Crowd responses to Oxford Circus incident

Londoners may well be breathing a collective sigh of relief after the evacuation of Oxford Circus tube station and subsequent lock-down of shops on Oxford Street, was unjustified because the feared terrorist attack did not materialize. However, as with most coverage of such incidents, media reports have been peppered with the use of irrationalist terms (such as 'panic' & 'stampede'), and it has resulted in a fair degree of media requests for crowd psychologists to do interviews (the ones I know of are listed below). Regular readers of this blog will no doubt be familiar that myself and others who work in the field of crowd safety management are highly critical of the use of such terms to describe crowd emergency behaviour & I regularly critique coverage of such emergencies (most recently, the failed bomb attack at Parsons Green tube station, this September). This time the media are particularly keen on using these terms, because there wasn't actually a 'real' emergency, and so this presumably confirms that people were over-reacting and hence 'panicking'. However, I will argue in this blog that it's very easy to write off crowd responses as 'panic' without properly examining the actual behaviors of those involved, or the social context in which they are operating.

I will start with the general point that it's misleading to describe incidents such as these as 'panic', because it implies irrational behaviour, and that if there are any casualties, it is the fault of those who 'panicked'. Furthermore, one could say that it is a perfectly logical act for people to flee what they believe is a potentially life-threatening incident. Ironically, under-reaction in emergencies (especially fires) can be a worse problem, as if people delay action to escape possible danger, this could decrease their chances of a successful evacuation. In his blog, John Drury provides a 10 point guide for journalists, and he highlights the irony that government advice during terrorist incidents is to 'Run, Hide, Tell', but when people do follow this advice, the media call it 'panic'! 

With relation to specifically happened at Oxford Circus, the media seem keener than usual to describe it as 'panic' because it turned out not to be a terrorist attack. However, the people involved in this incident would not have had the hindsight of external observers & so may not have known the apparent trigger for the incident (a fight on the platform), so fleeing when they saw a crowd surge towards them (rather than waiting to see what the threat is) may have been the most sensible thing to do given the information they had available at the time. A similar incident happened at Los Angeles (LAX) airport in August 2016 after people mistakenly believed that there had been gunshots and fled the airport terminal. I argued at the time that considering the general fear in the US of terrorist attacks at airports post 9/11, acting on such an assumption should not be seen as 'irrational', but normative behavior within a social context. 

Much has also been made of the role of rumours circulating on social media about this incident, and the singer songwriter Olly Murs got into a public spat with Piers Morgan after tweeting to his 7.83 million followers on Twitter that there had been gunshots in Selfridges. I would certainly advocate that the information disseminated in emergencies should be as accurate as possible, and that people with large numbers of followers on social media need to be careful with what they put in the public domain. However, I feel that it is a little unfair when people external to the event criticize those caught up in it for spreading inaccurate information, as they may have the hindsight that those within in it do not. It is also important to bear in mind that for rumours to spread in such incidents, they still have to be credible to be believed and acted upon by crowd members. One could say in the current context of fear of terrorist attacks in London post Westminster/ London Bridge, Parsons Green etc, that rumours of another attack could be considered a credible threat, especially since the authorities took the incident seriously enough to dispatch armed police to the scene. However, other less credible rumours (e.g. that Godzilla was charging up Oxford Street!) may not have been so readily believed. The crucial thing is for the authorities to be as open & honest as possible with the information they provide, so that people don't feel such a need to seek information about the incident from other sources. A common myth about emergencies is that if people become aware of a possible threat, then they will be come too fearful to act rationally and will therefore 'panic'. However, there is almost no evidence to support this idea, and withholding information in emergencies could even result in people delaying action to keep themselves safe, and so could ironically increase the danger.   

