Bystander effect
Claim assessed: The more bystanders there are, the less likely each of them is to help.
Published · how we work
In short
The bystander effect means that a person is less willing to help in an emergency when others are present than when they are alone. The classic explanation is diffusion of responsibility: each person feels that helping is someone else’s job. The effect is well documented, but it is often misread as if nobody in a crowd helps. Footage of real conflicts shows the opposite: someone almost always steps in.
What it says
The classic experiments. Darley and Latané (1968) started from the 1964 murder of Kitty Genovese, about which the press had written that 38 witnesses watched without helping. In their experiments and those that followed, which simulated various emergencies, the presence of other witnesses made each person step in less often. They explained this by diffusion of responsibility. We haven’t read the study; we describe it from Philpot et al. (2020) and Liebst et al. (2019).
The meta-analysis. Fischer et al. (2011) pooled the studies from the 1960s to 2010: over 7,700 participants and 105 measured effects. The average effect was small to moderate (g = −0.35). It was weaker when the situation seemed dangerous, when the perpetrator was present and when helping carried a physical cost. The authors think a dangerous emergency is recognised as an emergency faster. We have read only the abstract of this work.
Real assaults, bystander by bystander. Liebst et al. (2019) analysed 81 assaults in central Copenhagen, reported to the police and filmed by surveillance cameras, and observed 764 bystanders. They expected the effect to reverse, since in dangerous situations more people means more support. Instead they found the classic effect: the more bystanders there were, the less often each one stepped in. Far more important was whether the bystander knew someone involved in the conflict: their odds of stepping in were over 20 times those of a stranger. Even so, at least one bystander stepped in in 85% of the assaults.
What matters for the victim. Philpot et al. (2020) asked the question a victim asks: will anyone help? They analysed 219 real conflicts, from heated arguments to serious violence, filmed in the city centres of Amsterdam, Cape Town and Lancaster. They counted as intervention calming gestures, soothing touches, blocking or pulling the aggressor away, consoling the victim and practical help for an injured person. At least one bystander stepped in in 90.9% of the conflicts, usually several, 3.76 on average. The share didn’t clearly differ between the three countries, although central Cape Town is seen as much less safe. With each extra bystander, the odds that the victim received help rose by about 10%.
Example
Someone collapses on a crowded platform. Each person around may think someone else has already called the emergency number. On the other hand, with so many people present, it is likely that at least one of them will come over. The example is an illustration, not a study cited here.
How to apply it
The steps below are a practical way to apply it that we propose, based on the studies cited.
- Don’t assume someone else has already handled it. That is exactly the thought diffusion of responsibility describes.
- When you need help, ask a specific person. This is a deduction from the diffusion-of-responsibility explanation; the studies cited here didn’t test it.
- You don’t have to step in alone. In the conflicts filmed by Philpot et al., several people usually stepped in, and the meta-analysis suggests the other bystanders can also be a support.
- Don’t fear that the crowd will ignore you. In the real conflicts studied, someone almost always helped.
Limits and nuances
- The effect is about each bystander, not the group. A bystander helps less often when there are many, but with more people, the chance that at least one helps can rise. The two are often confused.
- The footage studies are correlational. They don’t show that the number of bystanders is the cause; situations with many and few bystanders may differ in other ways.
- It may be saturation, not just apathy. Liebst et al. suggest that beyond a certain number of people already helping, the others are no longer needed. The effect of each extra bystander shrank as the number grew: going from two to three bystanders mattered more than going from 12 to 13.
- What counted as help. Philpot et al. also counted small acts, such as a calming gesture. The footage had no sound, so purely verbal interventions couldn’t be counted.
- The settings are limited. The footage comes from city centres, often from areas with bars, and the Copenhagen assaults were reported to the police, so more serious. The results don’t automatically apply to other situations.
- The Kitty Genovese case. Philpot et al. note that the account of the 38 passive witnesses is questioned by more recent work, which we haven’t read.
- What we read. Philpot et al. (2020), in the accepted manuscript on OSF, and Liebst et al. (2019), in the published version from the University of Copenhagen repository, we read in full. Of Fischer et al. (2011) we read only the abstract. We haven’t read Darley and Latané (1968); we describe it from the other studies.
Sources
- John M. Darley, Bibb Latané (1968). Bystander intervention in emergencies: Diffusion of responsibility
- Peter Fischer, Joachim I. Krueger, Tobias Greitemeyer et al. (2011). The bystander-effect: A meta-analytic review on bystander intervention in dangerous and non-dangerous emergencies
- Lasse Suonperä Liebst, Richard Philpot, Wim Bernasco et al. (2019). Social relations and presence of others predict bystander intervention: Evidence from violent incidents captured on CCTV
- Richard Philpot, Lasse Suonperä Liebst, Mark Levine, Wim Bernasco, Marie Rosenkrantz Lindegaard (2020). Would I be helped? Cross-national CCTV footage shows that intervention is the norm in public conflicts
See also: Asch conformity experiments, Availability heuristic, Fundamental attribution error, Hawthorne effect, Milgram experiment