'Lack of information can turn a passive crowd into a stampede'
 

(updated 3rd Sept 2004)

An expert on how humans act in large groups, mathematician and consultant Dr. G. Keith Still, 47, is a regular visiting speaker at the government's Emergency Planning College at Easingwold, Yorkshire

Tuesday February 18, 2003

[Extracted in the Guardian of February 28 2003]

There is a scene in the movie "Marathon Man" where Laurence Olivier plays a sadistic dentist to a reluctant Dustin Hoffman, who is asked over and over by his tormenter: "Is it safe?" Several new cavities appear and they are no closer to an answer. The script also involves a lot of people running around in the middle of the night in their pyjamas. It's a good analogy to the situation emergency planners face on a daily basis. Although there are many reports of the "latest potential terrorist threat" we seldom read about the civil contingencies and the emergency planning departments around the country. Their work doesn't make front-page headlines until something goes horribly wrong. Yet every large planned event requires a team of local authorities, safety engineers and health inspectors all testing contingencies before the event is granted a licence. The police are consulted, security plans are evaluated and the only considerations are "Is it safe?"

Having worked with many international event planning teams, I can testify that when things go right it "was expected", while when things go wrong the emergency planners take the blame. It is a thankless task. To put emergency planning in perspective, here is an eye witness account from a London Underground passenger during January's derailment of part of a Central Line train.

"Just before Chancery Lane station we heard a scary loud sound and the whole compartment started wobbling and lights went off. There were shattering noises and it appeared that the entire train has lost its balance and is about to crash. There were screams and panic and the compartment was full of smoke or dust. Soon after, we were at the platform, but the doors wouldn't open. We were apprehensive that in case there is fire we wouldn't be able to get out. A young man smashed a window of the door and we could get out through it. There was bit of stampeding to get out of the station as an announcement was to leave the station as soon as possible. Unfortunately one of the escalators was not functioning. We climbed steps which was an ordeal. We breathed fresh air as we came out. You need the hand of god over you to undertake any travel these days. The whole incident might have been 30 minutes or more but it was like a nightmare." 

The eye witness, Dr Bhatnagar, is a family friend of a colleague of mine, Seema King, who is business manager of our consultancy company, Crowd Dynamics. Also in January, another colleague, this time from the Emergency Planning College at Easingwold in North Yorkshire, was in the control room during the police operation to detain a suspected terrorist at a flat in Manchester - the raid where Stephen Oake, a special branch officer, was stabbed to death. In October last year, a cousin of our systems engineer was killed in the Bali bombing. And I've stood at the edge of "ground zero" in New York a few weeks after the terrorist attack on the Twin Towers there in September of 2001. International terrorism touches our lives in many ways, but little is spoken of the activities of emergency planning and civil contingencies.

Panic is rare
Contingency planning for London Underground is a formidable task. There are over 408 kms of railway, 149 kms of deep tube, 32 kms of cut and cover, 273 stations, 408 escalators, 112 lifts, 507 trains in peak service and some 3,987 carriages in the fleet. Each day there are 3,100,000 passenger journeys. Each year that comes to approximately 927,490,000 passenger journeys and 6,716,000,000 passenger kilometres. Passenger safety is the only concern of the emergency planning manager. Planning for the Millennium Eve celebration took 18 months and over £10m plus project costs. It passed without incident - as I said, emergency planning is a thankless task. No one is complacent: January's underground accident at Chancery Lane should have been avoided, but with such a demand on the infrastructure, beyond its designed capacity, holes in the system will appear. Emergency planning takes many forms. Human factors are an important element in understanding how people will react to situations. The late Jonathan Sime was a Surrey-based environmental psychologist who made considerable contributions to this field. To quote from a paper of his:

"The most important finding of the research is the fact that the start-up time (ie, people's reaction to an alarm) is as (if not more) important as the time it takes physically to reach an exit . . . On average two-thirds of the time from onset of the alarm to reaching an exit was spent by people not moving at all. On average one-third of the time was spent in moving from seat positions to an exit."

In other words, people generally seek further information before they act. The majority of events unfold in a relatively predictable manner: panic is seldom a factor in emergency egress (evacuation) - unless the public lack information, are confined, threatened or confused. Lack of information can turn a passive crowd into a stampede. The consequences of a crowd crush can be much worse that the incident. Of course, whenever large crowds gather and move there is inherent risk to public safety. In east London on March 3 1943, Britain's worst civilian disaster saw 173 men, women and children lose their lives as people tried to get into an air-raid shelter at Bethnal Green underground station. Rocket fire near by frightened the crowd; a woman with a baby tripped on the staircase; in the push, others fell over her. It was crowd dynamics that produced the high death toll.

