Violence in Numbers

If you’ve attended a music festival, you probably know the thrill of being part of a crowd, singing the lyrics of a popular song together with thousands of fellow fans. As social beings, our desire to belong to a group is one of our fundamental needs. However, being in a group alters how we think and act. The unity of a group doesn’t just evoke pleasant feelings — it changes how individuals process responsibility, control, and consequences, sometimes leading to outcomes like violence and aggression. Psychology research has shown that social situations reduce an individual’s sense of agency or the feeling that one is in control of one’s actions. This diminished sense of control can, in turn, make individuals feel less accountable for their actions and lead to violent behavior.

Striking examples of aggressive group behavior can be observed during political rallies, such as the January 6 attack on the U.S. Capitol, or big sports events. For instance, a group of researchers from the Ifo Institute for Economic Research showed that violent crimes increase on average by 17% on football match days in Germany. The collective energy can push fans toward aggression, leading to clashes with rivals or police. 

While such mob attacks may involve hundreds or thousands of people, these patterns of aggressive behavior can be observed at much lower scales within a group of just a few people. Our research team has recently conducted a meta-analysis based on more than 100 published studies. We found that violent offenses, or offenses against individuals, are less likely to involve collaboration compared to property crimes like burglary, with around three-quarters of violent offenses committed by a single individual. However, when violent crimes are committed by a group, these groups tend to be larger than those involved in other types of crimes. Rather than cooperation being necessary for the crime commitment itself, we suggest that violent crimes emerge as a consequence of social dynamics.

The idea that social dynamics shape criminal behavior has been extensively discussed by criminologists, psychologists, and sociologists over the last decades. One widely accepted explanation is that being a part of a group reduces the responsibility of individuals. For instance, analyzing the survey data from around 1,200 criminal male adolescents, Zachary Rowan showed that not only did they report a lower sense of responsibility when they committed a crime in a group, but this decrease was also associated with the number of co-offenders. The more people were involved in crime, the less responsible they felt.

This diffusion of responsibility doesn’t just alter perception, it also influences behavior. Studies in social psychology show that when people believe they won’t be held solely responsible, they are more likely to take risks, act impulsively, or engage in morally questionable actions. In the context of group violence, this means that individuals may escalate aggression beyond what they would do alone, knowing that blame will be shared or diluted among the group.

Mathematically, it makes sense that as the number of participants in a group increases, the fraction of responsibility attributed to each individual diminishes. For example, in a team project, the workload and credit for success are typically shared among all members. Similarly, in a group crime, individuals may feel less personally accountable as the “share” of responsibility is distributed across the group. However, one limitation of offender surveys is that they rely on self-reporting after the crime has already occurred. Offenders might claim they felt less responsible as a way to minimize their perceived guilt. So how do we know that diffused responsibility is a genuine psychological phenomenon and not just a convenient excuse?

Luckily, neuroscience has an answer for this, and the answer is that social situations indeed influence an individual’s sense of agency. Frederike Beyer, a psychology researcher based in London, conducted two experiments using modern neuroimaging techniques, keeping this question in mind. Both experiments explored how social contexts influence individuals’ sense of control over outcomes in risky decision-making tasks. Participants faced scenarios where they had to act to minimize losses while managing uncertainty and unpredictable changes. In social conditions, responsibility for the outcome could be shared with another player, creating a dynamic of distributed accountability.

Results of the first experiment showed reduced brain electrical activity in trials when they had made a successful action, but with the presence of a co-player. Specifically, this reduced neuronal response was observed in the mid-central part of the skull around 250 milliseconds after the trial outcome was presented. This response, known as feedback-related negativity (FRN), is commonly associated with monitoring the consequences of one’s actions. The reduction in FRN suggests that individuals perceive an outcome as less within their control when acting in a shared context.

In the second experiment, Beyer and colleagues found increased neuronal activity in the precuneus and temporo-parietal junction (TPJ) brain regions while participants were taking action in trials with a co-player. Both of these regions have been linked to the processes related to self-consciousness. For instance, damage to the TPJ has been shown to cause out-of-body experiences, where individuals feel as though their sense of self is detached from their physical body.

Taken together, these findings suggest that our brains process actions differently in social contexts compared to when we act alone. This distinct mental processing occurs not only after outcomes are evaluated but also during the action itself. While these experiments do not directly address whether the reduced sense of agency scales with the number of co-players, they provide a fundamental framework for exploring how social dynamics influence our sense of control.

Although these findings paint a rather grim picture of group dynamics, it is important to note that groups can also inspire positive outcomes. Collective actions have led to monumental positive change, such as the March for Science and Fridays for Future movements. In such cases, the unity and shared purpose of a group amplify compassion, resilience, and innovation. Understanding both the positive and negative aspects of group behavior is key to fostering environments that encourage collaboration while mitigating the risks of violence.

References

1. Andres, L., Fabel, M., & Rainer, H. (2023). How much violence does football hooliganism cause? Journal of Public Economics, 225, 104970. https://doi.org/10.1016/j.jpubeco.2023.104970

2. Beyer, F., Sidarus, N., Bonicalzi, S., & Haggard, P. (2016). Beyond self-serving bias: diffusion of responsibility reduces sense of agency and outcome monitoring. Social Cognitive and Affective Neuroscience, 12(1), 138–145. https://doi.org/10.1093/scan/nsw160

3. Beyer, F., Sidarus, N., Fleming, S., & Haggard, P. (2018). Losing control in social situations: How the presence of others affects neural processes related to sense of agency. eNeuro, 5(1), https://doi.org/10.1523/eneuro.0336-17.2018

4. Rowan, Z. R., Kan, E., Frick, P. J., & Cauffman, E. (2021). Not (Entirely) guilty: The role of co-offenders in diffusing responsibility for crime. Journal of Research in Crime and Delinquency, 59(4), 415–448. https://doi.org/10.1177/00224278211046256

5. Klymentiev, R., Harvey, D., Rocha, L.E.C. & Vandeviver C. (2025). A systematic review and Bayesian meta-analysis of co-offending characteristics. Nature Human Behaviour. https://doi.org/10.1038/s41562-025-02244-z

6. Paloutzian, R. F. (1975). Effects of deindividuation, removal of responsibility, and coaction on impulsive and cyclical aggression. The Journal of Psychology, 90(2), 163–169. https://doi.org/10.1080/00223980.1975.9915771

7. Vives, M., Cikara, M., & FeldmanHall, O. (2021). Following your group or your morals? The In-Group promotes immoral behavior while the Out-Group buffers against it. Social Psychological and Personality Science, 13(1), 139–149. https://doi.org/10.1177/19485506211001217

8. Wallach, M. A., Kogan, N., & Bem, D. J. (1963). Diffusion of Responsibility and Level of Risk Taking in Groups. ETS Research Bulletin Series, 1963(1). https://doi.org/10.1002/j.2333-8504.1963.tb00956.x

9. Wikipedia contributors. (2025, April 15). Fridays for future. Wikipedia. https://en.wikipedia.org/wiki/Fridays_for_Future

10. Wikipedia contributors. (2025, April 9). March for Science. Wikipedia. https://en.wikipedia.org/wiki/March_for_Science

11. Wikipedia contributors. (2025, April 15). January 6 United States Capitol attack. Wikipedia. https://en.wikipedia.org/wiki/January_6_United_States_Capitol_attack