Why crowd behaviour matters - for psychology and society - my professorial lecture

The August 2024 riots: Empowerment of the xenophobes

The riots in some English towns this week came 13 years almost to the day after the wave of riots that swept England following the killing of Mark Duggan in London, 2011. There are some parallels, but also some crucial differences. The 2011 riots were driven to a significant extent by anti-police anger and resentment. The 2024 riots are driven by racism: xenophobia and Islamophobia. The other parallel is with what happened immediately after the Brexit vote in 2016 and the Trump victory in the same year. Each was followed by a spike in hate crimes, most of which were ‘racially’ motivated. The Trump victory was also followed by an upsurge in far-right street mobilizations.

A lot has been said about the role of ‘coordination’ in the August 2024 xenophobic riots. But there is a significant ‘spontaneous’ element that also needs explanation, as does the relation between coordination and spontaneity in collective events like these. This is where social psychology comes in.

A short answer to the question of why the August riots spread is meta-perception and collective empowerment of specific group identities. ‘Meta-perception’ refers to what people believe about what other people believe. Collective empowerment in events like these can have different sources.

First, it can arise from a belief about public opinion. With the unexpected Brexit and Trump results in 2016, xenophobes suddenly (and mistakenly) believed that their racist views were widely shared. They thought they would be supported if they took threatening and violent action against those they perceived as ‘foreign’. The Brexit and Trump hate-crime spikes were driven by empowerment, not just grievance.

Why might people at the first of the August wave of xenophobic riots (Southport, 30^th July) feel their views were widely shared? Social media has been highlighted. But note also that, just three days before, ‘thousands of people’ marched in London on the ‘Tommy Robinson’ ‘patriots’ demo. A big demonstration is a way of building a sense that ‘you are not alone’, that many others think like you, and so of capacity building.

Second, empowerment can arise within the events. Where police appear increasingly unable to prevent the rioting, people in further locations gain a sense of vicarious empowerment. They believe the police locally won’t cope either. Although activists have been identified as leading players in the far-right riots, where others share with them the police as a common outgroup, they too can become empowered. This helps explain why other people less committed to the xenophobic ideology also start joining in.

Moreover, enacting their xenophobic identity in a way they are not usually able (through attacks on shops, hotels etc. but also defeating the police) is itself empowering to them because it reverses normal social relations. It enhances their sense of agency. It fuels another kind of spread. Those involved now feel confident to do the same again, perhaps travelling to another location to do so.

Third, collective empowerment can arise from the expectation that like-minded others will come out onto the streets locally. The precondition for this expectation is the participant’s belief that there are others locally who broadly agree with them and who therefore constitute an ‘us’ or ‘we’. When there is a critical mass of such people, a riot can occur. When riots happen in other locations over ‘similar’ issues, such people increasingly believe that people will now come out on to their own streets locally. Belief that others like self will come onto the street provides expectations of support for ingroup normative action (in this case xenophobic action) and hence the intention to act. The social media calls from the far-right activists could help promote the expectation also, by creating an impression of widespread intentions.

Solutions to this crisis of xenophobic riots from a social psychological perspective address these empowerment processes at source. First, potential participants need to understand that there is not the wider public support they think there is for their views, that the opposite is the case. Second, prevent them mobilizing and marching, to limit that capacity-building. Third, prevent their actions (including smaller acts of hate) from having a tangible impact – prevent them from turning their subjective identity into objective reality – by negating and cancelling out their effects. And fourth, as it is particular identities that are empowered or disempowered, assert and support collective identities antagonistic to theirs. For example, well-organized and -attended groups and activities based on international class solidarity help to defeat racism and xenophobia on the streets by making such solidarity more realistic than the racist vision.

This analysis and these kinds of actions also address the wider problem of why people support tyrannical systems such as fascism. These movements gain much of their support from being seen able to put their beliefs into practice. Their capacity for organizing and achieving their goals legitimizes them and gives them credibility. Undermining their capacity to act and organize undermines their credibility in the eyes of others who might consider supporting them. 

