The Unwritten Rule
Stand on any commuter platform during rush hour and you'll see the same scene in every major city: hundreds of people within arm's reach of each other, phones up, eyes down, earbuds in, obeying an unwritten code that says you must not make eye contact, must not speak, and must treat the person six inches from your elbow as if they were furniture.
Nicholas Epley, a behavioral scientist at the University of Chicago, noticed this paradox on his own daily commute. Humans are among the most profoundly social species on the planet, yet packed into a train car at twice the density of a cocktail party, they act as if the other passengers don't exist. Either the silence was genuinely better, or people were systematically wrong about what would make them happy. He and Juliana Schroeder, then a doctoral student at Chicago Booth, designed nine experiments to find out.
The Commute Nobody Expected to Enjoy
In the first two experiments, Epley and Schroeder recruited commuters at Metra train stations and CTA bus stops in the Chicago area and randomly assigned them to one of three conditions: connect with a stranger near them, sit in solitude, or commute however they normally would. After the ride, each participant rated how positive, how happy, and how productive their commute had been.
The results ran in one direction only. Commuters assigned to connect with a stranger reported a significantly more positive experience than those who sat in solitude. They also rated their commute as no less productive, despite spending an average of fourteen minutes in conversation. The pattern held across trains and buses, among men and women, and regardless of commute length.
A separate group of commuters, asked only to predict how they would feel in each condition, expected the exact opposite: they forecast solitude as the most pleasant option and talking to a stranger as the least. The prediction and the reality pointed in completely different directions.
Why Everyone Stays Silent
Experiments 3a and 3b probed the mechanism. Commuters estimated what percentage of fellow passengers would be willing to have a conversation if approached. The average prediction hovered around 40 percent. The actual rate in Experiments 1a and 2a: one hundred percent. Every single stranger approached by a study participant was willing to talk.
People weren't choosing solitude because they enjoyed it more. They chose it because they assumed, incorrectly, that nobody else wanted to connect, and the belief functions as a self-reinforcing trap: because everyone assumes others would reject them, nobody initiates, and because nobody initiates, nobody ever learns the assumption is wrong. Experiments 4a and 4b tested this directly, comparing participants who had never tried connecting with a stranger against those who actually had, and the two groups delivered opposite predictions.
Experiment 5 moved to a university laboratory waiting room and tested the receiving end: participants who were talked to by a confederate reported equally positive experiences as those who had been instructed to start the conversation, which means connection was pleasant whether you initiated it or simply received it.
Seven Studies, 2,304 People, Same Answer
A single research group can be wrong, which is why the 2021 mini meta-analysis by Gillian Sandstrom and Erica Boothby, published in Self and Identity, matters: they aggregated seven studies totaling 2,304 participants and measured six specific pre-conversation fears, including worries that the conversation wouldn't be enjoyable, that the partner wouldn't like them, that they'd lack conversational skills, and the mirror-image set of concerns about whether the partner shared those same anxieties. Across every study, participants overestimated all six fears, and conversations went better than expected on every dimension measured.
The replication extended across oceans. Schroeder, Lyons, and Epley ran the original commuter paradigm with London-area train riders in a 2022 study published in the same journal as the original. Commuters who talked to strangers reported more positive experiences and learned more than those assigned to solitude. A 2025 field experiment by Li and colleagues at the University of British Columbia, with 563 adults, found that even a simple signal indicating willingness to chat increased the frequency, length, and depth of stranger interactions, with downstream benefits for social connection and happiness.
The Strongest Counterargument
The most credible critique targets the study's internal structure. Participants who drew the connection condition but failed to start a conversation were excluded from certain analyses. If the naturally reluctant gave up and dropped out, the remaining sample skews toward people comfortable talking, people who would enjoy it regardless. Epley and Schroeder addressed this by noting that compliance was high and the solitude and control groups provided clean comparisons, but the concern is legitimate. The study measured short-term mood on a single commute, not long-term well-being, and the settings were trains and buses in Chicago and London, not cultures where unsolicited conversation violates stronger social norms. Research on personality and solitude suggests that some introverts experience genuine restoration from being alone; a twenty-minute commute experiment may not capture the full picture for people whose social batteries drain differently.
What We Didn't Prove
The participants in the original study were overwhelmingly American commuters in a culture that nominally values friendliness. Whether the finding holds in Tokyo, Seoul, or Helsinki, where public transit silence carries different cultural weight, remains untested. The study measured self-reported mood, not physiological markers like cortisol or heart rate variability. No study has tracked whether repeated daily stranger connections accumulate benefits over weeks or months, or whether the novelty wears off. The 2021 meta-analysis pooled results primarily from Western, English-speaking samples. And the practical concern remains: some people on the train genuinely do not want to be spoken to, and no study measured the well-being cost borne by unwilling conversation targets who felt too polite to refuse.
The Bottom Line
The most replicated finding in the science of everyday social interaction is that people consistently underestimate how pleasant it will be to talk with a stranger. Nine experiments, a seven-study meta-analysis, and independent replications on two continents converge on the same conclusion: humans in close proximity choose silence not because it makes them happy, but because they wrongly believe the alternative would make them less so. If 7.9 million Americans ride public transit on a given workday and the average one-way commute is 27.6 minutes, that represents roughly 7.3 million person-hours per day spent in avoidable social disconnection. Over 250 workdays, that is 1.8 billion hours of commuting per year that could be made more pleasant with a single sentence.
What You Can Do
Say hello to one person on your next commute — not a monologue or a life story, just a single opening sentence. The research consistently shows the interaction will be more pleasant than you predict, for both of you. If you'd rather signal availability than initiate, a simple smile or brief eye contact substantially increases the likelihood that someone else will start the conversation. And when the conversation does happen, don't cut it short because you assume your seatmate wants it to end: Schroeder's follow-up work found that people consistently underestimate how long the other person wants to keep going, which means the biggest risk you face in talking to a stranger isn't rejection, but quitting too early.