Armed police on Oxford Street

Finally, even when crowd flight does occur, during such incidents, it's wrong to describe them as a 'stampede' because this implies unthinking animalistic behaviour. The research that I have done on crowd flight (Cocking, 2013) has found that people still tend to behave cooperatively when they flee (e.g. picking up others that fall over, waiting for friends/loved ones etc). This seems to be supported by footage of people fleeing the incident, as they don't appear to be pushing others as they get away, and many are not even running at full speed, but walking quickly, pausing and/or turning round to see what is going on etc. While there do appear to have been some casualties (16 hurt and 2 hospitalized) during the evacuation, given the likely number of people that would have been in the vicinity at 16.30 on a Friday four weeks before Christmas, this would appear to be a relatively small number, and injuries in crowd flight are usually because of crowd density or people accidentally falling over, not because they are deliberately pushing others over. 

So, to conclude, what we saw in Oxford Street this Friday, was not 'panic', and describing it as such is not only inaccurate, but also hinders detailed exploration of how we can safely manage such incidents in future. 

Media appearances by myself & John Drury:

Radio London Drivetime show 24/11 (interview from 2.48.00)
BBC news channel 25/11 at 12.10 (clip can be accessed via DropBox)
John Drury on LBC, 16.00 25/11
Radio5 Live, 25/11. (interview from 05.50)

References:
Cocking C. (2013) Crowd flight during collective disorder- a momentary lapse of reason? Journal of Investigative Psychology & Offender Profiling,10 (2) p.219-36. DOI:10.1002/jip.1389

Drury, J. and Cocking, C. (2007). The mass psychology of disasters and emergency evacuations: A research report and implications for practice. http://www.sussex.ac.uk/affiliates/panic/Disasters%20and%20emergency%20evacuations%20(2007).pdf 

Drury, J., Cocking, C., & Reicher, S. (2009a). Every one for themselves? Understanding how crowd solidarity can arise in an emergency: An interview study of disaster survivors. British Journal of Social Psychology 48.

Drury, J., Cocking, C., Reicher, S. (2009b). The nature of collective ‘resilience’: Survivor reactions to the July 7th (2005) London bombings. International Journal of Mass Emergencies and Disasters 27 (1) 66-95.

Parsons Green Tube attack

London saw another apparent terrorist attack on Friday 15th September- this time at Parsons Green Tube station in South West London, with 29 people injured in what must have been a horrifying experience for those affected. Mercifully though, it seems that the device failed to detonate fully, and so the casualties could have been much higher, with many fatalities had it exploded as intended, (given that the bomb went off on a packed tube train that can hold nearly 900 passengers during rush hour). The timing of the explosion was also fortunate, because it happened while the train was still above ground, and had it detonated fully while in a tunnel, the force of the explosion on the train and its occupants would have been much greater. So, while I wouldn't want to downplay the shock and terror that those on the train may have experienced, it looks like this incident could have been much worse.

'Stampede' or logical flight?
As with such incidents there has been blanket media coverage (including an interview I did for the BBC news channel on Friday evening), with terms such as 'panic' and 'stampede' being used liberally in both media and eye-witness accounts. For instance, BBC coverage describes a 'stampede' as people evacuated down the stairs. Those familiar with this blog will be aware that those of us who work in the field of emergency planning and response are critical of such terms, as they don't usually match up with detailed examination of what actually happens, but also because they imply that people in crowds behave selfishly and/or & irrationally. In the interview I did for BBC news, I made the point that it's not 'panic' to flee a potentially life-threatening risk, and in the current context (that this is the 5th major incident of its kind in Britain this year), it is not surprising that people rapidly fled as soon as it became apparent that there was an explosive device on the train. Previous research I have done (Cocking, 2013) has found that while instinctive crowd flight can happen (e.g. people run as soon as they see a crowd surging towards them), it is usually brief and people still co-operate with others during such flight, so the idea that people who fall over are then deliberately trampled by others as they flee is largely a myth.
However, there are also specific physical aspects of the incident at Parsons Green that need further exploration, as otherwise it is easy to  slip into using irrationalist narratives if one does not examine why certain things happened.  So, interviews with eyewitnesses include reports of people falling over each other as they evacuated down the stairs off the platform at Parsons Green, and it is possible that some injuries occurred this way (although we do not yet know how many and most of the 29 casualties appear to have been burns from the explosion). However, there are descriptions of people helping each other within the crush, and also mobile phone footage that people took of the bomb on train, so not everybody fled, and some people stayed on the platform. What happened as people ran down the stairs seems to be a crowd collapse, whereby a domino effect can happen in densely packed crowds (eg if someone falls over, then the physical pressure on those behind them means that they are forced over into the space created). This is what happened in the Bethnal Green tube station tragedy in 1943 (when 173 people died during a crowd crush as they ran down the station steps to escape an air raid), because someone fell over and others tumbled over them, creating a crowd collapse.

