes below the U.S. median.
Sixty-five percent are male, and 35 percent are female. Four percent are international students. One member is a varsity athlete.
Eighty-eight percent study STEM, and 65 percent are engineers.
Outgoing Charter president Conor O’Brien ’19 did not follow up on a request for comment.
Quadrangle Club (est. 1901)
Quadrangle members come from towns that have a median household income of $71,830 and an average of $81,256. Twenty-nine percent have hometown household incomes above $100,000. Forty-two percent have hometown incomes below the U.S. median.
Fifty-three percent are male, and 47 percent are female. Fourteen percent come from abroad.
Seventy percent study STEM subjects. One-third are engineers.
Outgoing Quadrangle president Sarah Spergel ’19 did not follow up on a request for comment.
Tower Club (est. 1902)
Tower members come from towns that have a median household income of $83,205 and an average of $89,256. Forty-five percent have hometown household incomes over $100,000. Forty-three percent have hometown household incomes below the U.S. median.
Fifty-three percent are male, and 47 percent are female. Seventeen percent are international students, of whom 66 percent hailed from South or East Asian countries. Three-and-a-half percent are varsity athletes.
Sixty-one percent study STEM subjects, and 35 percent are engineers. Twenty-seven percent are WHEP concentrators.
Tower president and ICC co-chair Rachel Macaulay ’19 declined to comment.
Terrace Club (est. 1904)
Terrace members come from towns that have a median household income $68,304 and an average of $81,352. Twenty-three percent have hometown household incomes over $100,000. Thirty-six percent have hometown household incomes below the U.S. median.
Fifty-one percent are male, and 49 percent are female. Thirteen percent are international students. Two percent are varsity athletes.
Fifty percent study STEM subjects. Nineteen percent are engineers.
Outgoing Terrace president Elizabeth Yu ’19 did not respond to a request for comment.
Cloister Inn (est. 1912; reopened 1977)
Cloister members come from towns that have a median household income of $83,958 and an average of $87,390. Thirty-nine percent have hometown household incomes over $100,000. Thirty-five percent have hometown household incomes below the U.S. median.
Males are 60 percent of the club, and females are 40 percent. Seven percent are international students, of whom over half are from the United Kingdom. Athletes are 64 percent of club members.
Fifty-seven percent study STEM subjects, and one-third are engineers.
“One of Cloister’s main recruitment goals this past year was to expand recruitment to non-varsity athletes,” outgoing Cloister president and ICC chair Hannah Paynter said in an interview. She explained that the club used to be smaller and hovered around consisting of 70 percent varsity athletes before it grew this year.
“I am happy to see that you found that number go down,” said Paynter, “because we found that too.”
I obtained the eating clubs’ membership rosters from the University’s meal exchange website. Paynter confirmed the list is accurate.
Tigerbook — a student-produced and student-maintained database service to which only University undergraduates have access — provided members’ concentrations, hometowns, and profile pictures.
I used the most recent U.S. Census Bureau data on area median household incomes (AMI) of American eating club members’ hometowns to estimate their socioeconomic levels. This proxy is most accurate when confined to a small, specific area, but of course, some members’ hometowns were listed as big cities, such as Philadelphia.
The New York Times found from anonymous tax records that Princeton students have average family incomes of $186,100. None of my income estimates came close to that of the Times’s values, so using the AMI probably significantly underestimates club members’ household incomes.
The Princeton athletics website provided rosters of all varsity teams to determine whether students were athletes.
I defined STEM concentrations as fields of study that the Office of the Dean of the Faculty grouped into the Natural Sciences and Engineering and Applied Science academic divisions.
A spring 2017 TI newsletter included a list of names and high schools for the newly accepted members of the Class of 2019. Some private schools have day costs and higher boarding costs. In such cases, I used the boarding cost in calculations.
I inferred members’ gender by their name and appearance. My crude method was, well, crude. It skipped over transgender, gender non-conforming, and other individuals who identify differently than their facial appearance normatively suggests.
Twitter also guesses users’ gender based on their profile information. My investigation’s gender estimates may be off by a few percentage points — though likely not by any huge margin. If this method is accurate enough for Twitter, then it is for my purposes as well.
I estimated the clubs’ racial demographics by last name and appearance. I will not release the specific results at this time because they suffer from my own biases and the same pitfalls as the gender estimates.
Liam O’Connor is a junior geosciences major from Wyoming, Del. He can be reached at firstname.lastname@example.org.