Six zombie ideas in crowd psychology

 

What are zombie ideas? These are ideas that keep coming back, even though they have been thoroughly refuted by the evidence. They should be dead, but they won’t stay dead! They keep coming back because they serve certain interests or prejudices (or both). Here are six zombie ideas in crowd psychology that keep cropping up in everyday talk, in the news, among policymakers and practitioners, and in academic publications. And here’s why they’re wrong.

1. De-individuation

The most distinctive claim in the ‘de-individuation’ family of theories was inherited from Gustave Le Bon – the idea that being anonymous leads to a loss of self and hence uncontrolled, anti-normative behaviour. This idea could not cope with the evidence that conditions of anonymity in fact are associated with a wide range of behaviours, including accentuation of pro-social behaviours. There is little evidence that anonymity leads to a ‘de-individuated’ state of reduced private self-awareness. Rather, anonymity makes group identities more salient and hence leads to more, not less, conformity to relevant situational norms.

Key reading: Postmes & Spears (1998)

Gustave Le Bon headstone

(Pierre-Yves Beaudouin / Wikimedia Commons / CC BY-SA 4.0)

2. Groupthink

While groupthink is supposedly a pitfall of small groups and organizations rather than crowds, I include it here as it’s another example of an anti-collectivist concept. It is used loosely by commentators to refer to any situation where group members prioritise the group’s own ideas over critical or external views. For example, some of those involved in decision-making at the height of the Covid pandemic have used the idea of groupthink to explain organizational failures in decision-making. The distinctive claim of the ‘groupthink’ concept is that highly cohesive groups will be subject to concurrence-seeking at the expense of critical inquiry, leading to faulty decisions. A big problem for this idea is that there is not much evidence that greater cohesiveness leads to worse decision-making. Rather than the tendency to ignore critical evidence being a function of groupness, it’s more likely to be an effect of particular group norms (for example that value loyalty).

Key reading: Aldag (2022)

3. Mass panic

Aside from the profound problems of judging whether behaviour counts as ‘panic’ in an emergency (what is reasonable behaviour in this situation?) and the related problem of trying to import a polysemic everyday term into scientific explanation, there is another basic problem. There is no evidence that people in crowds are typically uncontrolled, selfish or competitive in emergencies. The common finding of social support among people in emergencies adds to the problems of this concept.

Key reading: Clarke (2002)

4. Contagion

One of the most popular concepts in the social and behavioural sciences, ‘contagion’ is often used synonymously with spread and social influence. But there is little evidence that mere exposure alone is sufficient to prompt emulation. Group boundaries in the transmission of behaviours and emotions demonstrate this. Even for supposedly basic processes like so-called emotional contagion, reviews of the evidence suggest that the mimicry involved is not automatic, but rather relates to communication goals that already involve an emotional orientation to the other person.

Key reading: Drury et al. (2019)

5. The hooligan

The hooligan is a concept from sociology more than psychology, but it is a good example of a dispositional explanation. For the earliest beginnings of crowd psychology as a science, some have claimed that crowd conflict occurs through the convergence of certain kinds of individuals (usually with criminal, violent, or poorly socialized dispositions). From the 1960s urban riots in the USA to the 2011 English riots, proponents of such ideas have failed to produce the required evidence. In the football context, of course some groups seek conflict, but this in itself can’t explain collective behaviour. As Stott and Pearson explain, the 'hooligan' concept has little explanatory power: ‘disorder’ sometimes occurs when known ‘hooligans’ are not present; and when known ‘hooligans’ are present, ‘disorder’ doesn’t always take place.