Crushing toll
Over the years, we've all read about or seen many examples of these dynamics at work among people attending sports matches and many other mass events. Here are just some of the examples:

1964 Lima: 300 die in a stampede after goal disallowed, Olympic qualifying match.
1968 Buenos Aires: 74 die when a crowd stampeded after burning paper was thrown on to football terraces. Fans head towards a closed exit and are crushed against the doors.
1971 Glasgow: 66 die when barriers in Ibrox stadium collapse during football match.
1974 Cairo: 49 trampled to death as crowds break barriers.
1979 Lagos: 24 die as Nigerian football fans stampede during a light failure.
1981 Piraeus: 24 die in a stampede as Greek fans rush to leave sports ground.
1981 Sheffield: 38 injured during a crowd surge at Hillsborough stadium.
1982 Moscow: 340 die at European Cup match when fans, leaving stadium, try to re-enter after last-minute goal.
1982 Cali, Columbia: 24 die when drunken sports fans provoke a stadium stampede.
1985 Mexico City: 10 die trying to force their way into a stadium.
1985 Brussels: 39 die at Heysel stadium when riots break out and a wall collapses.
1988 Kathmandu: 70 die in stampede towards locked sports-stadium exits in a hailstorm.
1989 Sheffield: 95 die, over 400 injured, in crowd surge when police open gates at Hillsborough football stadium to alleviate crowding.
1990 Mecca: 1,426 die in stampede in overcrowded pedestrian tunnel during the annual Haj pilgrimage.
1991 Orkney, South Africa: 40 die along fences when cricket fans try to escape fighting.
1993 Hong Kong: 21 die in crowd crush in the early hours of January 1 when thousands of party-goers were caught off-guard in narrow sloping streets in what became known as the Lan Kwai Fong disaster; poor police planning and bad weather played a part.
1994 Mecca: 270 pilgrims die in a stampede during "stoning the devil" ritual.
1996 Guatemala City: 83 die at a World Cup qualifying match.
1998 Mecca: 150 Muslim pilgrims crushed to death in a stampede.
1999 Kerala state, south India: 51 killed in a stampede after part of a shrine collapsed during Hindu ceremony attended by 1.5m.
1999 Minsk, Belarus: 53 die when a crowd of 2,500 rushed to get out of the rain at the railway station.

Just last year, in September, 21 pupils were killed, most suffocated under a pile of bodies, when a guard rail gave way in a dark stairwell at a Chinese school during a power blackout. There is also the phenomenon of "crowd crazing", where marketing or promotions prompt the public into actions of the sort seen in January last year a Japanese shopping mall. According to news reports, the Tama Plaza Tokyu shopping centre in Yokohama offered a new year promotion of "lucky bags" (fukubukuro) containing merchandise more valuable than their retail price. A large crowd turned out. Anxiety and competitiveness developed, and when the doors of the centre opened mid-morning, the shoppers surged forward. The craze caused a crowd collapse as people toppled over each other. Ten people were injured. This sort of event is more alarming to emergency planner as areas not designed, planned or managed are attracting large crowds in often dangerous, competitive environments. Nowadays, however, there is an added element: media highlighting of newer and seemingly more dangerous threats to public safety from terrorism is changing people's reaction time when something untoward happens. Where people before would generally seek further information before acting, there is now an added risk from crowd reaction to a normal public announcement ("Please leave the station as soon as possible").

Recall the Chancery Lane witness: "There was bit of stampeding to get out of station."

It is important to raise awareness of the potential of a terrorist threat - but there also does need to be more than a passing media mention to the work of the civil contingency planners around the world. Raising the anxiety of the general population can only exacerbate public reaction, and in some cases over-reaction, to a situation. This can be more life threatening in an emergency than the incident, as I've highlighted in the various disasters around the world in recent decades. Crowd behaviour can be changed with appropriate information - specifically what to do in an emergency situation - and we are working with a number of authorities around the world to provide an understanding of crowd dynamics to this aim. At all times the information needs to be simple; during an emergency you do not have time to look up a wide range of alternatives in a big book of words. So the design and management of places of public assembly needs to take into account the human factors for both normal and emergency situations. The contingency planning needs to take into account the various human factors and the analysis of the situation needs to consider how people react to emergencies in an ever changing risk profile presented by the more alarming terrorist threats. As a regular visiting speaker at the Civil Contingencies College at Easingwold, it's my privilege to meet many of the dedicated men and women working in this field - fire officers, police, security groups, emergency planners and civil contingencies teams from around the United Kingdom (and the world). Courses on safety at sports grounds, safety in built and complex environments, and our own crowd dynamics workshops focus on the issues of emergency planning. The delegates have only one question, "Is it safe?" and if not, how do we make it safer.

Dr G. Keith Still holds a PhD in interdisciplinary mathematics from Warwick University. He has originated a number of innovations in crowd modelling and other fields, and is is the founder of Crowd Dynamics Ltd, an international consultancy: www.CrowdDynamics.com