The incident on Friday could also have been exacerbated by some specific physical factors relating to where it happened. So, for instance, Parsons Green is quite an old Victorian station on the District Line (it was built in 1880 before the deeper tunnels of the London Underground were created), and so many of its stations are above ground (see photo below). Therefore, people had to evacuate downwards to get out of the station (rather than upwards, had they been at a station in a tunnel), and the effects of anyone falling over in such a downward fleeing crowd would have been made worse by the effect of gravity. Secondly, it seems from reports that the device went off towards the rear of the train, and the exit at Parsons Green is towards the front of the train. So when people evacuated, most would have probably surged towards this available exit at the end of the platform, meaning the crowd surge was denser than had there been more than one obvious way to evacuate (it seems that some people evacuated along the tracks, but much less than those that fled towards the exit at the end of the platform).

Tube train above ground at Parsons Green station

Conclusion:
Previous work I have done with John Drury into mass emergencies, including 7/7 (see references below), found that people often behave much more resiliently than is expected of them, and I think yesterday's incident at Parsons Green was no exception. Yes, there were factors during the crowd evacuation that we need to consider if we are going to help ensure that future incidents are managed safely. However, I would argue that these are usually physical and/or logistical aspects of the space in which the evacuation happens (rather than any inherent problems with the way that those affected behave en masse), and using irrationalist terms such as 'panic' or 'stampede' do not help us advance our crowd safety management approaches. Finally, it is also worth bearing in the mind the general sense of community spirit, spontaneous acts of co-operation and generosity shown by others that happens after such incidents. For instance, a local resident in Parsons Green offered to put the kettle on for anyone affected, and a local Pizza company gave out free pizza and water to members of the emergency services on the scene (see photo). I think these are good examples of how adverse events can help bring communities together, and not divide them (as is often the intention behind the perpetrators), and we should remember this positive message in the aftermath of such incidents.

Stall set up by local Pizza company to support emergency services

References:
Cocking C. (2013) Crowd flight during collective disorder- a momentary lapse of reason? Journal of Investigative Psychology & Offender Profiling,10 (2) p.219-36. DOI:10.1002/jip.1389Drury, J. and Cocking, C. (2007). The mass psychology of disasters and emergency evacuations: A research report and implications for practice.http://www.sussex.ac.uk/affiliates/panic/Disasters%20and%20emergency%20evacuations%20(2007).pdf 

Drury, J., Cocking, C., & Reicher, S. (2009a). Every one for themselves? Understanding how crowd solidarity can arise in an emergency: An interview study of disaster survivors. British Journal of Social Psychology 48.

Drury, J., Cocking, C., Reicher, S. (2009b). The nature of collective ‘resilience’: Survivor reactions to the July 7th (2005) London bombings. International Journal of Mass Emergencies and Disasters 27 (1) 66-95.

Notting Hill Carnival & collective support

I was at the Notting Hill Carnival today, which was held this year in the shadow of the Grenfell Tower fire tragedy in June 2017. For those unfamiliar with the geography of Notting Hill, this is not mere hyperbole, and from where I watched the carnival go by on Ladbroke Grove, the gutted Tower block was clearly visible to carnival goers. This made it a more sombre and reflective event than previous carnivals I have experienced (I grew up in West London & so have been a frequent visitor to the Carnival over the years), but from what I saw, the collective support offered by the local community and carnival goers shows how people can come together to support each other after such tragedies.

Locals had asked that visitors to the carnival didn't take photographs of the Tower and to respect people's privacy around the estate where Grenfell Tower is. Therefore, my focus in this post is on examples of collective support afterwards, rather than the tragedy itself & I have copied below some of my photos that I think are good examples of such collective support. In a previous post I looked at how the local community came together to support each other after the Shoreham air disaster in 2015, and how a nearby bridge became a focus of messages of support. A similar thing has happened in the area close to Grenfell Tower (I also noticed this on a previous visit to a local church in July), and a popular image of  Grenfell as a tube logo has also emerged,with people wearing T-shirts with this logo in solidarity at today's carnival.