Key reading: Pearson & Stott (2022)

6. Mob mentality

An overarching zombie idea, that links many of the above, but which also includes the distinctive claims that in crowds people revert to a simpler, less intelligent, and more primitive or archaic psychology, under the influence of which behaviour tends to gullibility, barbarism, loss of control, and violence. The fundamental problem here is two-fold. First, if this is a real tendency it cannot easily explain the majority of crowds, which are peaceful and pro-social. Second, the suggestion of a universal tendency like this cannot explain the social form of behaviour when there is crowd violence. To explain the distinct targets of the sans culottes, urban rioters, football fans and many others, and the sophistication in even the most violent crowd, it makes better sense to refer to their identities, group norms, and values.

Key reading: Reicher (1984)

Case study: UK Covid mutual aid groups

 

By John Drury & Evangelos Ntontis

In 2020, tens of thousands of people got involved in Covid mutual aid and similar community support groups, with over 4000 new groups being set up in Spring of that year.

Who were they?

Many participants were new to volunteering or community action. Some groups were repurposed pre-existing community groups. Groups tended to be informal, distinct from the existing voluntary sector, and with no formal constitution. Some groups later applied for charitable status to access grants more easily. Local communities with more social capital tended to have more mutual aid groups.

What did Covid mutual aid groups do?

Mutual aid groups’ main activity was shopping to support those self-isolating or shielding. They also engaged in other community support activities, including fundraising, providing information, dog-walking, mental health support, and collecting prescriptions.  Some pointed out that mutual aid groups were crucial in the UK’s pandemic response. In addition, many groups sought to respond to other community needs beyond Covid, including food poverty and supporting refugees.

Understanding how Covid mutual aid groups sustained themselves

The Economic and Social Research Council funded research to examine how Covid mutual aid groups sustained themselves over time. Following the initial upsurge, participation in mutual aid groups dropped, particularly after ‘lockdown’ restrictions eased. For example, activity in Covid mutual aid groups on Facebook dropped by 75% by June from the high point of March 2020. Some volunteers left because they felt let down by local authorities, needed logistical infrastructure, felt overwhelmed, lacked capacity, or lost motivation due to return to ‘normality’. For the groups that continued, there was a need to sustain themselves and maintain volunteers’ engagement over time.

Interviews with organizers and a two-wave survey of volunteers indicated three types of factors that helped sustain groups. First, there was group scaffolding – such as access to funds, space for meetings and storage, computing facilities, and transport.

Second, there were group experiences which arose from participation and motivated further involvement -- including a sense of identity, wellbeing, empowerment, and skills acquisition. Finally, organizers employed various group strategies to enhance a sense of belonging and motivate volunteers – in particular, fostering a culture of care, holding social events, a flexible leadership structure, and regular communication.

Learnings: Implications for community resilience

Central government, local authorities, and local infrastructure organizations/ the formal voluntary sector can all help scaffold the group processes that sustain mutual aid groups.

Group scaffolding can comprise financial/ practical support, connections and links, and guidance / advice.

It is important that no ‘strings’ are attached to this external support, as it is precisely the identity of mutual aid groups as independent and informal that makes them trusted by communities.

Understanding collective flight responses to (mis)perceived hostile threats in Britain 2010-2019: a systematic review of ten years of false alarms in crowded spaces

Understanding collective flight responses to (mis)perceived hostile threats in Britain 2010-2019: a systematic review of ten years of false alarms in crowded spaces: (2022). Understanding collective flight responses to (mis)perceived hostile threats in Britain 2010-2019: a systematic review of ten years of false alarms in crowded spaces. Journal of Risk Research. 

HOW CAN WE SAFELY RE-OPEN LIVE EVENTS?

 New briefing from Independent SAGE: https://www.independentsage.org/how-can-we-safely-re-open-live-events/ 

Don’t personalise, collectivise!

Don’t personalise, collectivise!

The way we deal with the coronavirus is bound up with the way we think about society and about the individual. And the problem is that we are in danger of getting it wrong on all counts, with the consequence that we will be less effective in containing the virus. There is nothing new about us being wrong. But this time, lives are at stake.

