A vintage photo of a dorm at UW-Madison.
Lee Balterman / The Life Picture Collection / Getty

In 1926, the University of Wisconsin published a brochure advertising its new men’s dormitories. “Here … the man from the well-to-do home and the man who tends furnaces to buy his text-books will learn respect for each other across a common table,” the booklet read, “and the son of banker and farmer will find mutual understanding, of a winter’s evening, in give and take to the crackling of logs in a wide fireplace.”

A page in the University of Wisconsin’s 1926 brochure (Image courtesy of the UW-Madison Archives, #S1369)

The brochure reflected evolving attitudes toward college-student housing at the time. College deans had begun to worry that fraternities, which rose in popularity toward the end of the 1800s, were undermining the student experience, according to Carla Yanni, a professor of art history at Rutgers University and the author of the new book Living on Campus: An Architectural History of the American Dormitory. They were full of affluent white students, while many poorer students lived in off-campus boarding houses. Greek life was effectively promoting socioeconomic and ethnic segregation. Dorms, in turn, emerged as a sort of “social leveler,” Yanni told me.

Reflecting on that brochure today elicits a sense of déjà vu (minus the furnace-tending and log-crackling). In recent years, colleges across the country have expressed renewed interest in their democratic mission to build a diverse student body and expose students to their peers from different backgrounds. That duty entails enabling young adults to practice “the fine art of getting along with their fellows,” as Yanni puts it in her book.

Left to their own devices, many incoming freshmen are inclined to preselect their roommates—and when they do, many opt for peers they already know or to whom they assume they will relate. The advent of social-media platforms such as Facebook in the mid-2000s, some residence-life administrators told me, fueled that impulse. Many new students join groups online to connect with one another before freshman year, and they submit roommate requests based on the information they glean there. But in recent years, many colleges have started to restrict—if not altogether remove—incoming freshmen’s say in the roommate-matching process. Schools want to maintain careful control over that process so that they can fulfill higher education’s democratic mission at the micro-level of the dorm room.

Most residential colleges in the United States have some sort of official process for roommate matching in place. In the Association of College and University Housing Officers–International’s 2018–19 survey, 56 percent of respondents said their institution has incoming freshmen fill out a questionnaire to determine dorm assignments. Another 34 percent said their school asks new students to create a profile on the college’s virtual roommate-matching platform, some of which let students match themselves, some of which don’t. And while the most common criteria schools used to pair students were personal habits (such as cleanliness and smoking status), slightly more than half of the survey’s respondents reported considering characteristics such as academic interests and hobbies. Some colleges are so meticulous, they have incoming students take personality tests. North Carolina’s Davidson College, for instance, has for decades incorporated into its freshman-roommate assignments each student’s personality traits as determined by the Myers-Briggs test—especially where the individual falls on the extroversion-introversion scale.

Many schools, including Davidson, give students the option to request a specific roommate in advance, and consider those requests on a case-by-case basis. But some schools have recently stopped allowing student input on the process. For example, in 2013, NYU stopped allowing first-year students to choose their roommates in order to mix people from different parts of the country and the world. (NYU’s current policy says the university will consider but not guarantee students’ preferences.) Last year, Colgate University, a liberal-arts college in upstate New York, similarly stopped letting incoming freshmen preselect their living buddies. Some schools, such as Tufts University and Duke University, now randomly assign roommates to incoming freshmen.

Many incoming freshmen, understandably, want to choose their own roommates. But the common practice of students sorting themselves on social media is regarded skeptically by some student-affairs administrators. “There is a limitation to [what you can learn by] looking at Facebook, looking at profiles, looking at posts,” says Carlos Gonzalez, who oversees residential services at Northwestern University.

