Mass Gatherings

Abstract

Mass gatherings, which MeSH classifies under social behavior, are planned events—a pilgrimage, a sporting final, a festival, a rally—that assemble large numbers of people for a shared purpose. They are studied by disciplines that rarely met before the field matured: mass-gatherings medicine, concerned with the infectious, environmental, and trauma risks such events concentrate; a physics of pedestrian flow modelling how density and turbulence produce crowd disasters; and a social psychology of how shared identity governs whether a dense crowd feels dangerous or safe, and whether an emergency ends in a crush or in mutual aid. This article develops those threads around two unifying findings: that the lethal variable in a crowd is density and the flow it drives, and that the moderating variable is social identity. A worked example and three interactive demonstrations make the mechanisms manipulable.

Keywords: mass gatherings, crowd safety, mass-gatherings medicine

Every year the Hajj brings more than two million pilgrims to a few square kilometres of Mecca; a European football match seats tens of thousands behind a handful of exits; a festival crowd presses toward a single stage. These events deliver profound collective experience and, occasionally, catastrophe—the 2015 Mina crowd disaster, the 1989 Hillsborough disaster, the 2021 Astroworld crush. What makes a mass gathering dangerous is rarely the thing observers reach for first. It is not that crowds turn irrational; it is that people packed above a critical density lose the physical room to move, and that whether the people around them are experienced as a threatening mob or a community of fellow participants depends on who they take themselves to be (Memish et al., 2019).

The behavioural and health sciences approach mass gatherings through complementary questions. How should such events be defined and classified so that their risks can be planned for? What are the physical laws of a dense moving crowd, and where does flow collapse into a deadly crush? What illnesses and injuries do mass gatherings concentrate, and how are they surveilled and prevented? And why do some dense crowds feel safe and behave with striking solidarity while others panic—a question whose answer, it turns out, is not density alone but the social identity through which density is experienced. The sections below take these in turn, after first placing mass gatherings within the broader category of social behavior under which they are classified.

What Mass Gatherings Are

A mass gathering is, in the working definition the field has settled on, a planned, time-bounded event at which a number of people large enough to strain the planning and response resources of the host community assemble in a defined space for a common purpose—religious, sporting, musical, civic, or political. The World Health Organization and the mass-gatherings literature add that the threshold is relative, not absolute: a crowd of a few thousand may be a mass gathering for a small town and routine for a capital city, because what defines the category is the ratio of attendance to available resources, not a fixed headcount (Memish et al., 2019). The Medical Subject Headings vocabulary treats Mass Gatherings as a descriptor filed under both Social Behavior and crowding, reflecting its dual character as a social event and a problem of physical density.

What makes mass gatherings a distinct object of study, rather than merely a large instance of ordinary social life, is that they concentrate risks that are negligible at smaller scale. Pathogens that would infect a handful of people instead reach thousands and then disperse, with returning attendees, across the globe; a few degrees of summer heat become mass heat-stroke; a bottleneck that would inconvenience a small queue becomes, in a crowd of a hundred thousand, a fatal crush. The study of mass gatherings is therefore the study of what changes when human assembly is scaled up—epidemiologically, physically, and psychologically—and of how those changes can be anticipated and managed (Steffen et al., 2012).

Two organising ideas run through the sections below. The first is that the dominant physical hazard of a mass gathering is density and the flow dynamics it drives, not any supposed loss of reason in the crowd. The second is that the dominant psychological variable is social identity: whether attendees experience one another as a threatening mass or as fellow members of a meaningful collective determines how they feel, how they behave under pressure, and whether an emergency becomes a disaster. The history of the field is in large part the displacement of a folk theory with a long pedigree in crowd psychology—the panicking, irrational mob of Le Bon, 1896—by these two more exact accounts.

Dimensions of Mass-Gathering Research

No single discipline owns the mass gathering. The subject is studied along several dimensions at once—how events are defined and classified for planning; the physics of the dense crowd; the epidemiology and clinical burden the event concentrates; and the social psychology of safety, experience, and emergency behaviour. Each supplies a different lever on the same events, and the most useful analyses combine them. Table 1 summarises the dimensions, and the sections that follow develop each in turn.

