DON T PANIC Crowd behaviour in emergencies implications for

DON’T PANIC! Crowd behaviour in emergencies: implications for professionals Presentation for the EPC 3/9/2008 Dr Chris Cocking London Metropolitan University c. cocking@londonmet. ac. uk

Outline of Presentation n 1) Background aims of research 2) Examples of how behaviour in emergencies support our theories 3) Implications for emergency planners

Context of research n n Crowd behaviour theories have developed over timea) 19 th Century: Le Bon irrationalist approach b) 1960 s - 70 s: more rationalist approaches c) 1980 s to present: The Social Identity Approach- not rational or irrational, but transformation from individual to collective

Iroquois fire 1903 n ‘most of the dead were trampled or smothered, though many jumped or fell to the floor of the foyer. In places on the stairways, particularly where a turn caused a jam, bodies were piled 7 or 8 feet deep. Fireman and Police confronted a sickening task in disentangling them. [ ] The heel prints on the dead faces mutely testified to the cruel fact that human animals stricken by terror are as mad and ruthless as stampeding cattle’ (in Latane & Darley 1970) p. 53

What is panic? n n Difficult to define Is escaping a potentially fatal threat- panic or logical flight behaviour? ‘Collective flight based on hysterical belief’ (Smelser, 1963) ‘extreme and groundless fear’ (Quarantelli, 2001) -but debates whether it’s even a useful concept to describe behaviour in emergencies

The ‘Panic’ model n n n Part of the irrationalist tradition in crowd psychology a) Threat causes emotion to overwhelm reason b) Collective identity breaks down c) Selfish behaviours- pushing, trampling d) Contagion- these behaviours spread uncritically to crowd as a whole

Implications of the panic model n Assuming people will panic in disasters can affect emergency planning n Implications behind Civil Contingencies Act of potential threat of public gatherings (Drury, 2004) n n Counter-terrorism planning in US tends not to trust public to behave sensibly, assuming that they are uncooperative and prone to panic (Glass & Schoch. Spana, 2002) Over-protective government responses may stunt public’s own natural resilience (Wessely, 2005)

Problems with the panic model n n Mass panic is rare - noticeable by its absence in many different emergencies, such as; a) Atomic bombing of Japan during World War II b) Kings Cross underground fire (1987) c) WTC evacuation 9/11 ‘classic panic action or people behaving in an irrational manner was noted in [just] 1/124 (0. 8%) cases’ (Blake et al. 2004)

Panic during WWII? n n Not simply number of casualties, but intensity and unpredictability of bombings 80, 000 killed in London during Blitz, vs. 3000 in Coventry, but more psychological casualties in the latter Children often found it more upsetting to be separated from parents than air-raid itself Cf Jones et al (2006), Mawson (2005)

Under-reaction rather than over-reaction ‘When people die in fires, it’s not because of panic, it’s more likely to be the lack of panic’ Neil Townsend, Divisional Officer, London Fire Rescue Service

Social attachment model n n n Developed by Mawson (2005) Uses early psychological theories of maternal attachment In times of stress, people seek out attachment figures- known as affiliative behaviour Social norms rarely break down But these ties can have fatal consequencespeople tend to leave or die as a group

Social attachment model n Improves on panic model, and supported by evidence from disasters (Cornwell, 2001) n n n But problems remain; a) Implies that panic in a crowd of strangers is more likely b) Neglects possibility that strangers may co-operate with each other

The self-categorisation approach n n Disasters can create a common identity or sense of ‘we-ness’ This common identity results in orderly, altruistic behaviour as people escape common threat Increased threat can enhance common identity Supported by evidence from sociologists Clarke (2002); Tierney (2002)

Recent research n n n Project funded by ESRC from 2004 -7 at University of Sussex Can existing psychological models of crowd behaviour can be applied to emergencies? 3 different areas of research: interviews, room evacuations, and VR simulations

Interviewing disaster survivors Sinking of the Jupiter, 1988 & Oceana, 1991 Hillsborough football stadium disaster, 1989 Ghana football stadium ‘stampede’, 2001 Bradford football stadium fire, 1985 Fatboy Slim beach party, 2002 Harrods bomb, 1983 Canary Wharf evacuation, 2001

Results from interviews Common identity quickly emerges n Co-operative rather than selfish behaviour predominates n