And the roommates students want might not always be the best roommates for them, in schools’ eyes. As the Dartmouth College economist Bruce Sacerdote, who’s studied the effects of random roommate assignments, wrote last year in an analysis for The Conversation, “Natural instincts do not always benefit us in the long run.” “As human beings we naturally gravitate towards our comfort zone and find peers who look a lot like ourselves,” he went on. But as campus administrators have long argued, people (ought to) attend college not only to get a degree, but also to transcend their comfort zone—by engaging with people, disciplines, and ideas that diverge from what they are used to.

“I absolutely do not think we would have found each other [if we’d been allowed to choose],” Rachel Harshaw says of her college roommates. Harshaw graduated from New York’s Hamilton College in 2017. Incoming Hamilton freshmen fill out a detailed questionnaire, and a team in the college’s residence-life office pairs each student with one or more living buddies. This system assigned Harshaw to a suite of six freshmen total, none of whom appealed to her when she did her pre-matriculation internet sleuthing. One of them was a softball player who looked like she would be the outgoing, Greek-life, jock type; another was a popular, blond theater lover. Neither seemed like a potential friend to Harshaw, who’s shy; she was “terrified,” she told me.

“I’m a picky person; I’m a skeptical person,” said Harshaw, who now works as a program manager at the MGH Institute of Health Professions in Boston. “Going into my freshman year, I was so wanting to pick my own roommate.” Today she considers those suitemates her best friends. She currently lives with two of them in Boston. “I was so upset that they were being handed to me, and I was so nervous about it,” Harshaw said. “But at the end of the day, if I hadn’t taken those friendship risks, it’s kind of scary to think about where I’d be.”

Roommate assignments are a high-stakes endeavor, given how important that relationship can end up being. The bond can be so tight, one study found, that the two students—if they’re friendly—tend to adopt each other’s speech patterns. Even roommates who aren’t super close can have a profound influence on each other, if only because of how much time they spend together: While only 37 percent of the student participants in a 2006 study identified their roommate as their “best friend,” close to half of the respondents said they spent more time with their roommate than with any other friend. An immense body of research demonstrates the impact a student’s living buddy can have on that student’s grades, career choice, beliefs, and social habits (such as drinking behavior).

These stakes are a big reason colleges put so much stock into the roommate-matching process. Schools have an interest in ensuring that students thrive, and whom a freshman lives with can have an immense impact on her long-term success in college and beyond. As Sacerdote alluded to, a peer who on paper may seem appealing to a roommate-seeking freshman may not end up being the best match for that freshman—or may be so similar that living with her undermines the freshman’s ability to experience new things in college. “The purpose of living on campus is to be exposed to more diversity in all arenas of life,” Northwestern’s Gonzalez says, “whether it’s ethnicity or culture or socioeconomic status or where we’ve grown up.”

It’s hard to suss out how students’ satisfaction with college-assigned roommates stacks up against their satisfaction with self-selected ones, because apples-to-apples data don’t exist. But statistics from individual schools that are more hands-on with roommate matching indicate, at the very least, that most students are satisfied with their roommates, even when they don’t get to pick. Students at Northwestern and Davidson, for example, generally seem to buy into their college’s roommate-matching strategy, with an estimated one in 10 incoming freshmen at each school opting to request a dorm buddy. At Davidson, the majority of students go on to live with their freshmen roommates as upperclassmen, according to Jason Shaffer, the college’s associate dean of students.

And in a survey conducted by the research firm Skyfactor during the 2015–16 school year among 20,000 students at 15 participating institutions with varying roommate-assignment protocols, more than half of the respondents said they were happy with the roommate to whom they were assigned as a freshman; just one in 10 of them requested a roommate change that year.

Perhaps that’s because students, too, understand the stakes of a roommate relationship. After all, as Yanni noted, “a lot of students come to college for the social life, not for the academics.”

“College was so much for me, and for my roommates, about catapulting ourselves outside of our comfort zones,” Harshaw said, “and there were a lot of instances of catapulting that I was not happy with at the time [I started freshman year]. But, I mean, that’s the point of college, and they really enriched me and made me the person I am today.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.