Grenfell tube logo 

Messages of support on Ladbroke Grove 

I also noticed some interesting dynamics in the crowd behaviour at the carnival. For instance, the area closest to Grenfell Tower was designated as a quiet zone for reflection, and at 15.00 on both carnival days, there was a planned minute's silence. I did wonder beforehand how easy it would be to organise a minute's silence in such a large carnival crowd, but in the area where I was standing, word went around from about 14.50 onward that there was going to be a minute's silence. The crowd pretty much universally respected this & fell silent along with the emergency services personnel who were stationed there. The floats also fell silent when they passed the quite zone, and some had messages of support along the side The few people that didn't realize what was going on were told by others to be quiet & quickly did so. Once the minute was over, there was a spontaneous round of applause and yellow balloons were set off from local flats. This was quite a powerful experience to be part of and I felt this was a good example of the potential strength of unity that crowds can have in showing positive emotion & collective solidarity. This all fits with previous work I have done with John Drury on community resilience after disasters, where we found that people can come together and support each other much better than is often expected of them (see references below).

Quiet zone on Ladbroke Grove

Emergency services observing 1 minute silence

Conclusion:
The Grenfell story is far from over and the survivors & victim's families will need ongoing material and emotional support in their fight for answers and justice (the Red Cross are taking donations & support for those affected can be accessed here). I also wouldn't want to make any claims that such a journey will be easy or without setbacks. However, I hope that the examples of collective and community support that I saw at the Notting Hill carnival today, could help form a vital part of the collective healing process that is so desperately needed in this area of West London.

Messages of solidarity on a float

References:

Cocking, C (2013) Collective resilience versus collective vulnerability after disasters- a Social Psychological perspective. In R. Arora (ed.) Disaster Management: A Medical Perspective. CABI: Oxford, UK.http://bookshop.cabi.org/?page=2633&pid=2433&site=191

Cocking, C (2016) ‘Collective Resilience and social support in the face of adversity- evidence from Social Psychology’ in Kumar, U (ed.) 'Routledge International Handbook of Psychosocial Resilience'. Routledge, Taylor & Francis: UK.

Drury, J. (2012). Collective resilience in mass emergencies and disasters: a social identity model. In: Jetten, J., Haslam, C. and Haslam, S.A. (eds) The Social Cure: Identity, Health and Well-being. Psychology Press, Hove, UK.

London Bridge attack & crowd responses

Britain is waking up to the news that London has suffered another attack that has so far seen 7 people killed and nearly 50 hospitalized (a casualty hotline has been set up for those worried about loved ones- 0800 096 1233). Shortly after 22.00 on Saturday 3rd June, a van mounted the pavement on London Bridge, and began mowing down pedestrians. Then, three men got out and began attacking people with knives in the nearby Borough market on the Southern end of London Bridge until they were shot by armed Police. This was clearly a horrific attack, and my thoughts are with all those affected. I am familiar with the area, (it is a popular place for commuters to stop for a drink before travelling onwards from London Bridge railway station), and the thought of innocent bystanders being targeted at random while they were enjoying a Saturday night out, is particularly worrying. However, as horrific as this attack surely is, there is also evidence that shows how events such as these can see a co-operative spirit emerge.

Crowd flight?
This low-tech attack appears to be the third of its kind seen in London in recent years (the first being the murder of Lee Rigby in 2013, and in March 2017 the Westminster Bridge attack), where a vehicle is used to attack pedestrians. While such incidents are truly terrifying for those affected, closer inspection of events usually shows that describing them as 'mass panic' rarely matches with the available evidence. So, for instance, footage of people leaving the scene last night shows an orderly evacuation, where they are evacuating with a degree of haste, but this is far from headlong flight (some people are not even running), and I can see no evidence of selfish behaviour (such as pushing others) or people falling over in the rush. This fits with crowd responses to the truck attack in Nice in July 2016, and with research that I have done (Cocking 2013a) into crowd flight.  

