The crowd effect
Lesson 18 of 31 in the course “Brain Hacks”. Free, with animations and a quiz.
Why no one steps up in a crowd
Someone feels unwell on a busy street. Dozens of people walk past, a few glance over, but no one stops. It looks like everyone is heartless. In fact it’s the situation at work, not character. Psychologists call it the bystander effect. In a crowd, two autopilot thoughts switch on in everyone. One: “Someone will help.” Two: “Everyone’s calm, so it’s probably nothing.” Both sound reasonable, and both can leave a person without help. In this lesson: what the experiments showed, which famous story turned out to be a myth, what real camera footage shows, and how to make sure help comes to you.
The “seizure” experiment
In the 1960s psychologists ran this experiment. A student sat alone in a booth and joined a discussion over headphones. One speaker mentioned he was prone to seizures. Soon he began choking, calling for help, and then the line went silent. It was a recording, but the student didn’t know. The key part: some believed they were connected only to this person, others thought one or four more people were listening. Alone, 85% ran to get the researcher. With one other listener, 62%. With four, only 31%. Gender and personality type didn’t predict who helped. One thing did: how many others the student thought could hear it.
The smoke nobody reported
The same researchers went further. Students were filling in a survey when smoke began seeping into the room from a vent. Those sitting alone quickly got up and reported it 75% of the time. Those sitting with two actors told not to react reported it only 10% of the time. Nine out of ten kept writing, waving the smoke away and coughing. Afterwards they explained: “Probably steam” or “an air-conditioning leak.” That’s the second crowd mechanism. When it’s unclear what’s happening, we look at others. Everyone’s calm, so it’s nothing. But the others are looking at you too, and they see your calm.
Two mechanisms in one picture
Let’s put both mechanisms together. The first is diffused responsibility. When you’re alone, it’s all on you, and the brain finds no excuse. When others are around, it’s split: “Surely someone’s already calling.” The second is uncertainty. Most street situations are unclear: is that person lying down because they’re ill or just tired? We look at the faces around us, and everyone tries to look calm so as not to seem silly. Each person reads the others’ calm as a sign that all is fine. Important: this isn’t about bad people. Participants weren’t indifferent. Many were nervous and worried; the situation just nudged them to wait.
The myth of the 38 witnesses
These experiments started after a famous story. In 1964 a young woman was attacked at night in New York. A major newspaper wrote that 38 neighbours saw the attack and nobody did anything. The story went into every textbook. Later, journalists and historians rechecked the facts. Almost no one had seen the whole attack, many only heard screams and didn’t understand what was happening, and some did call the police. In 2016 the same newspaper admitted it had grossly exaggerated the number of witnesses and what they saw. The takeaway: the crowd effect itself is backed by experiments, but neat stories about it are worth checking. Even in textbooks.
Good news from the cameras
Lab experiments are one thing, the street is another. In 2019 psychologists studied street-camera footage: around two hundred real conflicts in the UK, the Netherlands and South Africa. In 91% of cases at least one bystander stepped in: separating people, calming them, leading someone away. And the more people around, the higher the chance someone helped. How does that fit the experiments? Each individual in a crowd really is more likely to wait. But with many people, there’s almost always someone who steps up first. So don’t assume people are bad. Helping is the norm. Sometimes it just needs a spark, and that spark can be you.
Hack: single out one person
If you or someone near you needs help, don’t shout “Somebody help!” That spreads responsibility across everyone again. Instead, pick one person, look them in the eye, point at them and be specific: “You, in the yellow jacket, call an ambulance!” That one line fixes both crowd problems at once. Uncertainty is gone: it’s clearly serious. Responsibility isn’t diffused: it’s clear who helps and what exactly to do. It works for small stuff too. “Who’ll do the slides?” in a group chat often gets silence, while “Lyosha, can you do the slides by Friday?” usually gets a “Yes.” The more specific the ask and the person, the better the odds.
Danya and the 30-person chat
Late one evening, in the class group chat, Masha writes: “I feel really bad.” Sad emojis and hearts appear under the message, but nobody replies. Everyone assumes someone has already messaged her privately. It’s the same crowd effect, just online: the bigger the chat, the easier it is to decide someone else will help. Danya catches himself thinking “surely someone already asked” and remembers this lesson. He messages Masha directly, then calls. Turns out she has a high fever and is home alone. Danya helps her get a doctor. Try it: next time you catch the thought “someone’s already done it”, check. Often that “someone” is you.
Check yourself
- In the “seizure” study, 85% helped when alone but 31% with four other listeners. The main reason?
Correct answer: Responsibility diffused: each waited for someone else. Yes. The more witnesses someone thinks there are, the less personal responsibility they feel. - Smoke fills the room, two calm people sit nearby. Only 10% reported it. Why?
Correct answer: Seeing calm neighbours, they decided it was nothing. Yes. In an unclear situation we look at others and read their calm as “everything’s fine”. - You feel unwell in a crowded park. What works best?
Correct answer: Address one person: “You in the yellow jacket, call an ambulance!”. Yes. A specific person and a specific request remove both the uncertainty and the diffused responsibility. - What did street-camera footage of real conflicts show?
Correct answer: In about 9 out of 10 cases someone stepped in. Yes. 91% of conflicts, and the more people around, the higher the chance someone helps. Helping is the norm.