Bystander effect
The more witnesses, the less likely anyone helps.
What it means
The bystander effect is the finding that the presence of other people reduces the probability that any single individual will intervene to help in an emergency. Two mechanisms combine to produce it: diffusion of responsibility, in which the felt duty to act is divided among everyone present so no one feels personally obligated, and pluralistic ignorance, in which each onlooker, seeing others remain calm and inactive, misreads the situation as a non-emergency. Latané and Darley's experiments demonstrated that help becomes slower and less likely as group size grows, even when the need is genuine. There are important boundary conditions: the effect weakens or reverses when the situation is unambiguously dangerous, when bystanders can identify with the victim or each other, and when someone is singled out by name, which collapses the diffusion of responsibility. It matters for emergency response, safety design, and bystander-intervention training, where the practical remedy is to make responsibility specific — addressing one identifiable person with a clear instruction rather than appealing to the crowd.
Where the story came from
The effect is inseparable from a story that later fell apart. In 1964 the New York Times reported that 38 witnesses watched Kitty Genovese's murder in Queens and did nothing, a claim that shocked the public and spurred Latane and Darley's research. Archival work by Manning, Levine, and Collins showed the figure was essentially invented: there is no evidence 38 people saw the attack, several never grasped what was happening, and at least one shouted at the assailant and drove him off temporarily. The parable was wrong, yet the laboratory phenomenon it inspired proved real. It is worth separating the durable finding from the sensational anecdote that still opens most textbooks, because the two are often confused for one another.
What the evidence shows
The seizure study is only the anchor. A meta-analysis by Fischer and colleagues, pooling 105 effects and more than 7,700 participants, put the overall effect at roughly g = -0.35, a moderate drop in helping as the group grows. Crucially, the size depends on the situation: it is largest when the emergency is ambiguous and low-cost, and it shrinks or reverses when danger is obvious, a perpetrator is present, or bystanders can physically back each other up. The tendency also appears early. Five-year-olds help a struggling adult less when other children are available, and a barrier condition showed this reflects diffused responsibility rather than mere shyness. So the effect is genuine but conditional, not a blanket law of crowds.
Where it breaks down
The strongest challenge comes from field data. Reviewing real street conflicts caught on public CCTV in the United Kingdom, the Netherlands, and South Africa, Philpot and colleagues found that in nine of ten incidents at least one bystander, and usually several, stepped in, and that a larger crowd made intervention by someone more likely rather than less. This inverts the classroom intuition: for visible, unfolding aggression, more witnesses raises the chance help arrives, because it takes only one person to break the freeze. The lab paradigm captures a real hesitation, but it over-represents ambiguous, private, low-stakes situations. The authors argue the field should stop asking why nobody helps and start studying what makes the intervention that usually happens succeed.
Turning a crowd into a helper
Latane and Darley framed helping as a five-step decision: a bystander must notice the event, interpret it as an emergency, accept personal responsibility, know how to help, and then act. A crowd can stall any step, as others' visible calm suppresses interpretation, their mere presence dilutes responsibility, and uncertainty about the right move delays action. Modern bystander-intervention programs, now common on campuses and in workplaces, target each juncture: they teach people to name ambiguous situations out loud, to expect the freeze, and to rehearse low-cost scripts so acting needs no improvisation. Assigning a task to one named person works because it collapses the responsibility step for everyone at once, but the deeper aim of training is to make each person less dependent on the crowd's cue to begin with.
Examples
A person collapses on a crowded street and everyone assumes someone else has already called for help.
A safety concern posted to a channel of forty engineers sits unanswered for a week; sent to one named engineer with 'can you look at this by Friday', it is fixed that day.
Nobody replies to the 'can someone take this?' email to the whole team, and each person's silence quietly reassures the others that it cannot really be urgent.
A distressing post reaches a group of five thousand members; every reader assumes that with so many watching, someone must have already reported it, and no one ever does.
A ward monitor alarms for a patient no nurse is specifically assigned to; each of several staff within earshot assumes a colleague nearer the bed will answer it first.
First described in Latané & Darley (1968).
Key references
- Philpot, R., Liebst, L. S., Levine, M., Bernasco, W., & Lindegaard, M. R. (2020). Would I be helped? Cross-national CCTV footage shows that intervention is the norm in public conflicts. American Psychologist, 75(1), 66-75. doi.org/10.1037/amp0000469
- Plotner, M., Over, H., Carpenter, M., & Tomasello, M. (2015). Young children show the bystander effect in helping situations. Psychological Science, 26(4), 499-506. doi.org/10.1177/0956797615569579
- Fischer, P., Krueger, J. I., Greitemeyer, T., Vogrincic, C., Kastenmuller, A., Frey, D., Heene, M., Wicher, M., & Kainbacher, M. (2011). The bystander-effect: A meta-analytic review on bystander intervention in dangerous and non-dangerous emergencies. Psychological Bulletin, 137(4), 517-537. doi.org/10.1037/a0023304
- Manning, R., Levine, M., & Collins, A. (2007). The Kitty Genovese murder and the social psychology of helping: The parable of the 38 witnesses. American Psychologist, 62(6), 555-562. doi.org/10.1037/0003-066X.62.6.555
- Darley, J. M., & Latane, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 8(4), 377-383. doi.org/10.1037/h0025589