Íse Murphy

When crowds ‘crush'

As the world population increases, so does the frequency of mass gatherings and mass events (Li et al, 2020). Instigated by a career in major events, I am fascinated by crowds, how they move, how they behave. When I began studying crowd safety, it became clear that most crowd disasters could have been avoided with simple strategies. However, in order to prevent, we must understand why they happen in the first place.

Over the last century in the UK, changes in legislation have supported the reduction of football related crowd disasters. Modernisation of communications and transport have allowed the sharing of training, skills and experience between countries across the world. However, disasters are still occuring, and since the pandemic, the event safety industry has faced significant issues such as loss of human resource, loss of experience and a shift in crowd behaviour, which all contribute to dangerous occurences. When disaster occurs, it is always due more than one reason, which I have previously discussed. However, this article focuses on the proximate cause, i.e. the trigger point of a crowd 'crush' or crowd turbulence (Helbing et al, 2007).

There are two main areas of academic study on crowds - physical (crowd dynamics) and psychological (crowd psychology). The reason a crowd come together can be fuelled by psychology (sport, politics, celebrations) but once this crowd form, dynamics swirls into the mix. Effective crowd management balances both crowd psychology and crowd dynamics.

Foule; How we perceive crowds

Over the last century, our perception of crowds and how they behave have evolved. The classical view, developed by notable names from France and Italy including Gustave Le Bon, Gabriel Tarde and Scopio Sighele perceived crowds as as collective mind, a single soul, void of reason and intellectually inferior to the individual (Le Bon, 1896). Tarde (1968) viewed crowds as 'hypnotic states' and a germ of society. By the end of the 19th century, crowds were viewed in a negative light. The French for crowd is foule, (one letter away from the English 'foul') which offers further insight to how crowds were perceived in the past. In fact, the study of crowds was in order to subvert and control them (Nye, 1975). However, modern researchers, spearheaded by Elias Canetti (Crowds and Power, 1962), have realised that crowds are far more rational, discriminate, not as violent as previously assumed and don't panic (Drury et al, 2009; Drury and Reicher, 2009; Templeton et al, 2015). Today, studies focus on shared social identity in crowds; where collective behaviour occurs when individuals shift from personal identity to identify as part of a group, resulting in reduction in stress, increase in well-being (Hopkins and Reicher, 2017), and mutual aid and cooperation in times of crisis (Drury et al, 2009; Yuan and Xiaoping 2018; Drury et al, 2019).

Trigger Density

So why, if crowds are not as irrational as previously assumed, does a crowd 'crush' happen? There are many factors to consider; time, density, flow, surface area, time of day, weather, communication, management, type of event, type of environment, etc. but it all narrows down to density. The proximate cause of a 'crush' is insufficient information and space (Sime, 1999).

Density itself is not a risk without a trigger. When density increases in a crowd, flow decreases, which can shift the dynamic of the crowd from moving to 'jamming' (Zheng et al, 2010). An increase or decrease in crowd density can potentially be the difference between life or death. At densities of 2 people per metre squared (ppm²), people can move relatively freely and sway their arms as they walk together in a crowd. At a density of 3-4ppm², crowd flow decreases as people have less room to walk. This density is typical of a queue or standing audience.

The trigger point hits when density increases to 5ppm² or more (Smith, 1995; Lee and Hughes, 2005; Zheng et al, 2010; Feliciani and Nishinari, 2018). What happens at this point is a phenomenon. There is a sudden shift in crowd dynamics from being a group of individuals who have total control over their own bodies, to a fluid mass that behaves as one (Fruin, 1993; Helbing and Mukerji, 2012).

It is important to note that this density itself, without a trigger, is not the cause of disaster, but is a risk that can be identified, measured, and managed to reduce the potential of harm. Feliciani and Nishinari (2018) note there have been controlled studies whereby this density has been reached, and exceeded, safely, with researchers identifying that it is when force is applied to this density, that safety risk increases significantly. When density and force are combined, the resulting crowd pressure is what causes injury and death (Helbing at al, 2007). In addition, the physical interaction between people transfers force between bodies which builds, resulting in crowd turbulence; a wave like movement where the direction and strength of force is sudden and hard to predict, resulting in excessive pressure on the lungs and loss of balance (Helbing and Mukerji, 2012).

