Skip to main content

Verified by Psychology Today


What Motivates Bystanders to Intervene in an Attack?

How identities influence decisions to help.

Key points

  • Scholars and laypeople alike underestimate the influence that shared identities can play in people’s decisions to help.
  • As our identities shift, becoming more or less expansive, different people are included within the boundaries of our concern.
  • There may be additional factors at work when attacks involve a male perpetrator and a female victim.
  • Research underscores the need for bystanders to intervene more forcefully to stop violence against women and sexual assault.

This weekend, the New York Times reported yet another story about bystanders failing to intervene, in this case while witnessing a woman being sexually assaulted on a train outside Philadelphia.

In this highly disturbing case, several riders apparently observed an attack lasting about eight minutes during which a man assaulted and raped a woman sitting in their car. A police officer ran onto the train and caught the man. But people were horrified at the lack of assistance from bystanders who had done nothing to help the woman—and some of whom allegedly recorded the attack on their phones.

“I’m appalled by those who did nothing to help this woman,” Timothy Bernhardt, the superintendent of the Upper Darby Township Police Department, said on Sunday. “Anybody that was on that train has to look in the mirror and ask why they didn’t intervene or why they didn’t do something.”

He added that investigators had received reports of some passengers recording the attack on their phones but that the police had not confirmed those reports.

Every time an event like this occurs (and they occur shockingly often), people—very rightly—ask why. How could people possibly just sit there, doing nothing to help while a fellow passenger, a fellow human being, is in dire need of assistance? Although as we note in our new book, The Power of Us, people do often, eventually, intervene, there are factors that seem to make this more or less likely.

Social psychologists tend to explain these events in terms of the “bystander effect,” a long-studied phenomenon in which people are often less likely to help others in need when there are lots of other people around, compared to when they are on their own or in the company of just a few others. People look at others to help them figure out how to respond in novel situations—and if they see a bunch of people not reacting, they may assume that the problem isn’t serious or that someone else should intervene.

This is the standard explanation of bystander behavior or, more accurately, the lack thereof.

The Role of Shared Identities

But scholars and laypeople alike underestimate the influence that identities, and specifically shared identities, can play in people’s decisions to help. Earlier this year in response to a previous attack with unresponsive bystanders, we wrote the following (see also a related piece in the Los Angeles Times):

"As our identities shift, becoming more or less expansive, different people are included within the boundaries of our concern. These boundaries are not static, but depend on how we’re thinking about ourselves at particular moments.

In the early 2000s, Mark Levine and colleagues studied the dynamics of helping with English football fans. Some were asked to think about their identities as fans of a particular team — specifically, how much they loved Manchester United. Others were instead asked to think about their more general identity as fans of the game — as football fanatics.

Then, walking across the university campus to complete a second component of the study, they ran into an injured stranger. Like the Good Samaritan, did they stop to help? Or did they just walk on by?

Remarkably, whether they helped depended to a great extent on what type of shirt the distressed stranger was wearing. Sometimes he was in a Manchester United jersey, other times he sported the Liverpool colors (Manchester’s biggest rivals), and at other times he wore just an ordinary plain shirt.

Fans who had been asked to think about their love for Manchester United, the more parochial identity, were significantly more likely to help the stranger if he was wearing a Man U jersey compared to either of the other two shirts.

However, those fans who had thought about their love for football more generally behaved very differently. Having brought to mind this broader identity, they were just as as likely to help the stranger whether he was wearing a Liverpool or a Manchester United shirt. Notably, they remained quite unlikely to help a stranger in an unadorned shirt — someone with whom they still didn’t share an obvious allegiance."

When Sympathy Overrides Distress

In 2018, Ruud Hortensius and Beatrice de Gelder proposed a new model of the bystander effect focused on how the brain responds during emergencies. They proposed that when people observe another person in need of help, two separate brain systems are activated. One system creates feelings of personal distress and produces freezing or avoidance responses—reactions that are not conducive to helping. The other system, which they claim tends to operate more slowly, creates feelings of sympathy, which can drive people to intervene.

According to their model, our immediate, instinctive reaction in emergencies is to freeze in a state of indecision or to run away if we can. These quick and unthinking responses can, however, be overcome if feelings of sympathy and concern for the victim, increased perhaps by a shared identity, are stronger and override them.

From this perspective, a particularly disturbing aspect of the recent example of bystander apathy is the (as yet unconfirmed) report that some may have recorded the attack on their phones. Stalling in indecision or trying to escape what one fears is dangerous are understandable, though hardly admirable responses. But recording an attack can be understood neither as freezing nor avoidance — and suggests a genuinely callous lack of sympathy that our models of bystander behavior do not really account for.

Stopping Violence Against Women and Sexual Assault

There may be additional factors at play when attacks involve a male perpetrator and a female victim, as was the case on the train outside Philly—factors which put women at further risk.

Research in the 1970s found that people were more likely to intervene in a violent fight between a man and a woman if they believed them to be strangers rather than a couple. This might have been because people tended to perceive the female victim as more desirous of help and the attacker as less likely to stay and fight if they believed that the man was a stranger to the woman being attacked.

This pattern is troubling because, as the researchers noted, people observing violence “seem likely to perceive the protagonists as dates, lovers, or married couples rather than as strangers, acquaintances, or friends.” Of course, we hope that norms have changed since the 1970s, such that bystanders are perhaps more likely to help today when they see violence involving couples. But this research underscores the need for bystanders to intervene more forcefully to stop violence against women.

Understanding why people fail to help assault victims in no way exonerates them. Rather, these insights should be used to train people how to intervene in moments of crisis. A recent review of relevant research highlights the importance of increasing awareness and encouraging bystanders to step up when someone is the target of sexual violence:

“Sadly, there’s very little evidence-based research on strategies to prevent or address sexual harassment. The best related research examines sexual assault on college campuses and in the military. That research shows that training bystanders how to recognize, intervene, and show empathy to targets of assault not only increases awareness and improves attitudes, but also encourages bystanders to disrupt assaults before they happen, and help survivors report and seek support after the fact.”

More from Dominic Packer, Ph.D. and Jay Van Bavel Ph.D.
More from Psychology Today