Table 1

Dimensions of Mass-Gathering Research

Defining and Classifying Mass Gatherings

The field's first task was to make “mass gathering” a category precise enough to plan around. The settled approach classifies events along a handful of axes that predict their risk profile: whether the crowd is seated or mobile (a mobile crowd, free to surge toward an attraction, carries far more crush risk than a ticketed, seated one); whether the venue is bounded or open; the event's duration (a multi-day pilgrimage concentrates fatigue, heat, and sanitation problems a single evening does not); and the attendees' demographics and purpose, which shape both the medical baseline and the behaviour. A religious pilgrimage, a rock concert, and a marathon are all mass gatherings, but each loads a different combination of these axes, and good planning begins by typing the event rather than treating “crowd” as one thing (Memish et al., 2019). This classificatory work is what allows a host to forecast, from the event type and size, the medical usage rate, the staffing, and the physical layout a safe gathering will require (Steffen et al., 2012).

Crowd Dynamics and the Physics of Density

The decisive advance in crowd safety came not from psychology but from physics. Dirk Helbing and colleagues, analysing video of the Jamarat Bridge at the Hajj where repeated fatal disasters had occurred, showed that a dense crowd behaves like a driven many-particle system with its own phase transitions. As density rises, the crowd passes from smooth laminar flow, through a stop-and-go regime of visible density waves, into crowd turbulence—irregular, involuntary shoving in all directions, in which people are swept off their feet and the pressures between bodies become lethal. Crucially, the transition is a property of density and geometry, not of the crowd's state of mind: the people were not panicking, they were being moved by forces they could neither see nor resist (Helbing et al., 2007).

The key quantitative tool is the fundamental diagram relating density, walking speed, and flow (Figure 1). Specific flow J—the number of people passing a line per metre of width per second—is the product of density ρ and mean speed v. Because speed falls as the crowd packs together, flow does not rise indefinitely with density: it climbs to a maximum at a critical density and then collapses, so that beyond the peak a denser crowd moves fewer people per second, not more. This counter-intuitive congested branch is the heart of crowd-safety engineering: once a venue drives its crowd past the critical density, adding more people reduces throughput and raises pressure at the same time (Moussaïd et al., 2011). Mehdi Moussaïd and colleagues showed that the self-organised lanes and stripes of pedestrian traffic can be reproduced from a few simple behavioural rules—each walker adjusting heading and speed to keep a clear path—and that the same rules predict where those patterns break down into the clogging that precedes a disaster. The knowledge domain has since grown into a mature, data-rich discipline of its own (Haghani, 2021). Demo 1 lets the reader move the density along the fundamental diagram and watch speed, flow, and crush risk change.

Mass-Gatherings Medicine and Public Health

Alongside the physics grew a clinical specialty. Mass-gatherings medicine, consolidated largely through decades of research on the Hajj, studies the distinctive health burden of large events: the amplification and global dispersal of infectious disease, the environmental toll of heat and exertion, and the trauma of crush and stampede. Ziad Memish and colleagues set out the field's scope, showing that a gathering functions as an epidemiological amplifier—concentrating susceptible and infectious people, then scattering them home across continents—so that surveillance at the event is also global disease surveillance (Memish et al., 2019). The non-communicable burden is just as real: at a summer Hajj, heat-related illness is a mass phenomenon, and studies of pilgrims' knowledge and behaviour feed directly into prevention (Yezli et al., 2019). The infectious risk drives international policy, from vaccination requirements to global coordination on emerging pathogens (Abubakar et al., 2012). The COVID-19 pandemic made the stakes vivid: the decision whether to hold, modify, or cancel mass gatherings became a defining public-health dilemma, weighing the events' role in seeding transmission against their social, economic, and political importance (McCloskey et al., 2020). The historical record of crowd disasters—compiled to extract their recurring causes—shows the same lesson repeatedly: most deaths come from crushing in bottlenecks, not from fire or from the hazard that triggered the movement (Soomaroo & Murray, 2012).

Social Identity, Density, and Safety

Physics explains when a crowd can crush; it does not explain why the same density is experienced so differently by different people. That is where social psychology enters, and its central finding is counter-intuitive. Hani Alnabulsi and John Drury surveyed pilgrims at the Hajj and found that crowd density did not simply reduce felt safety: its effect was moderated by social identification. Pilgrims who identified strongly with the crowd reported feeling safer and more comfortable as density increased, whereas those who identified weakly felt more endangered—a genuine crossover, not merely a difference in degree. For the high identifier, the press of bodies is the press of fellow pilgrims; for the low identifier, it is the press of a threatening mass (Alnabulsi & Drury, 2014).