Crowd Pressure on the body

When crowd density increases and we are no longer in control of our bodies, we are susceptible to this crowd turbulence. If crowd pressure continues to build on human bodies, injury occurs and risk of death is high unless that pressure is released. In scenarios where there is excessive force on the lungs, each time a person exhales, the continued force restricts the next inhalation. Slow death is caused by compressive asphyxiation. In scenarios where people lose their footing and fall over, they could be trampled by other people who would have no control over their own bodies either.

The tolerance level of pressure on a human chest is between 180 - 247 N/m (Smith and Lim, 1995). When a crowd 'crush' occurs, forces could reach up to 4,450 Newtons per metre (N/m) (Helbing et al, 2000), which can bend steel railings (Fruin, 1993). Individuals are unable to resist pressures of 162 N/m (young female) to 242 N/m (middle aged male) and human ribs will break under the pressure of 3,000 N/m (Li et al, 2020). This highlights just how much force is applied within that crowd.

In this scenario, there is almost no possibility for escape as due to this intense pressure and dynamic force, individuals have little control over their bodies to be able to find a means of escape. In addition, those at the back of the crowd who are not subject to the crowd turbulence simply might not know there is an issue only metres away from them, and are therefore unaware of the pressure in the centre of the crowd (Helbing and Mukerij, 2012). If the crowds in this example are not being monitored by a safety team, there little opportunity to manage the density before it becomes a risk, and no ability to direct or communicate with the crowd in order to resolve issues.

Preventing Disaster

Reducing the proximate risk is one approach, but does not solve the root cause of disaster. The work that I and many others do focuses on ways of preventing disaster, long before the event even takes place. There are many approaches that can be taken including but limited to; learning lessons of previous disasters, introduction and amendment of safety laws, the study of crowd psychology and crowd dynamics, the development of qualifications and training to improve the competency of crowd safety professionals, and the utilisation of high-tech simulation and crowd monitoring software available today.

In addition, simple strategies can also be adopted to reduce risk and avoid needless injury or death. Crowd Safety Risk Analysis includes the applications of risk assessment models such as Still's (2014) DIM-ICE (Design, Information, Management - Ingress, Circulation, Egress) and RAMP (Route Area Movement Profile) Analysis models which can quickly identify significant safety risks to a plan, location or capacity. If first pass analysis such as these were applied to some recent crowd disasters, it would have highlighted that the locations and capacities were not compatible.

Crowd Safety Management is a critical element in any event to ensure the safety of attendees. As discussed, crowds are dynamic, and so the risk assessment must also be dynamic. It is essential for crowd safety to be threaded into the four stages of an event life cycle; planning, licensing, delivery, and review, and on the day of event, crowds need to be managed by sufficient numbers of qualified and competent safety teams, working symbiotically with stakeholders and emergency services.

These strategies all contribute to safe management practices, however, when studying the root cause of disasters, the theme that consistently arises is safety culture; whereby the attitude (or lack thereof) towards safety by those in charge of an organisation is the underlying cause of almost all disasters (Keane, 1982 Stardust Fire; Couto, 1989 Aberfan Disaster; Paté-Cornell, 1992 Piper Alpha Disaster; Pidgeon, 1997 Chernobyl Disaster; Lea et al, 1998; Challenger and Clegg, 2011; Almeida and Schreeb, 2019).

If we truly want to prevent another disaster from occuring, we need to start there.

If this article was helpful and you wish to learn more about improving ingress, egress or zone ex at your event or venue, please send me an email


Almeida, M. M. de and Schreeb, J. von (2019) “Human Stampedes: An Updated Review of Current Literature.” Prehospital and Disaster Medicine, 34(1) pp. 82–88.

Canetti, E., 1962. Crowds and Power, translated by Carol Stewart. New York, NY.

Challenger, R. and Clegg, C. W. (2011) ‘Crowd disasters: a socio-technical systems perspective’. Contemporary Social Science. 2nd ed., 6(3) pp. 343–360.

Couto, R. A. (1989) “Economics, Experts, and Risk: Lessons from the Catastrophe at Aberfan.” Political Psychology. (International Society of Political Psychology), 10(2) pp. 309–324.