The   assumption, which is reflected in the advice being handed out to the public, is that the way to change behaviour is to appeal to individual interests. To make sure people take notice, personalise the message: ‘change your behaviour so that you will survive’. Surely that makes sense? Well no. It is precisely the wrong thing to do. Here’s why. 

At a practical level, those least at risk (young, fit, healthy) may well feel it isn’t worthwhile to make the necessary changes and so continue to act in ways that put the most vulnerable (old and infirm) at risk of infection. Additionally, at a moral level, we have the right to disregard dangers to ourselves and some even glory in being risk takers. It might be foolish, but it isn’t disreputable to ignore safety advice.

On top of this, if we frame things individually – look after yourself! – we run into difficulties when it comes to getting people to behave in ways that are inconvenient to themselves but benefit others (self quarantining, for instance). The same goes when it comes to distributing scarce resources (doctors time, medicines, hand gel etc.). If we prioritise the individual then the strongest rather than the neediest will win out. In both cases, the pursuit of self-interest is inefficient, it undermines the overall response to the crisis and many more will die. 

Our own research on emergencies (Drury et al., 2019)  shows that it is precisely when people stop thinking in terms of ‘I’ and start thinking in terms of ‘we’ – more technically, when they develop a sense of shared social identity – that they start to coordinate, support each-other and ensure that the neediest get the greatest help. Sometimes this sense of shared identity emerges by the very fact of experiencing a common threat. But messaging also matters. When a threat is framed in group rather than individual terms, the public response is more robust and more effective (Carter et al., 2013).

So, let’s look again at the coronavirus response. Instead of personalising the issue we need to collectivise it. The key issue is not so much ‘will I survive’ as ‘how do we get through it’. The emphasis must lie on how we can act to ensure that the most vulnerable amongst us are protected and losses to the community are minimised – after all, from a collective perspective, a loss to one is a loss to all.

If framed in this way, then it becomes important for everyone to wash their hands and cover their coughs because of the implications for others as well as for themselves. Moreover, while we might have a right to take risks for ourselves, we have a moral obligation to avoid imposing risks on others (especially those who are vulnerable and connected to us – just think how your driving changes when you have children in the car). Both of these considerations are powerful motivators of action (Reicher & Haslam, 2009).

What is more, once certain actions become communal issues subject to collective norms, then violating them invokes collective pressure. The best way to stop people going out when unwell or demanding resources they need less than others is not simply to change internal motivations but also to mobilise external disapproval. The feverish person who goes to work, the fit young person demanding access to A&E will be best dissuaded when the community comes together to make clear that these are not acceptable behaviours.

Once you collectivise the response to coronavirus, and once you create clear norms about maximising community well-being, then you become less reliant upon external forces such as the police to regulate behaviours – say around who is prioritised in getting medical help – with all the risk of clashes that entails. Instead, the community itself will constrain would-be deviants in their midst. As always, the best regulation is collective self-regulation (Reicher et al., 2004). 

The difficulty with this approach, of course, is that it is so much at odds with contemporary psychological commonsense, which insists that behaviour is governed by individual self interest. It is also at odds with social changes which relentlessly undermine communities and collectivities, seek to transform social groups into individual consumers and view every relationship as a market based interpersonal exchange. In this sense, perhaps coronavirus is a powerful wake-up call. 

We have to change the way we frame the epidemic.

We have to change we see the individual and society.

We have to collectivise – or we die.

Stephen Reicher, University of St. Andrews

John Drury, University of Sussex

References

Carter, H., Drury, J., Rubin, G., Williams, R. and Amlôt, R. (2013), "The effect of communication during mass decontamination", Disaster Prevention and Management, Vol. 22 No. 2, pp. 132-147. https://doi.org/10.1108/09653561311325280

Drury, J., Carter, H., Cocking, C., Ntontis, E., Tekin Guven, S., & 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 (141) doi: 10.3389/fpubh.2019.00141

Drury, J., & Alfadhli, K. (2019). Social identity, emergencies and disasters. In R. Williams, S. Bailey, B. Kamaldeep, S. A. Haslam, C. Haslam, V. Kemp, & D. Maughan (Eds). Social scaffolding: Applying the lessons of contemporary social science to health, public mental health and healthcare. London: Royal College of Psychiatrists.