This is one instance of a broader programme on the psychology of health and well-being at mass gatherings. Nick Hopkins and Stephen Reicher argued that collective events are not merely risks to be managed but can be beneficial, and set out a research agenda for how shared identity converts a crowd from a hazard into a source of support and positive emotion (Hopkins & Reicher, 2016). Longitudinal work at Indian pilgrimages bears this out: participation in the Magh Mela and similar events was associated with improved well-being rather than the deterioration a stress model would predict, and the mechanism was the intensification of shared identity (Tewari et al., 2012; Khan et al., 2016). Survey studies of crowd events further show that it is the subjective experience of crowdedness, filtered through identification, that drives positive emotion—not objective density alone (Novelli et al., 2013). Demo 2 lets the reader vary density and identification and watch felt safety cross over between high and low identifiers.

Disasters, Crushes, and Collective Resilience

When a mass gathering does end in catastrophe, the popular explanation—that the crowd panicked and trampled itself—is almost always wrong. The historical pattern is that disasters are failures of space and flow: a crowd driven past its critical density in a bottleneck, a blocked exit, a surge toward a single point (Soomaroo & Murray, 2012). Even the archetypal “panic” events dissolve on inspection. Norris Johnson's reconstruction of the 1979 crush at a concert by The Who found no competitive, every-person-for-themselves stampede: people tried to help those who fell, and the deaths were caused by crowd pressure against locked doors, not by a collapse of social order (Johnson, 1987).

The positive account comes from the social identity approach to crowds, the framework that replaced the irrationalist tradition by explaining crowd action through the shared identity of the participants rather than a loss of reason (Reicher, 2001). In a comparative study of survivors—including the 2005 London bombings—Drury and colleagues found that a sense of common fate in an emergency transforms a physical crowd of strangers into a psychological group, and that this emergent shared identity motivates the orderly, mutually supportive, often self-sacrificing behaviour survivors actually report (Drury et al., 2009). Drury's integrative review consolidated the evidence that shared identity is the mechanism converting a crowd from a source of danger into a source of support, and the same framework reframed emergency planning away from “mass panic” toward treating the public as a capable resource (Drury et al., 2013; Drury, 2018). That people who share an identity literally move in coordination—a measurable behavioural signature of a psychological crowd—closes the loop back to the physics: solidarity is not only felt, it changes how a crowd flows (Templeton et al., 2018). Demo 3 lets the reader vary the degree of shared identity in a crush and read off how coordinated mutual aid rises as competitive pushing falls.

Worked Example

The mechanisms become concrete in three small exercises, each matching one of the demonstrations. The numbers are illustrative—they capture the qualitative shape of the published findings rather than reproducing any single reported statistic.

The fundamental diagram and critical density. Model walking speed as a function of density with a simple linear law, v = v₀(1 − ρ/ρ_max), taking a free-walking speed v₀ = 1.3 m/s and a jam density ρ_max = 5.4 persons/m². Specific flow is J = ρv = v₀ρ(1 − ρ/ρ_max), a downward parabola that peaks at the critical density ρ_c = ρ_max/2 = 2.7 persons/m². There, v = 1.3×(1 − 2.7/5.4) = 0.65 m/s and J = 2.7×0.65 = 1.76 persons/m·s—the maximum the walkway can carry. Push the crowd to ρ = 4.5 persons/m² and speed falls to v = 1.3×(1 − 4.5/5.4) = 0.22 m/s, so flow drops to J = 4.5×0.22 = 0.98 persons/m·s: half again as many people, yet little more than half the throughput, and dangerously high pressure. This is why the congested branch is the killer (Helbing et al., 2007). Demo 1 lets the reader move ρ and read off v, J, and the regime.

Identity moderates felt safety (the Hajj crossover). Represent felt safety S (0 to 1) as a function of normalised density d (0 to 1) and social identification s (0 to 1), with S = 0.5 + d(0.8s − 0.4). The sign of density's effect flips at s = 0.5. For a strong identifier, s = 1, at maximum density d = 1: S = 0.5 + 1×(0.8 − 0.4) = 0.90—density makes the crowd feel safer. For a weak identifier, s = 0, at the same density: S = 0.5 + 1×(0 − 0.4) = 0.10—the identical crush feels dangerous. Same physical crowd, opposite psychology, because the moderator is who the surrounding people are taken to be (Alnabulsi & Drury, 2014). Demo 2 lets the reader move d and s and watch the two curves cross.

