On March 13, 1964, in New York, a young woman named Kitty Genovese was attacked in the street, not far from her apartment in the Kew Gardens neighborhood of Queens. The assault lasted several dozen minutes. She died from her injuries. A few weeks later, the New York Times published a sensational article: thirty-eight neighbors had allegedly watched the scene from their windows without calling the police, intervening, or even opening their doors.

This story has endured for decades. It became the starting point for one of the most frequently cited concepts in social psychology: the bystander effect—or bystander effect. The idea is simple and disturbing: the more people who are present during an emergency, the less likely each person is to intervene.

An experiment that changed how human behavior is studied

Two American psychologists, John Darley and Bibb Latané, were so struck by the Genovese case that they decided to test it scientifically. In 1968, they published a series of experiments that became famous in psychology textbooks.

In one of them, a participant was placed alone in a room with a microphone to take part in a discussion. Over the intercom, the participant heard one of the other participants—actually a recorded voice—simulate an epileptic seizure. The result: when subjects believed they were the only witness, they intervened in 85% of cases. When they believed five other people were also hearing the same scene, the rate fell to 31%.

The experiment was repeated in different forms, with variations involving smoke seeping under a door or an actor pretending to fall in a hallway. The result was consistent: the presence of others—even when merely assumed—held back individual action.

Two mechanisms at work

Darley and Latané identified two main psychological forces that explain this phenomenon.

The first is the diffusion of responsibility. When we face an emergency alone, we know that if we do not act, no one will. The moral responsibility rests entirely on our shoulders. But as soon as other people are present, that responsibility becomes fragmented. Each person thinks, more or less consciously, “someone else will take care of it.” When every witness makes this implicit calculation, the result is that no one acts.

The second mechanism is social influence. Faced with an ambiguous situation, we watch other people’s reactions to decide how to interpret what we are seeing. If those around us remain calm and do not move, we tend to conclude that the situation is not really serious. Without realizing it, we conform to a collective norm of indifference.

The Genovese myth: when reality is more complicated

Yet we need to pause here. The story of Kitty Genovese, as told in the famous 1964 New York Times article, turned out to be largely inaccurate.

Historians and researchers—including Joseph De May Jr., who examined court records and original testimony—showed that the figure of “38 passive witnesses” was greatly exaggerated. Many neighbors had neither seen nor heard anything clearly. Some had heard screams but interpreted them as an ordinary street dispute. At least one person had called the police. And a friend of Kitty Genovese held her in her arms during her final moments.

This is not a minor anecdote. The bystander effect certainly exists as a real phenomenon—the experiments by Darley and Latané clearly demonstrate it. But it was built on a founding myth whose facts did not match reality. That is a lesson in itself: major theories sometimes grow out of powerful stories rather than verified facts. And once a story becomes embedded in culture, it can survive corrections for a very long time.

How more recent research adds nuance

A meta-analysis published in 2011, covering more than 7,700 participants across dozens of experiments, confirmed that the bystander effect is real—but less systematic than previously believed. In clear and physical emergencies—someone falls, someone is bleeding—the inhibition is much weaker than in ambiguous situations.

Researchers also analyzed CCTV footage from public spaces in the United Kingdom, the Netherlands, and South Africa. In the vast majority of recorded conflicts, bystanders had indeed tried to intervene. This finding tempers the idea of universal human passivity in the face of violence.

Why it still concerns us

The bystander effect is not limited to physical emergencies in the street. It operates in offices, families, and online groups. How many times have we seen a problematic message in a group chat and said nothing, assuming someone else would respond? How many times have we let a hurtful remark pass in a meeting, thinking that our silence was shared and therefore acceptable?

The diffusion of responsibility does not need a New York alleyway to work. It thrives wherever there is a group—which means almost everywhere in our lives.

Knowing about this mechanism is not enough to protect us from it: most people who took part in Darley and Latané’s experiments had no idea what they were demonstrating. But that knowledge can at least encourage us to resist the temptation to think “someone else will do it”—and to ask ourselves, from time to time, whether that someone is simply us.