Bystander Effect: What Research Shows About Why People Fail to Help

The bystander effect, the tendency for individuals to be less likely to offer help in emergency situations when other people are present, is one of the most cited and most practically significant findings in social psychology. First systematically studied by Bibb Latane and John Darley following the 1964 murder of Kitty Genovese in New York City, the phenomenon has been investigated in hundreds of studies and has influenced thinking about emergency response, workplace misconduct, bystander intervention training, and civic responsibility. Understanding what research actually shows, including both the robust findings and the important limitations, provides a more nuanced picture than popular accounts typically offer. The Kitty Genovese murder was widely reported at the time as a case in which 38 witnesses watched the attack without calling police, a framing that became foundational to Latane and Darley's research program and entered public consciousness as a symbol of urban indifference. Subsequent historical investigation has found that the original reporting significantly overstated the number of witnesses who saw the full attack and that some witnesses did call for help, though these revelations do not undermine the subsequent experimental research that documented the bystander effect using controlled laboratory conditions. Latane and Darley's experiments systematically varied the number of apparent bystanders to staged emergencies and found consistent evidence that the presence of others reduced the probability that any individual would intervene and increased the time before intervention occurred. These effects were replicated across many variations including smoke filling a room, a person falling from a ladder, a robbery, and a medical emergency, establishing the generality of the finding across emergency types. Two psychological mechanisms were proposed and supported by research. Diffusion of responsibility refers to the reduced sense of personal obligation to help when others are present, as each individual assumes that someone else will or already has responded. Pluralistic ignorance refers to the tendency to look to others for cues about how to interpret ambiguous situations, and when all bystanders look calm, to conclude that the situation is not an emergency even when it is. Research manipulating these factors, for example by removing the ability to see others' responses or by explicitly assigning responsibility to one bystander, confirms that both mechanisms contribute to the bystander effect. The bystander effect has important limits and boundary conditions that subsequent research has established. Research finds that the effect is weaker or absent in situations involving clear emergencies with unambiguous need for help, in situations where the victim and bystander know each other, in smaller and more cohesive groups, and in situations where the costs of intervention are low. Research on real emergency situations, including analysis of data from bystander intervention in actual emergencies, finds that bystanders intervene more often than laboratory studies might predict, particularly in dangerous situations where the need for help is clear and unambiguous. Online bystander behavior is an extension of the traditional bystander effect that has generated research attention as social media and digital communication have created new contexts for witnessing and responding to harmful situations. Research on online bystander behavior finds patterns consistent with the traditional bystander effect: the presence of many viewers or commenters is associated with less individual intervention in situations of online harassment or bullying. However, online environments also create new possibilities for intervention including reporting harmful content, supporting victims through messages, and counter-speaking against harmful behavior. Bystander intervention training programs have been developed for contexts including campus sexual violence prevention, workplace harassment, and emergency response. Research on these programs finds that they increase bystander intervention intentions and, in some studies, actual intervention behavior. The Green Dot program for sexual violence prevention and Bringing in the Bystander have been evaluated in quasi-experimental studies and found to increase bystander intervention. These programs focus on building the knowledge, motivation, and skills to overcome the psychological barriers to intervention that the bystander effect produces.