Hillsborough 1989

I don’t think people did lose control of their emotions [ ] they were clearly in control of their own emotions and their own physical insecurity, I mean [] you’re being crushed, you’re beginning to fear for your own personal safety, and yet they were [ ] controlling or tempering their emotions to help try and remedy the situation and help others who were clearly struggling

‘Fat Boy Slim’ Brighton 2002

n n ‘People were helping people up and helping people down it was a very different atmosphere from any other gig that I’d ever worked before’ ‘It was like a massive rave party where everybody felt they knew each other where they could go up to each other hug total strangers and they were in such close proximity to each other and all you could see was people sticking their arms round each other and grinning and you know it was oh God it’s a bit packed isn’t it that sort of…those conversations were going on but not complaints about it’

Room evacuation studies n n n Simulated role-plays of room evacuations with smoke and time pressures Some evidence of common identity emerging in response to shared fate But study suffered from lack of realism

VR evacuation programme n n Joint project with computing scientists at Universities of Nottingham & RMIT (Australia) Many good simulations of crowd flow, but most don’t consider psychological theories of crowd behaviour Evidence for link between sense of groupness and helping Discussions with potential users (e. g. Home Office/Sci. Tech) to market it as a training tool

Research into 7 th July, 2005 Press reports and web-logs n Web based questionnaire for eyewitnesses of bombings n Interview study of survivors n Results support our theories n

Chronology of events on the tube on 7/7 n n Blasts followed by darkness and silence Screaming- people try to work out what’s going on Smoke & soot clear- attempts to help/ comfort others, & escape- some delay because of fear that tracks are live Passengers wait approx 30 mins. for rescue, and walk in orderly fashion along tracks when directed

Response to 7/7 Individual fear and distress, but no mass panic n Evacuations characterised by orderly, calm behaviour n Many reports of altruism, cooperation, and collective spirit of Londoners/ UK as a whole n

Orderly evacuation

The myth of Panic n n Many accounts of ‘panic’ But what actually is panic, and what is logical flight behaviour? Need to look at what people actually do, and decide if it is indeed ‘panic’ More than just semantics, as it could affect emergency evacuation planning

Panic? n n ‘There was no real panic - just an overwhelming sense to get out of the station quickly’ ‘Almost straight away our packed carriage started to fill with smoke, and people panicked immediately. Thankfully there were some levelheaded people on the carriage who managed to calm everyone down’

Unity n n ‘I felt there was a real sense of unity. We were all trying our best to find a way out of there and reassure each other’ ‘One of the things which struck me about this experience is that one minute you are standing around strangers and the next minute they become the closest and most important people in your life. That feeling was quite extraordinary’

Co-operative behaviour n n n ‘Many people kept calm and tried to help one another to see if anyone was injured’ ‘I was very aware of people helping each other out and I was being helped myself’ ‘Passengers with medical experience were found, I found a tool box and we smashed a window, allowing the medical guys to enter the other train’

Wider contexts n n n Do different nationalities/ cultures respond differently in emergencies? Common belief that if panic doesn’t happen, it’s due to a specific national identity (e. g. ‘British Bulldog spirit’), and panic may happen elsewhere, but little evidence to support this We expected some minor cultural variations at start of project, but the more we looked, the less differences we found

Panic on 9/11?

Asian Tsunami, Thailand 2004 n n When tsunami hit, divisions between local Thais and Western holidaymakers were forgotten and people co-operated Reports of fighting between tourists to get ferries from islands afterwards, but collective identity could have diminished by then and less important than desire to get home

Hurricane Katrina, Sept 2005 n n n Initial reports of mass looting, gang-rapes, and murders in Superdome, New Orleans But these reports were later seen to be wildly exaggerated- very little evidence to support them: crime rate in period after Katrina actually dropped Local Police chief resigned when scale of exaggeration became clear

Fear of mob has fatal results?