Collective resilience in a crush. Model the fraction of a dense emergency crowd engaged in coordinated mutual aid—steadying neighbours, passing people through, moving as one—as a function of shared identity s (0 to 1): H = 0.15 + 0.7s, with competitive pushing the complement 1 − H. In a mere physical crowd with no shared identity (s = 0), H = 0.15 + 0.7×0 = 15% cooperate and 85% push for themselves—the mass-panic image. Let a common fate forge a shared identity (s = 1): H = 0.15 + 0.7×1 = 85% cooperate and only 15% push selfishly. The same crush, facing the same danger, resolves in opposite ways depending on whether a shared identity has emerged—which is why disaster research finds mutual aid, not stampede, to be the norm (Drury et al., 2009). Demo 3 lets the reader vary s and trace cooperation against competitive pushing.

Discussion

Read across its disciplines, the science of mass gatherings converges on a single corrective to folk intuition. The folk theory holds that large crowds are dangerous because they make people irrational—that density breeds panic and panic breeds catastrophe. The evidence inverts almost every clause. Density is indeed the central hazard, but it kills through physics, not psychology: beyond a critical density the crowd's flow collapses and the pressures between bodies become lethal while the people themselves remain, as far as they can tell, perfectly rational and often actively trying to help (Helbing et al., 2007; Johnson, 1987). “Mass panic” is rare; what disasters reveal is usually a failure of space, flow, and management, dressed up after the fact as a failure of nerve (Soomaroo & Murray, 2012).

The psychological contribution is not to restore the panicking mob but to explain the striking variability that physics leaves open: why the same density is comfortable at one gathering and terrifying at another, and why some crowds in extremis cooperate while others fragment. The answer is social identity. When attendees share a meaningful identity, dense crowds are experienced as safe and even restorative, participation improves well-being, and emergencies call forth collective resilience and mutual aid; when they do not, the same conditions read as threat (Alnabulsi & Drury, 2014; Hopkins & Reicher, 2016; Drury, 2018). The practical synthesis, now embedded in mass-gatherings medicine and crowd-safety engineering alike, is that a safe gathering is one designed for both truths at once: kept below critical density by its physical layout, and planned around—rather than against—the shared identity of the people it assembles.

Current Directions

The most visible recent work was forced by the COVID-19 pandemic, which turned every mass gathering into a public-health decision and made the modelling of event-driven transmission an urgent, policy-facing science (McCloskey et al., 2020). The pandemic also sharpened a longer-running question in the field: how to weigh the measurable risks of assembly against the equally real benefits—social, psychological, economic—that the social identity research had begun to quantify, rather than treating crowds as pure hazard (Hopkins & Reicher, 2016).

A second thread is methodological. Crowd dynamics has matured into a data-rich discipline with its own bibliometric self-study, drawing on high-resolution tracking, simulation, and the analysis of real disasters to move from description toward prediction (Haghani, 2021). The behavioural-signature work—showing that people who share an identity coordinate their movement measurably—points toward integrating the psychological and physical models that have so far run in parallel, so that a single framework might predict both how a crowd feels and how it flows (Templeton et al., 2018). Open questions remain about how far findings drawn heavily from the Hajj and from Western sporting and music events generalise across cultures and event types, and about how best to translate the identity findings into layouts, stewarding, and communication that actively build rather than erode the crowd's own solidarity.

Common Misconceptions

Crowd disasters are caused by people panicking and stampeding.

Most are not. The typical fatal crowd event is a crush produced by density and flow—a bottleneck, a blocked exit, a surge past the critical density—while the people involved remain rational and often try to help. Investigations repeatedly find failures of space and management, not a collapse of nerve (Johnson, 1987; Soomaroo & Murray, 2012).

A denser crowd always moves more people.

Only up to a point. Because walking speed falls as density rises, flow peaks at a critical density of roughly 2.7 persons/m² and then declines: beyond the peak, adding people reduces throughput while raising pressure. The congested branch of the fundamental diagram is precisely the dangerous regime (Helbing et al., 2007).