Bystander responses:
Reports are coming in of the speed and efficiency of the response by the emergency services, which by all accounts was amazingly quick. For instance, the London Ambulance Service reports that they were on the scene within 6 minutes, and deployed over 80 medics to help the injured. The attackers were also shot dead within 8 minutes of the start of the incident. However, as with such events, there is always a delay between the incident beginning and the emergency services deploying on the scene (no response can ever be instantaneous), and so in the minutes (or even seconds) before they arrive, we often see heroic acts by bystanders as well. So for instance, in an interview with an eye-witness, he describes his efforts to warn people what was happening and how he also intervened by throwing things at the attackers. I have also heard reports on Radio5 Live that people were directing Police to where the attackers were in Borough market, and an eye-witness talked about how people reacted; 'everyone seemed to adapt to it very quickly and respond to it as it happened'. On a broader level, in the aftermath of the attack, the hashtag  #sofaforLondon appeared on social media, as Londoners began offering a space to stay for anyone unable to get home after the attack. People are also using the Facebook safety check service to reassure others that they are OK.

I believe that all these examples illustrate well the concept of collective resilience, something that emerged from research I did with John Drury (see references below) into mass emergencies, whereby spontaneous co-operation quickly emerges among those affected, and the idea that people will be too shocked or 'panicked' to help each other just isn't backed up by what happens on the ground. So, bystanders in emergencies could actually be seen as a form of 'zero-responders' (Cocking 2013b) that could help strengthen the official response to such incidents. I hope that this can help show that a positive aspect of our shared humanity can emerge from such awful attacks, as people come together to help others in times of need.

Responders tend to the injured 

References:

Cocking C. (2013a). Crowd flight during collective disorder- a momentary lapse of reason? Journal of Investigative Psychology & Offender Profiling. 10 (2) p.219-36.

Cocking, C. (2013b). The role of "zero-responders" during 7/7: Implications for the Emergency Services. International Journal of the Emergency Services, 2 (2) 79-93.

Drury, J. and Cocking, C. (2007). The mass psychology of disasters and emergency evacuations: A research report and implications for practice. 
http://www.sussex.ac.uk/affiliates/panic/Disasters%20and%20emergency%20evacuations%20(2007).pdf 

Drury, J., Cocking, C., & Reicher, S. (2009a). Every one for themselves? Understanding how crowd solidarity can arise in an emergency: An interview study of disaster survivors. British Journal of Social Psychology 48.

Drury, J., Cocking, C., Reicher, S. (2009b). The nature of collective ‘resilience’: Survivor reactions to the July 7th (2005) London bombings. International Journal of Mass Emergencies and Disasters 27 (1) 66-95.

Manchester attack and crowd resilience

The city of Manchester and the whole of the UK are reeling in shock in the aftermath of the  murderous attack on the Manchester Arena after the Ariana Grande concert on 22/4/17, with 22 confirmed dead and 59 injured. As I write this, the UK terrorist threat level has been raised to its highest level, 'critical' (meaning another attack may be imminent), and soldiers have been ordered to support the Police on the streets to free up more armed officers. It is difficult to find the words to describe such a horrific attack that was clearly designed to be as shocking as possible and divide communities in its wake. I will show in this blog how I feel that such dark events can bring people closer together, both during the incident and in its aftermath. As is common with media coverage of such incidents, the words 'panic' and 'stampede' are freely used to describe the crowd response, suggesting that the people affected behave irrationally. However, work that myself and colleagues have done in this area has found that the reality is often more complex, with people behaving in ordered ways that are governed by the context in which they find themselves.