Hajj pilgrimage, Saudi Arabia n n Some of largest crowds on earth travel to Mecca/ Medina for the Hajj each year Some tragic accidents, but overwhelming majority of pilgrims unaffected Fatalities usually due to physical pressure of crowd rather than any ‘irrational’ behaviour Need to overcome some fatalistic cultural beliefs (e. g. ‘it’s God’s will’, & ‘you go straight to heaven if you die on the pilgrimage)

Different behaviour depending on perception of threat? n n Some evidence that people may be more fearful/anxious if threat is man-made (e. g. terrorist attack vs ‘act of God’) But no evidence that this results in more overt ‘panicked’ behaviour. Seems to be largely that people fear that others may panic if they realise threat is manmade and so don’t say so even if they believe it is But common identity appears to emerge in both situations- no evidence yet of any

General conclusions n n Little evidence for mass panic in emergencies The ‘panic’ model should not be used in planning emergency responses Any selfish behaviour is confined to individuals and rarely spreads Risks associated with crowds are usually due to physical constraints and lack of info rather than their inherent ‘selfishness’

OK, BUT SO WHAT? Possible applications of the research

More info rather than less n n Very little evidence that people panic if made aware of a threat and some shows the opposite (Proulx & Sime 1991) Use of radio code words (e. g. Mr Sands etc) good for keeping professional composure, but no evidence people stampede if they hear ‘FIRE!’ If info is given in clear ways that people can safely act upon to escape threat, they usually do Deliberately withholding info could cause problems in any future emergencies, as people may not trust accuracy of messages

Delivery of information is important n n Info needs to be clear, unambiguous, delivered confidently, and come from believable source that crowd identifies with (e. g. recent advice to RUSI re CBRN incidents) Spokesperson should be appointed with sole duty to communicate with public (London Assembly 2006) n This could depend on type of crowd; e. g. commuters, football fans

Hurricane Gustav ‘You need to be scared, you need to be concerned, you need to get your butts moving out of New Orleans now! … We are ordering a mandatory evacuation of the city of New Orleans starting in the morning at 8 am on the West Bank… we give you 4 hours to evacuate’ (Ray Negin, Mayor of New Orleans. 31 st August 2008)

Crowds can be part of the solution rather than part of the problem People may delay own evac to help others Appeal to the crowds’ common humanity- ‘We’re in this together’ Don’t address commuters as atomised ‘customers’ Influential leader figures may emerge from crowd, who can help rescue effort

Spontaneous leaders on 7/7

Plan emergency response n n n Take the possibility of emergency seriously- don’t think ‘it won’t happen to us!’ Train staff in knowledge of location, and how to relay information effectively in emergencies Don’t say ‘don’t panic’, as it can create expectation of panic (Durodié & Wessely, 2005) Practice can make a real difference; e. g. WTC evac rates- 1993 vs 2001 (99% of those below planes’ impact escaped-USA Today)

Possible problematic behaviours by individuals n n n Delaying exit to safety/ finishing mundane tasks-’freezing’ or ‘disassociation’ People tend to leave by route they entered, even if closer exits are available Crowd members can be unaware of physical pressure that others may suffer People unwilling to leave area, or passers-by rubber-necking Attempts to breach cordons (worried parents, single-minded commuters, etc)

But… n n We don’t make any claims that everyone behaves like heroes Some individuals may behave in selfish or even ‘panicked’ ways But this is isolated and almost never spreads Others usually intervene to calm situation

Summary n n n Crowds in emergencies behave in ways that are consistent with their social identities and governed by the social norms of the situation The ‘panic model’ is largely a myth Evidence gathered from many different emergencies supports our theories

Future plans n n n We’d like to look in more detail at concept of ‘resilience’ Lot of research on trauma after emergencies- less on how we cope (only 25 -33% of disaster survivors get PTSD) Can the common identity that emerges from disasters help shield people from PTSD? Open question, but area needs more research

Thank you for listening n Any questions/ comments? Full Report available at: http: //www. sussex. ac. uk/affiliates/ panic/applications. html n Or e-mail me; n c. cocking@londonmet. ac. uk n

References: n n n n Blake et al (2004). Proceedings of Third International Symposium on Human Behaviour in Fire. Clarke L (2002). Contexts. Cornwell B (2001) The Sociological Quarterly, 44, 617 -638. Drury J (2004) The Psychologist. Durodié B & Wessely S (2002) The Lancet. Glass T & Schoch-Spana M (2002) Clinical Infectious Diseases Jones et al (2006) Journal of Risk Research Latane & Darley (1970) The Unresponsive bystander. Le Bon G (1968)The crowd: A study of the popular mind London Assembly (2006) Report of the 7 th July Review committee. Mawson A (2005) Psychiatry Proulx G & Sime J (1991) Fire Safety Science Quarantelli E (2001) The Sociology of Panic Smelser N (1962). Theory of Collective Behaviour. Tierney K (2002) Strength of a city: A disaster research perspective on the WTC attack.