High density is simply unpleasant and unsafe for everyone.

Density's psychological effect depends on identity. At the Hajj, pilgrims who identified strongly with the crowd felt safer and more positive as density rose, while weak identifiers felt endangered—a crossover, not a uniform decline. The same physical crowd is two different experiences (Alnabulsi & Drury, 2014).

Attending a huge crowd event is bad for one's well-being.

Not necessarily. Longitudinal studies of pilgrimages find that participation in a mass gathering can improve well-being, because the intensified shared identity it produces is a source of social support and positive emotion rather than only of stress (Tewari et al., 2012; Hopkins & Reicher, 2016).

Glossary

Collective resilience.

The mutually supportive, coordinated behaviour a crowd of strangers displays in an emergency once a shared identity emerges from their common fate; the empirical rule in crowd disasters, in place of mass panic.

Common fate.

A shared predicament—above all a shared danger—that turns a physical crowd of strangers into a psychological group with a common identity; the trigger for collective resilience in emergencies.

Critical density.

The density at which a crowd's specific flow reaches its maximum (roughly 2.7 persons/m² in simple models); beyond it, flow declines and crush risk rises as more people yield less throughput.

Crowd turbulence.

The chaotic, involuntary shoving in all directions that emerges at very high density, in which people are moved by pressures between bodies rather than by their own volition; the immediate physical cause of many crush deaths.

Fundamental diagram.

The relationship between crowd density, walking speed, and flow; because speed falls as density rises, flow climbs to a peak at the critical density and then collapses toward zero as the crowd jams.

Jam density.

The density at which a crowd is packed so tightly that movement halts and specific flow falls to zero; the upper limit of the fundamental diagram.

Laminar flow.

Smooth, orderly crowd movement at low to moderate density, before the onset of stop-and-go waves and crowd turbulence.

Mass gathering.

A planned, time-bounded event assembling a number of people large enough to strain the host community's planning and response resources, in a defined space for a shared purpose.

Mass panic.

The popular model of emergency crowds as reduced by fear to a selfish, antisocial stampede; largely a myth, since most crowd deaths are caused by density and flow, not by a breakdown of social order.

Mass-gatherings medicine.

The public-health specialty, consolidated largely around the Hajj, that studies and mitigates the infectious, environmental, and trauma risks concentrated by large events.

Medical usage rate.

The expected number of attendees requiring medical care per unit of attendance at an event; a planning metric forecast from the event's type, size, duration, and demographics.

Social identification.

The extent to which a person defines themselves as a member of a given group; at a mass gathering it moderates whether dense crowding is experienced as safe and supportive or as threatening.

Social identity approach.

The framework that explains crowd behaviour through the shared social identity of the participants rather than a loss of reason; it accounts for both felt safety under density and collective resilience in emergencies.

Specific flow.

The number of people passing a line per metre of its width per second; the product of density and walking speed, and the quantity a safe venue tries to keep below the maximum of the fundamental diagram.

Stop-and-go waves.

Visible density waves travelling backward through a dense crowd as movement alternates between halting and surging; the transitional regime between smooth flow and crowd turbulence.

Key Researchers

John Drury

(Professor of Social Psychology, University of Sussex). Leads the Crowds and Identities research group; with Hani Alnabulsi showed that social identification moderates the effect of crowd density on safety at the Hajj, and established collective resilience as the typical response to mass emergencies. ORCID · Google Scholar · Wikipedia

Dirk Helbing

(Professor of Computational Social Science, ETH Zurich). Physicist and computational social scientist whose empirical and simulation studies of crowd disasters established the fundamental diagram and the stop-and-go and turbulence regimes that explain fatal crushes. Google Scholar · Faculty Page · Wikipedia

Nick Hopkins

(Emeritus Professor of Psychology, University of Dundee). Social psychologist of collective identity who, with Stephen Reicher, set the research agenda on the well-being effects of mass gatherings and the Indian pilgrimage studies. ORCID · Google Scholar · Faculty Page

Ziad A. Memish

(Alfaisal University and King Saud Medical City, Riyadh; Adjunct Professor, Rollins School of Public Health, Emory University). Founding figure of mass-gatherings medicine, lead author of the field-defining Hajj public-health literature on infectious and environmental risk.