Drury, J., Cocking, C., & Reicher, S. (2009). The nature of collective resilience: Survivor reactions to the 2005 London bombings. International Journal of Mass Emergencies and Disasters, 27(1), 66-95.

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.

Reicher, S. D., & Haslam, S. A. (2009). Beyond help: a social psychology of social solidarity and social cohesion. In M. Snyder, & S. Sturmer (Eds.), The Psychology of Prosocical Behaviour Oxford: Wiley-Blackwell. 

Vilas, X., & Sabucedo, J. M. (2012). Moral obligation: A forgotten dimension in the analysis of collective action. Revista de Psicología Social, 27(3), 369-375.

Reicher, S., Stott, C., Cronin, P. and Adang, O. (2004), "An integrated approach to crowd psychology and public order policing", Policing: An International Journal, Vol. 27 No. 4, pp. 558-572. https://doi.org/10.1108/13639510410566271

Stott, C., Adang, O., Livingstone, A., & Schreiber, M. (2008). Tackling football hooliganism: A quantitative study of public order, policing and crowd psychology. Psychology, Public Policy, and Law, 14(2), 115-141. doi:10.1037/a0013419

Originally published in The Psychologist: https://thepsychologist.bps.org.uk/volume-33/april-2020/coronavirus-psychological-perspectives

The social relations of 'stop & search'

https://www.theguardian.com/commentisfree/2019/jan/31/police-stop-and-search-riots-2011-london

How participation in collective action can change lives, and how those changes endure over time

by Sara Vestergren

https://blogs.sussex.ac.uk/crowdsidentities/2018/08/07/how-participation-in-collective-action-can-change-lives/ 

Responding to People’s Emotional Needs in an Emergency

https://www.festivalinsights.com/2018/07/responding-peoples-emotional-emergency/

A 10-point listicle on journalist accounts of the public responses to events in Oxford Street, November 2017

A brief commentary for journalists and others on the events in Oxford Street, Friday 24th November 2017, when members of the public responded to fears of gunfire or a terrorist attack.

1.    In the footage from Oxford Street, many people are looking where they are going, assessing the situation, or otherwise moving in an orderly way, or standing around etc. Only a few people scream.

2.    Research evidence on emergency evacuations shows that coordination and order is common. Pushing and trampling occurs rarely and tends to be in narrow exits that are unfamiliar. Coordination reduces when people don’t see themselves as a ‘we’ or ‘us’.

3.    It is not clear that the number of injuries reported is particularly high for the number of people involved. 

4.    Research on fires shows that people often underestimate danger and/ or delay their exit (to respond, to stay with others). This, not over-reaction, is a main cause of fatalities.

5.    Research shows that there isn’t indiscriminate ‘contagion’. People attend most to, and follow the example of, those they judge to be relevant for the context, who are often people they define as similar to self.

6.    The word ‘panic’ was common in the news accounts, but researchers reject the term because it is hard to evidence.

7.    What is gained by saying ‘panic’ (rather than 'fleeing') is the implication that the behaviour is an overreaction, is unreasonable. But what should people do when the information they have is that there is a threat? ‘Sudden fear’ or ‘fleeing’ are more neutral terms.

8.    Journalists seemed to describe responses as ‘panic’ because it wasn’t after all a real emergency; but in many real emergencies people don’t know whether it is real or not (including fires and terrorist attacks).

9.    A first irony of the journalists’ use of the term ‘panic’ for the public response is that the government’s advice is to ‘run’ (this is not necessarily an endorsement).

10. A second irony is that the image of a vulnerable panic-prone public, rather than collectively resilient, is precisely what ISIS and others seek to achieve.