'Stampedes?'
From what I have seen of crowd responses in the Manchester Arena, I think it is problematic to simply describe people's behaviour as a 'stampede' (something I have looked at in previous blogs on crowd flight).  It appears that as the concert was finishing, a lone suicide bomber walked into the foyer and detonated his device. The initial response to the blast appears to have been an eerie silence, then people began screaming once they realised what had happened, and then they fled rapidly from the venue. Mobile phone footage from inside the venue and the train station shows the crowd flight after the blast, and while you can see a degree of urgency amongst people leaving, there's no evidence of people behaving selfishly. Research I have done into crowd flight (Cocking, 2013) has shown that it is misleading to describe such incidents as a 'stampede' because it implies that people are behaving like animals with no consideration for their fellow human beings. Instead, people tend to help each other out when they are able, and I have not yet seen any footage of people behaving selfishly (eg pushing people or trampling over others). I'm mindful that it is possible that fear in a crowd composed largely of young people could be greater than in a crowd of adults (which is where I have done most of my research), but I have not yet seen any evidence that any such increased fear significantly increases incidences of selfish behaviour.

'Panic'?
'Panic' is another word that is commonly used to describe such incidents, both by the media and in eye-witness accounts of the incident. Now, I'm not claiming that this incident was anything other than utterly horrific and terrifying for those involved, especially because there were many young people and children at the venue (and for some it was possibly their first gig). However, to describe such events as 'panic' doesn't fully explain the full complexity of what goes on, and like the term 'stampede' it implies that people are behaving irrationally and/or selfishly, when the evidence suggests otherwise. So, within the incident, I heard reports of people grabbing their relatives and running, forming human chains to help wheelchair users, and worried parents going against the crowd flow to enter the venue to find their children. There is also a moving interview on Channel 4 news where two parents describe how they looked after an injured child before they had found their own children. I would say that all of these examples are the opposite of 'panicked' behaviour, and instead show how people co-operate with each other (often to help complete strangers) in emergencies.  

Bystanders help the injured 

Collective Solidarity:
Much has been made of how people have come together in the aftermath of this tragedy, and I have heard journalists and politicians refer to the 'spirit of Manchester' and how such a tragedy will bring people together.  This fits with the the work I have done with John Drury that shows a sense of collective resilience can emerge from such incidents. So, for instance, I saw reports of people giving out water to victims, queuing up to donate blood, taxi drivers not charging for fares, people organizing lifts to get fans back to Liverpool, and one bystander leading 50 children to safety and arranging for a local hotel to put up those who couldn't get home. This collective solidarity also continued after the immediate attack, with a vigil in Manchester's Albert Square the day after the incident, and a planned rally by the far-right English Defence League (EDL) in the Arndale Centre was shouted down by Mancuniams. Now I accept that there could be a degree of political and media rhetoric here, (as a clear aim of such terrorist attacks is to divide communities), but the numerous examples of people spontaneously showing such collective solidarity, leads me to conclude that this is a real phenomenon above and beyond any journalistic hyperbole.

Mancunians confront the EDL

References:
Cocking C. (2013) Crowd flight during collective disorder- a momentary lapse of reason? Journal of Investigative Psychology & Offender Profiling,10 (2) p.219-36. DOI:10.1002/jip.1389


Cocking C (2016) Brussels terror attack victims show how humans help each other in times of crisis. Published online in The Conversation, 22/3/16. https://theconversation.com/brussels-terror-attack-victims-show-how-humans-help-each-other-in-times-of-crisis-56707

Cocking, C. & Drury, J. (2014) Talking about Hillsborough: ‘Panic’ as discourse in survivors’ accounts of the 1989 football stadium disaster. Journal of Community and Applied Social Psychology, 24 (2) 86-99. DOI: 10.1002/casp.2153;http://onlinelibrary.wiley.com/doi/10.1002/casp.2153/abstract

Drury, J. and Cocking, C. (2007). The mass psychology of disasters and emergency evacuations: A research report and implications for practice.http://www.sussex.ac.uk/affiliates/panic/Disasters%20and%20emergency%20evacuations%20(2007).pdf 

Drury, J., Cocking, C., & Reicher, S. (2009a). Every one for themselves? Understanding how crowd solidarity can arise in an emergency: An interview study of disaster survivors. British Journal of Social Psychology 48.

Drury, J., Cocking, C., Reicher, S. (2009b). The nature of collective ‘resilience’: Survivor reactions to the July 7th (2005) London bombings. International Journal of Mass Emergencies and Disasters 27 (1) 66-95.