Mehdi Moussaid

(Research Scientist, Center for Adaptive Rationality, Max Planck Institute for Human Development, Berlin). Behavioural scientist of crowd dynamics who showed that pedestrian self-organisation and the breakdown that precedes disasters follow from a few simple behavioural rules. Google Scholar · Wikipedia

Stephen D. Reicher

(Bishop Wardlaw Professor of Social Psychology, University of St Andrews). Originator of the social identity approach to crowds, senior author of the mass-gatherings well-being programme and of the reframing of crowd emergencies as occasions for collective resilience. Google Scholar · Faculty Page · Wikipedia

Frequently Asked Questions

What is a mass gathering?

A mass gathering is a planned, time-bounded event—such as a pilgrimage, a sporting final, a festival, or a rally—that brings together a number of people large enough to strain the host community's planning and emergency resources. The defining feature is the ratio of attendance to available resources, not a fixed headcount (Memish et al., 2019).

What is mass-gatherings medicine?

It is the public-health specialty, developed largely through research on the Hajj, that studies and manages the health risks such events concentrate: the amplification and global spread of infectious disease, environmental threats such as heat, and the trauma of crush and stampede (Memish et al., 2019; Steffen et al., 2012).

What actually causes crowd disasters?

Most crowd disasters are caused by density and flow, not by panic. When a crowd is driven past its critical density—at a bottleneck, a blocked exit, or a surge toward one point—flow collapses and the pressure between bodies becomes lethal, while the people involved typically remain rational and try to help (Helbing et al., 2007; Soomaroo & Murray, 2012).

What is the fundamental diagram of crowd flow?

It is the relationship between crowd density, walking speed, and flow. Because speed falls as density rises, flow increases only up to a critical density of roughly 2.7 persons per square metre and then declines—so beyond that peak a denser crowd moves fewer people per second, which is the dangerous regime (Helbing et al., 2007; Moussaïd et al., 2011).

Does mass panic really happen?

Rarely. The “mass panic” image is largely a myth. In real emergencies people orient toward helping those around them, and a shared identity born of common fate produces collective resilience, making mutual aid—not selfish stampede—the typical response (Johnson, 1987; Drury et al., 2009).

Why does the same dense crowd feel safe to some people and dangerous to others?

Because social identification moderates the experience of density. Research at the Hajj found that pilgrims who identified strongly with the crowd felt safer and more positive as density rose, whereas weak identifiers felt more endangered—so the same physical crowd is psychologically two different situations (Alnabulsi & Drury, 2014).

Can attending a mass gathering be good for a person?

Yes. Longitudinal studies of large pilgrimages find that participation can improve well-being rather than harm it, because the intensified shared identity the event produces is a source of social support and positive emotion (Tewari et al., 2012; Hopkins & Reicher, 2016).

How did COVID-19 change the study of mass gatherings?

The pandemic turned the decision to hold, modify, or cancel an event into a central public-health dilemma, weighing the role of gatherings in seeding global transmission against their social, economic, and political value, and it accelerated the modelling of event-driven disease spread (McCloskey et al., 2020).

References

Abubakar, I., Gautret, P., Brunette, G. W., Blumberg, L., Johnson, D., Poumerol, G., Memish, Z. A., Barbeschi, M., & Khan, A. S. (2012). Global perspectives for prevention of infectious diseases associated with mass gatherings. The Lancet Infectious Diseases, 12(1), 66–74. https://doi.org/10.1016/S1473-3099(11)70246-8

Alnabulsi, H., & Drury, J. (2014). Social identification moderates the effect of crowd density on safety at the Hajj. Proceedings of the National Academy of Sciences, 111(25), 9091–9096. https://doi.org/10.1073/pnas.1404953111

Drury, J. (2018). The role of social identity processes in mass emergency behaviour: An integrative review. European Review of Social Psychology, 29(1), 38–81. https://doi.org/10.1080/10463283.2018.1471948

Drury, J., Cocking, C., & Reicher, S. (2009). Everyone for themselves? A comparative study of crowd solidarity among emergency survivors. British Journal of Social Psychology, 48(3), 487–506. https://doi.org/10.1348/014466608X357893

Drury, J., Novelli, D., & Stott, C. (2013). Representing crowd behaviour in emergency planning guidance: ‘Mass panic’ or collective resilience? Resilience, 1(1), 18–37. https://doi.org/10.1080/21693293.2013.765740