Drury, J., Cocking, C. and Reicher, S. (2009) “The psychology of crowd behaviour in emergency evacuations: Results from two interview studies and implications for the Fire and Rescue Services.” The Irish Journal of Psychology, 30(3–4) pp. 59–73.

Drury, J. and Reicher, S. (2009) “Collective Psychological Empowerment as a Model of Social Change: Researching Crowds and Power.” Journal of Social Issues, 65(4) pp. 707–725.

Drury, J., Carter, H., Cocking, C., Ntontis, E., Guven, S. T. and Amlôt, R. (2019) “Facilitating Collective Psychosocial Resilience in the Public in Emergencies: Twelve Recommendations Based on the Social Identity Approach.” Frontiers in Public Health, 7 p. 141.

Feliciani, C. and Nishinari, K. (2018) “Measurement of congestion and intrinsic risk in pedestrian crowds.” Transportation Research Part C, 91, June, pp. 124–155.

Fruin, J. J. (1993) “The Causes and Prevention of Crowd Disasters.” Engineering for Crowd Safety. (Elsevier), August, pp. 99–108.

Helbing, D., Farkas, I., Vicsek, T., 2000. Simulating dynamical features of escape panic. Nature 407 (6803), 487–490.

Helbing, D., Johansson, A. and Al-Abideen, H. Z. (2007) “The Dynamics of Crowd Disasters: An Empirical Study.” arXiv.

Helbing, D. and Mukerji, P. (2012) “Crowd disasters as systemic failures: analysis of the Love Parade disaster.” EPJ Data Science, 1(1) pp. 1–40.

Hopkins, N. and Reicher, S. D. (2017) “Social identity and health at mass gatherings.” European Journal of Social Psychology, 47(7) pp. 867–877.

Lea, W., Uttley, P. and Vasconcelos, A. C. (1998) “Mistakes, Misjudgements and Mischances: Using SSM to Understand the Hillsborough Disaster.” International Journal of Information Management, 18, September, pp. 345–357.

Le Bon, G. (1896).The crowd: A study of the popular mind. Dunwoody, GA: Norman S. Berg

Lee, R. S. and Hughes, R. L. (2005) “Exploring Trampling and Crushing in a Crowd.” Journal of Transportation Engineering, 131(8) pp. 575–582.

Li, X., Song, W., Xu, X., Zhang, J., Xia, L. and Shi, C. (2020) “Experimental study on pedestrian contact force under different degrees of crowding.” Safety Science, 127 p. 104713.

Nye, R. A. (1975).The origin of crowd psychology: Gustave Le Bon and the crisis of mass democracy in the third republic. London: Sage.

Paté-Cornell, M. E. (1992) “Learning from the Piper Alpha Accident: A Postmortem Analysis of Technical and Organizational Factors.” Risk Analysis, 13(2) pp. 215–232.

Pidgeon, N. (1997) ‘The Limits to Safety? Culture, Politics, Learning and Man-Made Disasters’. Journal of Contingencies and Crisis Management, 5, March, pp. 1–14.

Sime, J. D. (1999) “Crowd facilities, management and communications in disasters.” Facilities, 17(9/10) pp. 313–324.

Smith, R. A. and Lim, L. B. (1995) “Experiments to investigate the level of ‘comfortable’ loads for people against crush barriers.” Safety Science, 18(4) pp. 329–335.

Still, G.K. (2014) Introduction to Crowd Science, London: CRC Press.

Tarde, G.D., 1968. La philosophie pénale (Penal Philosophy). Patterson Smith, Montclair.

Templeton, A., Drury, J. and Philippides, A. (2015) “From Mindless Masses to Small Groups: Conceptualizing Collective Behavior in Crowd Modeling.” Review of General Psychology, 19(3) pp. 215–229.

Yuan, C. and Xiaoping, Z. (2018) “Emergence of cooperation during an emergency evacuation.” Applied Mathematics and Computation, 320, March, pp. 485–494.

Zheng, X., Sun, J. and Zhong, T. (2010) “Study on mechanics of crowd jam based on the cusp-catastrophe model.” Safety Science, 48(10) pp. 1236–1241.

#crowd safety #psychology