Haghani, M. (2021). The knowledge domain of crowd dynamics: Anatomy of the field, pioneering studies, temporal trends, influential entities and outside-domain impact. Physica A: Statistical Mechanics and its Applications, 580, 126145. https://doi.org/10.1016/j.physa.2021.126145

Helbing, D., Johansson, A., & Al-Abideen, H. Z. (2007). Dynamics of crowd disasters: An empirical study. Physical Review E, 75(4), 046109. https://doi.org/10.1103/PhysRevE.75.046109

Hopkins, N., & Reicher, S. (2016). The psychology of health and well-being in mass gatherings: A review and a research agenda. Journal of Epidemiology and Global Health, 6(2), 49–57. https://doi.org/10.1016/j.jegh.2015.06.001

Johnson, N. R. (1987). Panic at “The Who Concert Stampede”: An empirical assessment. Social Problems, 34(4), 362–373. https://doi.org/10.2307/800813

Khan, S. S., Hopkins, N., Reicher, S., Tewari, S., Srinivasan, N., & Stevenson, C. (2016). How collective participation impacts social identity: A longitudinal study of the Magh Mela. Political Psychology, 37(3), 309–325. https://doi.org/10.1111/pops.12260

Le Bon, G. (1896). The crowd: A study of the popular mind. T. Fisher Unwin. (Original work published 1895 as Psychologie des foules)

McCloskey, B., Zumla, A., Ippolito, G., Blumberg, L., Arbon, P., Cicero, A., Endericks, T., Lim, P. L., & Borodina, M. (2020). Mass gathering events and reducing further global spread of COVID-19: A political and public health dilemma. The Lancet, 395(10230), 1096–1099. https://doi.org/10.1016/S0140-6736(20)30681-4

Memish, Z. A., Steffen, R., White, P., Dar, O., Azhar, E. I., Sharma, A., & Zumla, A. (2019). Mass gatherings medicine: Public health issues arising from mass gathering religious and sporting events. The Lancet, 393(10185), 2073–2084. https://doi.org/10.1016/S0140-6736(19)30501-X

Moussaïd, M., Helbing, D., & Theraulaz, G. (2011). How simple rules determine pedestrian behavior and crowd disasters. Proceedings of the National Academy of Sciences, 108(17), 6884–6888. https://doi.org/10.1073/pnas.1016507108

Novelli, D., Drury, J., Reicher, S., & Stott, C. (2013). Crowdedness mediates the effect of social identification on positive emotion in a crowd: A survey of two crowd events. PLoS ONE, 8(11), e78983. https://doi.org/10.1371/journal.pone.0078983

Reicher, S. (2001). The psychology of crowd dynamics. In M. A. Hogg & R. S. Tindale (Eds.), Blackwell handbook of social psychology: Group processes (pp. 182–208). Blackwell. https://doi.org/10.1002/9780470998458.ch8

Soomaroo, L., & Murray, V. (2012). Disasters at mass gatherings: Lessons from history. PLoS Currents, 4, RRN1301. doi:10.1371/currents.RRN1301. https://pmc.ncbi.nlm.nih.gov/articles/PMC3271949/

Steffen, R., Bouchama, A., Johansson, A., Dvorak, J., Isla, N., Smallwood, C., & Memish, Z. A. (2012). Non-communicable health risks during mass gatherings. The Lancet Infectious Diseases, 12(2), 142–149. https://doi.org/10.1016/S1473-3099(11)70293-6

Templeton, A., Drury, J., & Philippides, A. (2018). Walking together: Behavioural signatures of psychological crowds. Royal Society Open Science, 5(7), 180172. https://doi.org/10.1098/rsos.180172

Tewari, S., Khan, S., Hopkins, N., Srinivasan, N., & Reicher, S. (2012). Participation in mass gatherings can benefit well-being: Longitudinal and control data from a North Indian Hindu pilgrimage event. PLoS ONE, 7(10), e47291. https://doi.org/10.1371/journal.pone.0047291

Yezli, S., Mushi, A., Yassin, Y., Maashi, F., & Khan, A. (2019). Knowledge, attitude and practice of pilgrims regarding heat-related illnesses during the 2017 Hajj mass gathering. International Journal of Environmental Research and Public Health, 16(17), 3215. https://doi.org/10.3390/ijerph16173215